? ".^ SMITHSONIAN INSTITUTIONBUREAU OF AMERICAN ETHNOLOGYBULLETIN 174AN INTRODUCTION TO KANSASARCHEOLOGYBy WALDO R. WEDEL WITHDESCRIPTION OF THE SKELETAL REMAINS FROMDONIPHAN AND SCOTT COUNTIES, KANSASBy T. D. STEWART UNITED STATESGOVERNMENT PRINTING OFFICEWASHINGTON : 1959 For sale by the Superintendent of Documents, U. S. Government Printing 0?SceWashington 25, D. C. - Price f3 (paper cover) LETTER OF TRANSMITTAL Smithsonian Institution,Bureau of American Ethnology,Washington, D. (7., September 30, 1958.Sir: I have the honor to submit the accompanying manuscript,entitled "An Introduction to Kansas Archeology," by Waldo R.Wedel, with "Description of the Skeletal Remains from Doniphanand Scott Counties, Kansas," by T. D. Stewart, and to recommendthat it be published as a bulletin of the Bureau of American Ethnology.Very respectfully yours, Frank H. H. Roberts, Jr.,Director.Dr. Leonard Carmichabl,Secretary, SmitJisonian Institution. C^vc, CONTENTS PAGEForeword xiiiIntroduction 1The environmental setting 3The historical background 19Historic Indian tribes in Kansas 47Siouan tribes in Kansas 48The Kansa 50The Osage 54Caddoan tribes in Kansas 58The Pawnee -.. 58The Wichita 60Athapascan and Shoshonean peoples of western Kansas 68The Plains Apache 69The Comanche 75The Padouca 77Other tribes in western Kansas 78The Kiowa and Kiowa Apache 78The Arapaho and Cheyenne 80Summary 81Previous archeological work in Kansas 82Archeological investigations in northeastern Kansas 98The Doniphan site (14DP2) 98Cache pits 102House sites 105Burials 109Material from the house sites (Nebraska Aspect) 112Pottery 112Doniphan ware 112Chipped stone 116Ground stone 117Antler and bone 118Shell 118Material from cache pits and burials (Kansa?) 118Faunal remains 118Vegetal material 119Pottery 119Chipped stone 122Ground stone 122Bone and antler 124Glass, metal, and other trade materials 125Summary and comparisons 127Sites near Wolf Creek, Doniphan County 131The Fanning site (14DP1) 131House site 135Cache pits 136in ^?S-'?r1?S'jAN2 8 196aINSTITllTiON rv BUREAU OF AMERICAN ETHNOLOGY [Bull. 174 Archeological investigations in northeastern Kansas?ContinuedSites near Wolf Creek, Doniphan County?ContinuedThe Fanning site?Continued pageMiddens 138Vegetal remains 139Faunal remains 142Molluscan remains 143Pottery 143Fanning Plain 145Fanning Trailed 148Miniature Vessels 150Other sherds 150Perforated pottery disks 153Work in chipped stone 153Work in ground and pecked stone 157Work in bone and antler 162Metal and glass objects 166Summary and correlations 167Hilltop burial sites 172Ras Matherson site 172A. W. Guthrie site 174Miscellaneous surface finds in northeastern Kansas 175Sites near Manhattan, in Riley and Pottawatomie Counties 176The Griffing site (14RY21),Jn.Riley County 178House test 1 179House 1 179Other excavations 181Artifacts 182Pottery 183Riley Cord-roughened 183Work in chipped stone 185Work in ground stone 186Other stone objects 187Comment 187Kansa Village site (14P024), Pottawatomie County 187House 1 190Cache pits 191Burials 191Faunal remains 192Molluscan remains 192Vegetal remains 193Artifacts of native manufacture 193Objects of White manufacture 194General comments 197Other sites and materials in the Manhattan area 197Site 14P025 (Brous). 198Site 14P026 (Dike) 200Mound materials 203Fremont Point mound No. 1 203Fremont Point mound No. 2 _ 205Mound near Rocky Ford, Pottawatomie County 205General comments on the Manhattan locality 208 CONTENTS VPAGEArcheological investigations in central Kansas 210Sites on Little Arkansas River, Rice County - 210The Tobias site (14RC8) 211Mound 4 214Mound 6 215Mound 17 215Basin 1 219Basin 2 222General comments 226Food remains 231Pottery 233Geneseo Plain 233Genesee Simple Stamped 237Geneseo Red Filmed 239Exotic wares 245Other pottery objects 245Objects of antler 246Objects of bone 249Objects of chipped stone 264Objects of ground and pecked stone 276Unworked stone 290Objects of shell 291Perishable materials 294Objects of European manufacture 295The Thompson site (14RC9) 298Mound 1 300Mound 2 301Cache pits 302The artifacts 302Food remains 303Pottery 304Trade wares 308Clay pipe fragment 309Work in antler 310Objects of bone 311Objects of chipped stone 313Objects of ground and pecked stone; pigments 317Objects of European manufacture 319Other sites (Hayes, Major) 321Site on Cow Creek, Rice County 323Malone site (14RC5) 323Cache pits 32 6Food remains 329Pottery 330Objects of antler and bone 332Objects of chipped stone 335Objects of ground and pecked stone; pigments 337Objects of shell 339Objects of White manufacture 340Trait comparison of sites on Little Arkansas River and Cow Creek,Rice County 340 VI BUREAU OF AMERICAN ETHNOLOGY [Bull. 174Archeological investigations in central Kansas?Continued paqbSites on Walnut River, Cowley County 344Larcom-Haggardsite(14COl) 347Elliottsite (14C02) 349CountryClub site(14C03) 351Artifacts from the Cowley County sites 357Food remains 357Pottery 359Exotic wares 363Artifacts of bone 365Objects of chipped stone 367Objects of ground and pecked stone 373Unworked stone; pigments 375Objects of shell 376R6sum6 ; summary of artifacts found 377Archeological investigations in western Kansas 379Sites on Salt Creek, in Lane and Scott Counties. 379The Pottorff site (14LA1) 381Area A (Houses 1 and 2) 384Area B 388Area C 391Area D 394Other excavations 396Artifacts from Occupation A (Upper Republican Aspect) 397Food remains 397Pottery 398Objects of bone 399Objects of chipped stone 402Objects of stone other than chipped 405Objects of shell 407Artifacts from Occupation B (Woodland) 407Food remains 407Pottery 408Objects of bone and shell 410Objects of chipped stone 411Objects of ground stone 412Future investigations 412The Risston site (14SC4) 413Site on Walnut Creek, Lane County 416The Walter site (14LA2) 416Sites on Ladder Creek, Scott County 422 "El Cuartelejo" (14SC1) 424Test No. 1 426Test No. 2... 429Test No. 3 432Roasting pit (pit 3) 435Other excavations 435Irrigation ditches 437Later investigations 438Food remains 439Pottery _. 441Exotic sherds 444Pottery pipes - 445 CONTENTS VnArcheological investigations in western Kansas?ContinuedSites on Ladder Creek, Scott County?Continued "El Cuartelejo"?Continued pageObjects of antler and bone 448Objects of chipped stone 454Objects of ground and pecked stone 458Objects of unworked stone 460Objects of shell 461Objects of European manufacture 461Discussion and correlations 462Young burial site (14SC2) 468Surface finds near Shallow Water, Scott County 475Aboriginal chert quarries 476Cowley County quarries 476Chase County quarries 480Petroglyphs 482Other sites and localities 494Sites near Bazaar, Chase County (14CS1) 494Sites near Marion, Marion County 497Site 14PT1, Pratt County 503Site near Salina, Saline County (14SA1) 512Sites in Tuttle Creek Reservoir area on Blue River 523Sites near Neodesha, Wilson County 526Classification of sites _ 534Summary of cultures represented 536Preceramic materials 536The Woodland Complexes 542Late Woodland-Mississippi Complexes 557The Great Bend Aspect 571The Dismal River Aspect 589The Oneota Aspect 600Time perspective 615Interpretations - 620Conclusion 637Literature cited 642Appendix. The faunal lists from Kansas 662Description of the Skeletal Remains from Doniphan and Scott Counties,Kansas, by T. D. Stewart _ 669Explanation of plates 1-97 683Index 695 ILLUSTKATIONSPLATES(All plates follow page 694)1. Views at the Doniphan site, 14DP2.2. Burials 1-4 at Doniphan site, 14DP2, presumably Kansa.3. Burials, some slab covered, in situ at Doniphan site.4. Restored pottery vessels from house 2, Doniphan site.5. Shell, bone, and antler artifacts, Doniphan site.6. Catlinite and limestone artifacts, Doniphan site.7. Iron and brass artifacts, Doniphan site.8. Fanning site, 14DP1, house 1 and restored pottery vessel.9. Trailed and plaiuware potsherds and vessel handles. Fanning site.10. Trailed potsherds and miniature bowl. Fanning site.11. Rimsherds of Lower Loup Focus types, Fanning site.12. Miscellaneous artifacts from Fanning site.13. Brass and iron (6, c) objects. Fanning site.14. Burial mound excavations in Doniphan County.15. Surface finds of early-type projectile points from Doniphan and SewardCounties, Kans.16. Excavations near Manhattan.17. Riley Cord-roughened potsherds from GriflSng site.18. Miscellaneous objects from Kansa village site, 14P024.19. Miscellaneous iron and brass objects from Kansa village site, 14P024.20. Stone, bone, and shell objects from burial mounds near Manhattan.21. Tobias site (14RC8), Rice County.22. Excavations at Tobias site, Rice County.23. Excavations at Tobias site. Rice County.24. Excavations at Tobias site. Rice County.25. Miscellaneous features in mound 17 complex, Tobias site.26. Restored pottery vessels of type Geneseo Plain, from Tobias site. Rice County.27. Restored pottery vessels from Tobias site. Rice County.28. Restored pottery vessels, type Geneseo Simple Stamped, from Tobias site,Rice County.29. Restored pottery vessels from Tobias site. Rice County.30. Rimsherds, handles, and sherd disks from Tobias site. Rice County.31. Bone and antler implements from Tobias and Malone (c) sites. Rice County.32. Bone awls and polishing (?) tools from various Little River Focus sitesin Rice County.33. Bone projectile points and bipointed objects (o-t) from Tobias site,Rice County.34. Worked antler strips and fragments, and cut mammal bone fragment(l), from Tobias site, Rice County.35. Bone and antler artifacts from Tobias site. Rice County, and fragmentof bone paint applicator (h) from Larcom-Haggard site, CowleyCounty. CONTENTS rX 36. Bone and iron (k, I) objects from various Great Bend Aspect sites inRice and Cowley (&, c) Counties.37. Chipped projectile points from Tobias site, Rice County.38. Stemmed {b-g) and plain-shafted (h-m) chipped drills from Tobias site,Rice County.39. Expanded-base chipped drills from Tobias site, Rice County.40. Chipped knives from Tobias and Thompson {d, k, I) sites. Rice County.41. Top and side views of chipped end scrapers from Tobias site. Rice County.42. Grooved mauls and shaft polishers from various sites in Rice and CowleyCounties.43. Grinding stones from Tobias {a, d, f, g, h) and Thompson (&, c, e) sites,Rice County.44. Sandstone disks and abrading stones from Tobias and Malone {f,h,i) sites,Rice County.45. Stone pipes from Tobias site and vicinity, Rice County.46. Miscellaneous objects from Rice and Cowley (No. 456) Counties.47. Miscellaneous objects from Rice and Doniphan (d, g) Counties.48. Thompson site, 14RC9, Rice County.49. Pueblo and other exotic potsherds from various sites in Rice and CowleyCounties.50. Oxidized chain mail fragments, pit 4, Thompson site.51. Larcom-Haggard site, 14C01, Cowley County.52. Arkansas City Country Club site, 14C03, Cowley Coimty.53. Restored pottery vessels, type Cowley Plain, from various sites in CowleyCounty.54. Restored pottery vessels, type Cowley Plain, from Larcom-Haggard site.Collection of Louis Essex.55. Socketed scapula hoes from Cowley County sites.56. Bone and stone artifacts from Cowley and Lane County sites.57. Pottorff (14LA1) and Risston (14SC4) sites on Salt Creek, on Lane-ScottCoimty line.58. Houses 1 and 2 excavations, PottorfE site, 14LA1.59. Excavations at the Pottorff site, 14LA1.60. Pottery from Upper Republican and Woodland occupations, PottorfE site.61. Bone and shell objects, Pottorff site.62. Projectile points, drills, and end scrapers, Pottorff site.63. Knives and other chipped forms, Pottorff site.64. Stone and bone objects from Occupation B (Woodland), Pottorff site.65. Scott County State Park and site, 14SC1.66. Excavations at 14SC1, in Scott County State Park.67. Roasting or baking pit No. 3, Scott County State Park.68. Restored pottery vessels from 14SC1, Scott County.69. Pottery objects from Scott County site 14SC1. Actual size.70. Bone and shell artifacts from 14SC1, Scott County.71. Bone artifacts from 14SC1, Scott County.72. Projectile points, drills, and end scrapers from 14SC1, Scott County.73. Stone, iron (g), and pottery (h) objects from 14SC1, Scott County.74. Young burial site 14SC2, Scott County.75. Burials in situ, Young burial site, Scott County.76. Petroglyphs at Inscription Rock, 14EW1, Ellsworth County.77. Petroglyphs from various localities in Kansas.78. Petroglyph sites in Rice County.79. Anthropomorphic and other petroglyphs at Spriggs Rocks, Rice County. X BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174 80. Potsherds from the Roniger site, 14CS1, In Chase County.81. Potsherds from Pratt site, 14PT1.82. Chipped and ground stone objects from Pratt site, 14PT1.83. Bone and ground stone objects from Pratt site, 14PT1.84. Miscellaneous specimens from various sites in Kansas.85. Salina burial pit, 14SA1, Saline County.86. Salina burial pit, 14SA1, Saline County.87. Pottery vessels from Salina village and burial site, 14SA1.88. Pottery vessels from Salina village site, 14SA1.89. Pottery vessels from Salina burial pit, 14SA1.90. Pottery vessels from Salina burial pit, 14SA1.91. Pottery from Salina village and burial site, 14SA1.92. Objects of stone, pottery (i), antler {h), and shell from Salina village andburial site, 14SA1.93. Neodesha "fort" 14WN1, Wilson County.94. Four views of the Doniphan site skull No. 3.95. Four views of the Doniphan site skull No. 4.96. Four views of the Doniphan site skull No. 10.97. Four views of the Doniphan site skull No. 11.TEXT FIGURES PAGE1. Map showing location of certain archeological sites in Kansas 42. Historic Indian tribes in relation to physiographic divisions in Kansas. 63. The Central Plains, showing historic tribal locations, shifting rainfallaverages, and major grassland divisions 174. Map of principal known Kansa village sites during the 18th and 19thcenturies 505. Contour map of portion of Doniphan site, 14DP2, Doniphan County. _ 1036. Floor plans of houses 1 and 2, Doniphan site.. 1067. Sherds from Doniphan site 1158. Pottery pipes from Doniphan site 1169. Handle sherd of Lower Loup type, Doniphan site 12110. Stone club head from Doniphan site 12311. Bone or antler object from Doniphan site 12512. Lead cross from Doniphan site 12613. Contour map of Fanning site, 14DP1, Doniphan County 13314. Contour map of portion of Fannings ite, showing excavated features. 13415. Plan of house 1, Fanning site 13516. Vessel shapes from Fanning site 14417. Rim profiles from Fanning site 14618. Pottery handles from Fanning site 14719. Trailed potsherds from Fanning site 14820. Trailed and punctuate potsherds from Fanning site 14921. Pottery disks and miniature vessel, Fanning site 15122. Rimsherds of Lower Loup Focus type, Fanning site 15223. Chipped end scrapers from Fanning site 15624. Chipped stone artifacts from Fanning site 15825. Sandstone sharpening block from Fanning site 15926. Notched "sinew" stone from Fanning site 16027. Map of portion of Griffing site, 14RY21, Riley County (facing) 17828. Plan of house 1, Griffing site 181 CONTENTS XIPAGE29. Map of Kansa village site, 14P024, near Manhattan 18930. Floor plan of Kansa earth-lodge site, house 1, 14P024 19031. Artifacts from Brous site, 14P025, Pottawatomie County 19932. Artifacts from Dike (?) site, 14P026, Pottawatomie County 20133. Miscellaneous artifacts from the vicinity of Manhattan 20934. Map to show location of certain archeological sites in central Kansas. _ 21135. Map showing location of certain archeological sites in and near RiceCounty 21236. Sketch map of Tobias site, 14RC8, Rice County 21337. Plan of mound 17 and associated features, Tobias site 21638. Plan of excavations in basins 1 and 2, Tobias site 21939. Little River Focus vessel shapes from Tobias site 23640. Handle sherds from Tobias site 23841. Rimsherds from Tobias site 24042. Miniature pottery vessels, Tobias site 24243. Trailed sherds from Tobias site 24444. Molded clay lump from Tobias site 24645. Hollowed antler artifacts from Tobias site 24746. Bone artifacts from Tobias site 25347. Bone awls and polishers (?), Tobias site 25448. Bison epiphyseal hide-rubbing tool, Tobias site 26149. Drills from Tobias site 26750. Chipped knives, beveled and notched, from Tobias site 27151. Chipped knives, beveled and unbeveled, from Tobias site 27252. Stemmed knives from Tobias site 27253. Chipped point and beveled knives from Tobias site 27354. Side scrapers from Tobias site 27455. Mano from Tobias site 28156. Shaft polishers from Tobias and Paint Creek sites 28457. Sandstone sharpening blocks, Tobias site 28658. Notched and incised shell, Tobias site 29259. Sketch map of Thompson site, 14RC9, Rice County 29960. Rim and handle sherds, Thompson site 30661. Incised pottery pipe fragment, Thompson site 31062. Loaf-shaped clay object, Thompson site 31063. Chipped stone artifacts, Thompson site 31664. Sketch map of part of Malone site, 14RC5, Rice County 32765. Chipped stone artifacts from Malone site 33666. Sandstone arrowshaft-smoothers, Malone site 33967. Map of Arkansas City and nearby archeological sites 34768. Plan and cross section of mound 1, Arkansas City Country Club site,14C03, Cowley County 35469. Vessel shapes, type Cowley Plain, from Cowley County sites 36070. Pottery handles, type Cowley Plain, Cowley County sites 36271. Trailed and punctate sherds, Cowley County sites 36472. Chipped-stone artifacts, Cowley County sites 36973. Beveled and unbeveled knives, Cowley County sites 37074. Chipped scrapers and ax, Cowley County sites 37275. Map showing location of certain archeological sites in western Kansas. 38076. Plan of houses 1 and 2, Pottorfl site, 14LA1, Lane County 38477. Artifacts from Occupation A (Upper Republican), Pottorflf site 40078. Plan of Scott County pueblo ruin 425 Xn BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174PAGE79. Map of excavations at Scott County State Park, 14SC1 42780. Cross sections of roasting pits, 14SC1 43481. Miniature pottery vessel, 14SC1 44382. Pottery pipe fragments from 14SC1 44783. Worked and polished antler tip, 14SC1 44984. Rib implements from 14SC1 45285. Chipped drills from 14SC1 45686. Mano from 14SC1- 45887. Chipped artifacts from Young burial site, 14SC2, Scott County 47188. Petroglyphs at Inscription Rock, Ellsworth County 48489. Petroglyphs from localities 2, 3, and 4, central Kansas 48790. Petroglyphs at Spriggs Rocks, 14RC1, Rice County 48991. Petroglyphs from locality 6, 14RC10, Rice County 49092. Petroglyphs from locality 7, 14RC11, Rice County 49193. Pit-and-line markings near Arkansas City 49394. Dentate-stamped sherd from vicinity of Marion 50295. Crockett Curvilinear Incised sherd and restoration, Salina burial pit,14SA1 ----- 52096. Fluted point and fragments, from Doniphan and Seward Counties. _ 53797. Heavy chipped-stone artifacts from Kansas River valley 53998. Chipped-stone artifacts and corner tang knife, Kansas River valley. _ 54099. Chipped-stone artifacts, Kansas River valley 541100. Miscellaneous sherds from Wyandotte County 543101. Dentate-stamped and other sherds, Trowbridge site, 14WY1,Wyandotte County 545102. Miscellaneous sherds and pottery artifacts, Trowbridge site, 14WY1,Wyandotte County 546103. Bone, stone, and antler artifacts, Trowbridge site, 14WY1, Wyan-dotte County 548104. Miscellaneous specimens from eastern Kansas 557105. Map showing location of certain Central and Southern Plains archeo-logical sites and complexes 573106. Suggested chronology of certain archeological sites and complexes inKansas 613107. Four stereographic views of the Young site skull No. 3 674108. Four stereographic views of the Young site skull No. 5 675109. Simplified stereographic views of the Lansing skull and the Youngsite skull No. 5 682 FOREWORDThis study began as a routine report on three seasons of archeologi-cal fieldwork in Kansas, conducted by the writer for the United StatesNational Museum in 1937, 1939, and 1940. The fieldwork was essen-tially a survey; it involved some surface collecting and the limitedsampling of certain selected sites, but no comprehensive excavations.The sampling was designed to test the archeological possibilities invarious sections of the State, with the possibility in mind of futureextended investigations if and where warranted.As with so many other programs of the time, completion of this onewas made impossible during the war years. Both fieldwork and writ-ing came to a halt, the latter to be resumed in 1945. Shortly camefurther interruption when the Smithsonian Institution entered intoan agreement with the National Park Service to participate in thearcheological salvage work linked with the post-war Federal water-control program. The writer was detailed by the National Museum forfour summers and intermittently otherwise to the newly organizedRiver Basin Surveys and was assigned the task of organizing andsetting in motion the salvage program in the Missouri Basin. Notuntil 1950 was there again opportunity for resuming the Kansasreport.During the long delay that followed completion of the fieldwork,the conviction grew that a routine report was no longer adequate.Many of the archeological materials recovered clearly stood in closerelationship to materials on various time levels in adjacent States.Some of them as clearly called for further researches in the accumulat-ing literature bearing on the ethnohistory of the Central Plains.Finally, it was abundantly evident that much of the previous work inKansas, though carried on intermittently and often sketchily reported,had definite bearing on problems raised in our work from 1937 to 1940,and that this should be fully utilized wherever possible.Out of these and other considerations finally emerged the idea ofexpanding the report to include a comprehensive review of the avail-able ethnohistorical, archeological, and geographical data bearing onthe aboriginal occupancy of Kansas. It is the writer's hope that inclu-sion thus of a considerable body of ancillary data, much of it fromexceedingly obscure sources, will simplify the labors of future workersin the area, besides adding justification to this study as an introductionto the archeology of Kansas. XIII XIV BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174The list of individuals to whom I owe a debt of gratitude is a longone. For their ready endorsement of my initial proposal in 1937 andfor thus making possible the fieldwork on a continuing basis, I wishto thank especially the following: Dr. A. Wetmore, then assistantsecretary in charge of the United States National Museum; F. M.Setzler, head curator of anthropology ; and my immediate administra-tive superior, Neil M. Judd, curator of archeology. Thanks are alsodue Dr. Eemington Kellogg, director of the National Museum, forapproval of subsequent shorter trips to Kansas for follow-up work onvarious pertinent problems.Much of the success which attended our survey efforts, consummatedin the field for less than $4,000 dollars, must be attributed to thesplendid cooperation I received from the various members of the fieldparties. Few in number and sadly undei-paid by present standards,these men uncomplainingly worked long hours?digging by day, cata-loging at night?and accepted cheerfully the usual inconveniences ofcamp life, of recurrent dust storms, and of frequent changes of loca-tion. To Marvin F. Kivett, my field assistant during all three seasons,I am particularly indebted. Of him I should like to say that in a quar-ter century of activity in Plains archeology, I have known no workerwho has shown better field teclmique, keener insight, higher personaland professional integrity, or deeper comprehension of the task athand. In addition to Kivett, the 1937 field party included KennethG. Orr, E. G. Slattery, and Hugh Stabler. Slattery was a member ofthe 1939 party, which also included Karl Schmitt, J. M. Shippee,Henry Hornblower II, and W. B. Oswald. Shippee and Slatterywere again members of the party in 1940, which included also JohnGiles as cook. Philip Drucker was a welcome addition to the 1939group for two weeks; and Louis Essex volunteered his services forseveral days during the work in Cowley County. To any of theabove who may read these lines, I send greetings and thanks for ajob well done.To the following property owners, lessees, tenants, administrators,business firms, and others, I am grateful for excavation privileges onone or another of the various sites investigated : George Meidinger,Pete Leiffering, J. O. Corcoran, Ed Haid, Henry Otto, Kansas Ever-green Nurseries, Ras Matherson, A. W. Guthrie, B. E. Hale, DaveLeahy, Jr., Howard Young, C. D. Pottorff, James Risston, E. W. Wal-ter, Ed Tobias, C. F. Thompson, Mr. and Mrs. John Malone, HomerPeverley, John Goff, Alfred Larcom, Ernest Muret, T. N. Haggard,and John Boggs and other officials of the Arkansas City Country Club.With gratifyingly few exceptions, the above cheerfully extended per-mission for us to work when we called on them, and in but two instanceswere restrictions placed on the extent of digging we could do. FOREWORD XVAppreciation must also be expressed to numerous individuals whohelped in various other ways. Some of these gave access to importantcollections of local material ; others acted as guides to sites that mightotherwise have escaped our notice, or functioned as intermediariesin our negotiations with reluctant landowners. In northeastern Kan-sas, Fenn Ward of Highland was most helpful, as he has been throughcorrespondence since the close of our fieldwork. Mr. H. M. Trowbridgeof Bethel made available for examination an important Hopewelliancollection that originated near his home in Wyandotte County. Incentral Kansas, Mr. and Mrs. G. L. Whiteford of Salina providedmany helpful suggestions, supplied numerous photographs, and per-mitted study of an exceptionally important collection of prehistoricmaterials from a large burial ground and village site investigated bythem. Horace Jones, newspaperman and enthusiastic historian ofLyons, made our work in Eice County immeasurably easier and vastlymore productive than it otherwise would have been ; and through thecolumns of his paper, he enabled us to locate and examine many pueb-loan potsherds from local sites that otherwise would not have come toour attention. In southern Kansas, Bert Moore of Winfield was a fre-quent visitor to our excavations, guided us to other sites in Kansas andOklahoma, and assisted unendingly with local contacts. Others whomerit special notice here include Dr. Norman L. Roberts, Manhattan ; Stanley Dienst, Coffeyville; Thomas M. Galey, Wichita; B. F. Mc-Daniel, Dexter ; Mr, and Mrs. J. E. Harclerode, Ottawa ; A. W. andEarl Monger, Larned; George C. Estep, Garden City; William O.Leuty, Ellsworth; Phil Hohl, Bushton; Lowell Peverley, Geneseo;and Eugene Wing, Pratt. The extent and nature of my indebtednessto these individuals will be evident in other sections of this report.To a host of professional colleagues, anthropological and other,in the Smithsonian and outside, I am heavily indebted for expertopinions, identifications, and numerous helpful suggestions as thisstudy progressed. Among present and former members of the Na-tional Museum staff, the following should be mentioned in connectionwith specified identifications: A. Wetmore and Herbert Friedmann,bird bones; R. Kellogg, D. H. Johnson, H. M. Setzer, and Ray Gil-more, mammal remains; C. V. Morton and A. C. Smith, plant re-mains; J. P. E. Morrison, moUuscan remains; E. P. Henderson,minerals; G. A. Cooper, invertebrate fossils; C. M. Watkins andM. L. Peterson, early White trade and contact materials; T. D.Stewart and M. T. Newman, human skeletal materials ; John C. Ewers,ethnoarcheological items; D. M. Cochran, reptile bones; E. D. Reidand L. P. Schultz, fish remains. Outside the Museum staff, E. A.Goldman and S. P. Young, Fish and Wildlife Service, supplied im-portant observations on dog remains. R. W. Brown and L. G. Henbest, XVI BUREAU OF AMERICAN ETHNOLOGY [Boll. 174U. S. Geological Survey, identified certain paleobotanical and min-eralogic items. F. H. H. Roberts, Jr., Bureau of American Ethnology,kindly made available unpublished and restricted materials collectedin Kansas by the River Basin Surveys and not yet reported in print.Mrs. M. C. Blaker, archivist of the Bureau of American Ethnology,patiently tracked down for me various manuscripts in her files.From outside the Washington area, invaluable assistance was pro-vided by the following, among others : H. P. Mera, Stanley Stubbs,A. V. Kidder, and Marjorie F. Tichy, on matters involving Plains-Southwestern relationships ; A. M. Withers, on eastern Colorado ma-terials; A. D. Krieger and R. E. Bell, southern Plains materials;Carl H. Chapman, Missouri materials ; J. O. Brew, Carlyle S. Smith,and Roscoe Wilmeth, Kansas materials ; Marvin F. Kivett, Nebraskamaterials ; Volney H, Jones, ethnobotanical items ; Arthur Woodward,trade beads; and Stephen V. Grancsay, identification of chainmailfragments. J. H. Gunnerson generously granted me permission toread an unpublished manuscript awaiting publication by the Bureauof American Ethnology. The University of Nebraska Library loanedfour manuscript theses prepared in the Department of Anthropology.Dr. Walter H. Schoewe, of the State Geological Survey, Universityof Kansas, kindly provided the base maps used in figures 1 and 2.Nyle H. Miller, Kansas State Historical Society, supplied helpfulinformation on various historical matters.With many of those named in the preceding paragraphs, I havehad fruitful discussions that materially clarified points about whichI was in doubt, and for this I am most grateful. In the final stagesof the manuscript, I have had the benefit of helpful criticism fromMr. George Metcalf, aid in the Division of Archeology, who cheer-fully read long sections and offered his comments and views. R. G.Paine, formerly aid in the Division of Archeology, drafted severalof the site and other maps included in this report. All drawings ofspecimens are the work of Betty E. Baker, scientific illustrator inthe Department of Anthropology. The arduous task of typing themanuscript was accomplished at various times by three secretariesin the Division : Mrs. Leta B. Loos, Miss Lucy H. Rowland, and Mrs.Jeraldine M. Whitmore.I cannot fail to acknowledge here my heavy debt to one of thepioneers in Plains archeology, the late A. T. Hill of Lincoln, Nebr.,and to the professional who afforded me my first opportunity to workin the Plains, Dr. William Duncan Strong. Mr. Hill took a deeppersonal interest in the formulation of my field program, visited ourcamp at a number of the sites, and contributed many helpful sugges-tions for guidance of the work, especially in western Kansas. ToStrong and to Hill, I owe in large part my initial indoctrination into FOREWORD XVIIthe field of Plains archeology; and I trust the present study willreflect no discredit on our associations beginning more than a quarterof a century ago.Finally, to my patient and forbearing wife, Mildred Mott Wedel,go my sincere and affectionate thanks for urging the study on its waywhen growing curatorial responsibilities seemed overdemanding andthe road ahead looked all but impassable. I am indebted to herespecially for numerous suggestions in the section on ethnohistory,as also in that reviewing the archeological complexes of eastern Kansas.She has made available much mipublished information on Oneotamaterials, and has read long sections of manuscript and proofs. Thereader, no less than I, will be grateful to her for the index, whichshe painstakingly compiled. Here, as elsewhere, I gratefully acknowl-edge the assistance received, regret the possible omission of deservingnames and credits, and cheerfully assume responsibility for any misuseI may have made of the assistance so readily extended. 484172?5?- AN INTRODUCTION TO KANSAS ARCHEOLOOY By Waldo R. WedelINTRODUCTIONWlien I was invited, shortly after joining the National Museum staffin 1936, to submit a proposal for a field project for the following sum-mer, my immediate selection of Kansas as the locus of operations wasbased on several considerations. For one thing, probably no State inthe Great Plains region was so inadequately known, ethnologically orarcheologically. Most of the tribes?Siouan, Caddoan, and other ? known to have inhabited the State in historic times prior to reservationdays had passed from the scene when trained etlmologists began theirstudies. No ethnohistorical work of any note had been undertaken.As to archeology, notable contributions had been made here and thereby a handful of individuals, but mostly before 1910. No generalsystematic attack had been made on the prehistory of the region norhad anyone attempted to place the State in its proper position in theemerging picture of Plains prehistory. No local institution or organ-ization was engaged in any sort of anthropological research within theState ; and the large-scale archeological work relief operations of the1930's had no counterpart in Kansas.Yet there was ample reason to believe that the notion, still lingeringin some quarters, of the Great Plains as a region "largely bare ofarcheology," was no more applicable to Kansas than it was to Ne-braska. As Strong pointed out in his classic introduction to Nebraskaarcheology, perusal of the limited literature available showed thatKansas was "an extremely promising archeological field that up to thepresent has received nothing even approaching systematic investiga-tion." Having participated in five productive and stimulating seasonsof fieldwork in southern and western Nebraska from 1930 to 1934, Iwas entirely convinced of the accuracy of this trenchant observation.Realization thus of the possibilities, grafted onto lingering memoriesof fruitful boyhood "relic"-hunting trips in central Kansas, providedan initial stimulus for laimching sustained field investigations whenthe opportunity at last arrived. 1 2 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174There were other considerations. The accumulation of systematic-ally gathered information on the archeology of neighboring Stateswas imparting new insights into the background and antecedents ofhistoric Plains Indian culture. Lack of equally sound and up-to-dateinformation on Kansas was increasingly troublesome to students con-cerned with the wider relationships of the prehistoric Plains peoples.A significant proportion of the archeological resources of the region,long recognized by students as "an important laboratory for anthro-pological research" was thus unavailable for scientific examination.Inferences were possible, of course, but these were manifestly unsatis-factory. After all, there was a long record of "inferences" that thePlains had no worthwhile archeology, if indeed they were inhabitablebefore adoption of the horse by the Indian.The geographic location of Kansas, straddling the shifting border-line between the humid East and the arid West, was another factor.In the middle 1930's, much of the Great Plains was in the tighteninggrip of a record-breaking drought. Widespread and spectacular duststorms, recurrent crop failures, and extensive dislocation of the humaneconomy could be observed on every hand. Without subscribing toany theory of environmental determinism, it still seemed pertinent toinquire whether the archeological record might not show that theearlier prewhite populations were in some measure likewise influencedby the vagaries of the regional environment.The primary objectives of the National Museum survey of Kansas,then, may be summarized as follows : (1) To sample the archeological materials in various sections of the State,with a view to determining their general nature and their variability throughtime and space ; (2) To identify wherever possible the remains of historic tribal groups, suchas the Kansa, and thus to obtain some insights into their antecedents ; (3) To assess the materials found with respect to their similarities to, anddifferences from, the archeological complexes already recognized in adjacentStates, such as Nebraska, where geographical conditions were broadly similar;(4) To gather information bearing ou the manner in which the native popula-tions, at various periods and in various regions, adjusted their way of life totheir natural environment ; (5) To integrate these findings, tentative though they might be, into thelarger picture of native man's occupation of the Great Plains ; (6) To suggest problems that might profitably be attacked next.The core of this study is, as previously suggested, the data and ma-terials gathered in three seasons of fieldwork. This consisted bas-ically of 15 to 18 weeks annually of survey and limited excavationduring the summers of 1937, 1939, and 1940. A few of the localitieswere revisited in subsequent years, but visits were incidental to otheractivities and without opportunity or means for further excavation.In 1954, I spent approximately 10 days in brief observations at sitesand localities not previously worked, a prime consideration then being Wbdel] an mTRODUCTION TO KANSAS ARCHEOLOGY 3the effects of intense and prolonged drought on surface water sup-plies, natural vegetation, and the corn crop. All of these field obser-vations, however brief, as well as the incidental opportunities to viewadditional local collections, have been drawn on for this paper.In the sections which follow, I have first reviewed at some lengththe environmental setting, the ethnohistorical background, and thepublished record of previous archeological work in Kansas. Detailedpresentation of results of the National Museum's fieldwork is fol-lowed by a short section classifying the archeological data in terms ofthe larger complexes recognizable. Another section endeavors to sum-marize in some detail the nature, distribution, and wider relation-ships of these larger complexes, and this is followed by a discussionof the chronological situation. The final sections offer a tentative re-construction of Kansas prehistoiy, and a review of the achievementsof the survey in terms of the objectives listed above.THE ENVIRONMENTAL SETTINGTo the traveler from the wooded lands of the Eastern UnitedStates, journeying through Kansas along any of the great transcon-tinental thoroughfares, the general impression is likely to be one ofmonotonous sameness of landscape. It is a typical Plains State?anarea of low relief, with no mountains, few trees, and little surfacewater. The regional variations that occur from east to west are notusually very striking or abrupt, and easily elude the casual observerwho stays on the main highways and rail routes. To those who in-quire more closely into the details of topography, native flora andfauna, climate, and other aspects of the environment, it soon becomesapparent, however, that there are natural variations which bear,directly or indirectly, on man's utilization of the land.^Kansas is an approximate parallelogram of some 82,200 square mileslying in the heart of the United States. Its northeastern boundary for100 miles is the winding trench of the Missouri River; the southeasterncomer touches the Ozark plateaus. Thence it stretches westwardmore than 400 miles between the 37th and 40th parallels of northlatitude far into the Great Plains. Its surface rises gradually andirregularly from less than 700 feet above sea level near the southeast-ern comer on the Oklahoma State line to just over 4,100 feet on theColorado line near the northwest corner. This surface has been de-veloped on a generally simple yet varied geologic structure, and so isby no means uniformly flat and featureless. Down its long unevenslope, streams have furrowed deep wide valleys through which the * The descriptive section which follows is based largely on Adams. 1902 ; Cockrum, 1952 ; Fenneman, 1931; Flora, 1948; Gates, 1937 and 1940; Johnson, 1900; Parker, 1911;Schoewe, 1949 and 1951 ; Smith, 1940 ; and Snow, 1872. 4 BUREAU OF AMERICAN ETHNOLOGY [BULL. 174 ^^ c5 q; Wbdbl] AN INTRODUCTION TO KANSAS ARCHEOLOGY north half of the State drains eastward via the Smoky Hill-KansasRiver system and the Osage into the Missouri, and the south half byway of the Arkansas into the Mississippi. Along these valleys, thinwinding ribbons of hardwood timber run westward from the oak-hickory forests of the central Mississippi valley to interfinger withthe grassy uplands of the Great Plains. Thus, because of its geo-graphic location and its considerable east-west extent, what is essen-tially a Plains State exhibits a diversity of terrain, climate, soils,native flora, and fauna greater than is commonly supposed. (Fig. 1.)In terms of the major physical divisions of the United States,Kansas lies almost wholly within the Interior Plains (Fenneman,1931 ; Schoewe, 1949) . The exception is a small gently sloping area ofsome 50 square miles in the extreme southeastern corner of the State,bounded on the west by Spring River, which is assigned to the OzarkPlateau. As manifested in Kansas, the Interior Plains comprise twodivisions or provinces, both of which extend far beyond the borders ofthe State (fig. 2) . These are the Central Lowland, including roughlythe eastern third of Kansas, and the Great Plains, to which the remain-ing two-thirds are assigned. The Central Lowland in northeasternKansas, which was subjected to glacial activity, is characterized bysubmaturely to maturely dissected till plains ; south of Kansas River FiGUEB 1.?Map showing location of certain archeological sites in Kansas.Numbers on map refer to sites in text, as follows:1?14P024, Kansa site, ca. 1800-30,Pottawatomie Co.2?14RP1, Pawnee site, pre-1800;Republic Co.3?14DP2, Doniphan site; DonipanCo.4?14DP1, Fanning site, DoniphanCo.5?14JW1, Warne site, Jewell Co.6?14ML1, Glen Elder site, MitchellCo.7?14RC8, Tobias site. Rice Co.8?14RC9, Thompson site. Rice Co.9?14RC3, Hayes site, Rice Co.10?14RC2, Major site. Rice Co.11?14RC5, Malone site. Rice Co.12?14MP1, Paint Creek site, McPher-son Co.13?14C01, Larcom - Haggard site,Cowley Co.14?14C02 (Elliott) and 14C03 (Ar-kansas City Country Club), Cow-ley Co. 15?14C05, Maple City quarries, Cow-ley Co.16?14SC1, "El Cuartelejo," Scott Co.17?14PT1, Pratt site, Pratt Co.18?14WN1, Neodesha "fort," WilsonCo.19?14LA1, Pottorff site. Lane Co.20?14LA2, Walter site, Lane Co.21?14RY21, Griffing site, Riley Co.22?140T5, Minneapolis site, OttawaCo.23?14SA1, Whiteford site. Saline Co.24?14CS1, Roniger site, Chase Co.25?14PH4, Woodruff ossuary, Phil-lips Co.2&?14SC2, Young burial site, ScottCo.27?14WT1, Trowbridge site, Wyan-dotte Co.28?Younkin mound, Geary Co.29?Lansing Man, Leavenworth Co.30?Twelve-mile Creek, Logan Co.31?14P026, Dike site, PottawatomieCo. BUREAU OF AMERICAN ETHNOLOGY [BULL. 174 U PhP-,SO -1^ ffi WEDEL] AN INTRODUCTION TO KANSAS ARCHEOLOGY 7 are unglaciated scarped plains beveling faintly inclined rock stratawith the main streams entrenched. The western limit of the CentralLowland corresponds approximately to a line running irregularlysouth and west from northeastern Washington County to southwesternHarper County. "West of this and north of the Arkansas River valleyis a rough maturely to submaturely dissected plateau and Plains coun-try ; and another heavily eroded tract known as the Red Hills appearssouth of the stream. Between these two, in and around the northwardbend of the Arkansas, is an area of much lower relief termed the GreatBend Lowland. All of this was assigned by Fenneman to his PlainsBorder division ; but Schoewe and others have suggested inclusion ofthe Great Bend Lowland with the Central Lowland province and des-ignation of the broken areas to north and south as the Dissected HighPlains. In these latter areas, the east- and south-flowing streams havecut deeply into the edge of the uplands and thus by headward erosionare steadily widening the lowlands on the east at the expense of theHigh Plains. West of this zone of active erosion are the High Plainsproper, comprising somewhat more than a third of the State. Thesewere long thought to be the broad interstream remnants of smooth fluv-iatile plains that in Tertiary times extended from the Rocky Mountainseastward far across Kansas and Nebraska. Recent studies have indi-cated, however, that "the area has been extensively modified by severalcycles of erosion and deposition since Pleistocene time, and it seemsclear that most of the present High Plains surface was shaped in latePleistocene and Recent times" (Frye, 1946, p. 73) . A treeless countrycarpeted with short grass sod, having a semiarid climate and littlesurface water, this region has long influenced the popular conception ofthe landscape of the trans-Missouri West generally.To visualize more clearly the opportunities offered, and the limita-tions imposed, by the environment here to human occupancy, it will benecessary to examine more closely the region briefly characterizedabove. Several considerations should be borne in mind.For one thing, eyewitness accounts of the Kansas landscape go back,at most, some four centuries ; and for more than half this period, theyare sketchy in the extreme. The archeological record, on the otherhand, indicates a span of human existence covering several thousandyears. For most of this span, the environmental setting must be in-ferred from geologic, paleontologic, pedologic, and other evidence.Secondly, Kansas occupies a transitional zone climatically intermedi-ate between the humid East and the arid West. Like all such inter-mediate regions, it has long been subject to climatic fluctuations ofgreater or lesser magnitude. That some of these may have affectedman in prehistoric times, as in historic, is recognized ; but the degreeand duration of these influences and their effects on population move-ments are still obscure. Thirdly, from geologic and paleontologic 8 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174 studies (Frye and Leonard, 1952) it seems clear that the grassy plainsof Kansas are not a recent development, much less do they owe theirvegetational and faunal peculiarities to human activity primarily.Buried or "fossil" soils correlated with all interglacials are identifiedas grassland soils, indicating that throughout Pleistocene times thevegetational cover has been grass rather than trees; and moUuscanfossils point to a climatic control like the present one since Bradyan(mid-Wisconsin) times. As a part of the greater central NorthAmerican grassland, the Kansas landscape has a geologic and climato-logic basis ; and its vegetative and faunal characteristics long antedateman's demonstrable presence on the scene. (See also Wedel, 1957.)There is, in short, no present evidence that the human activities ofthe past few thousand years have been carried on in an environmentalsetting significantly different from the present, as for example, in awoodland as contrasted to a grassland with tree-fringed watercourses.The High Plains include some 30,000 square miles in westernKansas (Johnson, 1900; Schoewe, 1949, pp. 311-330). They arecharacterized by broad monotonously flat uplands and poorly de-veloped surface drainage. The horizon line almost everywhere is oneof phenomenal flatness and uniformity. The uplands are dotted withshallow saucerlike depressions ranging in diameter from a few toseveral thousand feet, and these may contain water for days or evenweeks after heavy rains. Larger depressions, such as Scott Basinand Shallow Water Basin south of Scott City, receive the runoff fromintermittent streams and here, too, ephemeral lakes sometimes formbefore the waters sink into the ground. Circular sinkholes with steepto precipitous walls, partially filled with water and sometimes of con-siderable size, result from the collapse of solution caverns formed byunderground waters. All these, however, are comparatively minorphenomena and do not materially affect the general impression offlatness. Relief is greater along the stream valleys. The Smoky Hill,Arkansas, and Cimarron Rivers traverse the region in bluff-linedvalleys; and along the first of these, erosion of the chalk beds hasproduced badlands areas and occasional buttes, pinnacles, and otherpicturesque features. Where the Smoky Hill and its usually shorttributaries have cut through the unconsolidated Tertiary silts, sands,and gravels into the impervious underlying rock formations, seepsand permanent springs occur and good water may be found in limitedquantity. North of the Smoky Hill, the Saline and Solomon Rivershave their sources in these plains, as do three creeks?Prairie Dog,Sappa, and Beaver?which run northeast through attractive terrace-lined valleys (cf. Fremont, 1845, p. 109) to empty into the RepublicanRiver just north of the Kansas-Nebraska State line. All of these aredry or intermittent in their upper reaches, and have wide shallowvalleys; they develop a permanent flow as they approach the 100th Wbdbl] an introduction TO KANSAS ARCHEOLOGY 9 meridian, where their valleys have been deepened to the contact be-tween Tertiary sediments and the older rocks. The Arkansas andCimarron Rivers, rising in mountainous regions outside the State, maybe classed as permanent streams; but their flow usually diminishesmarkedly toward the east and often sinks into the thick beds of sandthrough which they rmi. Sand dunes, both stabilized and active,border the south side of the Arkansas from the Colorado line eastwardto Dodge City or beyond, and also cover a large tract south of theCimarron.The High Plains are the short-grass country. The uplands, wherenot broken out for cultivation, are dominated by buffalo and gramagrasses, both of which are highly resistant to drought and close graz-ing and, in addition, cure well in the dry fall to produce palatableand highly nutritious winter forage. Yucca, prickly pear cactus, andsage are locally abundant, especially along the rougher valley mar-gins; in sandy localities, they have replaced the original bluestemgrass destroyed by drought and overgrazing. Where seepage andsprings result in live water, the creek banks have scattering standsof hackberry, cottonwood, and willow, with occasionally some elm,green ash, and box elder.^ Around the springs and in shaded rocky * The character of the tree growth In the western plains has been discussed at somelength by Kellogg (1905). With reference to western Kansas, he makes the followingobservations (Kellogg, 1905, pp. 23-24) : "The valley type of forest is confined entirelyto the valleys of watercourses and draws or their adjacent bluffs, and consists wholly ofbroadleaf species, with the exception of a few individuals or of small areas of red cedar.No body of timber of this type Is known to grow on the uplands of western Kansas andNebraska, though scattering trees are often found in the heads of draws almost to theupland level. In the valleys there are occasionally areas of a few acres in extent on whichtrue forest conditions prevail, but the general form is that of a narrow belt of trees a fewrods in width, holding closely to the banks of a stream or the bottom of a draw. . . ."The trees generally are low, with spreading crowns, though dense stands sometimesoccur in moist situations which produce good, clear trunks. Cottonwoods may attain aheight of 75 feet, but the maximum height of the other species is In the neighborhood of50 feet, while many mature trees do not grow even that high. Diameters, too, are mostlysmall, except in the case of cottonwood, which frequently attains a diameter of 3 to 4 feet,and sometimes over 5 feet. A green ash at Hays, Kans., is 44 inches in diameter 2 feetabove ground, and a white elm on Eagle Creek, near Paradise, is 62 inches in diameterat breastheight, but these are exceptional cases. Aside from the cottonwoods, the averagediameter of the broadleaf species of western Kansas and Nebraska is probably not far from6 inches."The Cimarron, Arkansas, Smoky Hill, Republican, and Platte rivers are preeminentlysandy streams with shifting beds, so that timber growth is either wholly absent from themover long stretches or consists only of scattering cottonwoods and willows. As previouslymentioned, the smaller streams which are less sandy support other species, principallygreen ash, white elm, hackberry, and box elder."In Kansas, as far west as the 100th meridian, nearly every stream has its fringe oftimber, and there is considerable diversity in species ; but both the quantity and varietydecrease westward, until at the Colorado line timber is very scarce, and occurs mainly onthe Smoky Hill and Arkansas rivers. Along the tributaries of the former stream. InRussell County, near the center of the State, there are found cottonwood, willow, red andwhite elm, black walnut, burr oak, hackberry, box elder, green ash, red cedar, red mulberry,and wild china ; while in Wallace County, on the western border, there are only a fewgroves of cottonwood and willow. The best westward extension of timber in Kansas is onBeaver Creek, in Rawlins County. The creek is bordered by a thriving growth of greenash, with a lesser number of several other species. The timber reaches to some distanceabove Atwood, not far from the point where the bed of the stream first carries permanentwater." 10 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174places, poison ivy flourishes. Thickets of wild plum, along withsome elderberry, chokecherry, and other shrubs bearing edible fruitsgrow in the ravines and along the creeks. The native fauna, ifsomewhat less varied than farther east, was nevertheless formerlyabundant. The grassy uplands supported large numbers of bisonand pronghorn antelope, and these in turn were preyed upon byprairie wolves. Present also were white-tailed and mule deer. Plainsgrizzly, coyote, swift fox, black-footed ferret, badger, prairie dog,white-tailed and black-tailed jackrabbit, western cottontail, and num-erous smaller rodents. Beaver, raccoon, and apparently elk, occurred.Bird life included a variety of small ground-nesting species, alongwith such larger forms as bald eagle, raven, crow, various hawksand owls, sharp-tailed grouse, and prairie chicken. Ducks and geesemay have been obtainable from time to time as seasonal migrants.The coyote and many of the smaller mammals can still be foundhere; the larger carnivores and the ruminants have long since beenexterminated.In terms of human occupancy, the High Plains are a region oflimited possibilities. They have a low and variable precipitation,averaging between 15 and 22 inches annually. There is an abund-ance of sunshine, along with strong southerly (and therefore, dry-ing) summer winds and a high rate of evaporation. Soils are gen-erally light and fine-textured, and tend to blow readily when thevegetative cover is destroyed. West of the 100th meridian, grain-farming on the uplands is likely to be precarious if water is notavailable to supplement the rainfall in dry years. Much of theregion is given over to stock-raising. The population is much sparserthan in regions to the east. Ethnohistorical and archeological evi-dence suggests that in the past, as also during the 1800's, huntingeconomies were more likely to succeed here than horticultural ; but itis equally clear that where, or when, conditions were at all favorable,native farming peoples were present. We shall return to this mat-ter again.In marked contrast is the Plains Border or Dissected High Plainsregion to the east. It comprises somewhat less than the central thirdof the State. North of the Arkansas, the eastern front of the HighPlains has been irregularly and deeply dissected by stream erosionto produce a broken landscape of high plateaulike uplands, promi-nent and often sharply indented east-facing sandstone or limestoneescarpments, conspicuous headlands, isolated buttes, hills, and roll-ing lowland plains. The eastern part consists of the Smoky Hills,carved mainly in the Dakota sandstone; to the west are the BlueHills, consisting chiefly of limestones and shales. The lower Re-publican, Solomon, Saline, and Smoky Hill Rivers flow in a general Wbdbl] an introduction TO KANSAS ARCHEOLOGY H easterly or southeasterly direction across this belt. Characteristically,they have flat-floored, bluff-lined valleys as much as 200 feet deep,but seldom exceeding 2 or 3 miles in width ; and in this region, theyusually carry water throughout the year. Numerous spring-fedtributaries indent the valley margins; and, like the main streams,their courses are commonly terrace lined. The bottom-land soils areusually deep, well drained, and fertile ; their corn-producing capacityis limited only by the amount of rainfall they receive. Near Con-cordia and Minneapolis, respectively in the drainage of the Repub-lican and the Solomon, salt marshes and springs impart a brackishquality to the creeks which drain them ; and the Saline River to thesouth is reputed to be one of the saltiest streams in the United States.Near Cawker City, on the Solomon, is the famous mineral spring stillknown by its Indian name, Waconda or Great Spirit Spring (McCoy,1840, p. 411; Grinnell, 1893, p. 358; Patrick, 1906). All of thesederive their mineral content from the saliferous shales of the Dakotasandstone.South of the Smoky Hill River, the escarpments turn westwardand fade away in the general level of the High Plains. To the southis the Great Bend Lowland, a broad flat to rolling plain representinga local base-level of the Arkansas River. It includes the roughlytriangular area lying within the great bend of the Arkansas betweenDodge City and Wichita, as well as the nearly flat land northeastof the river around McPherson and Newton. In this latter areaand on the north side of the Arkansas, alluvial and terrace depositspredominate with a belt of dune sands north and east of Hutchinson.South of the river is a poorly drained area of sand dunes and sandyplains, among which are found salt marshes, ponds, and sloughs. Tothe southeast, the sandy soils give way to "hard lands" and alluvialbelts with greater agricultural potential. Tributaries of the Arkan-sas in this section include Pawnee River, Walnut Creek, Cow Creek,and Little Arkansas River, all entering on the left, or north, bank.These generally have firm banks, wide fertile bottoms, a dependableflow of water, and more or less hardwood timber. On the right, orsouth, bank the Arkansas is joined by Rattlesnake and other smallcreeks emptying their brackish waters into the main stream aboveHutchinson; and by Ninnescah River, Bluff Creek, and the Chikas-kia River farther to the south.South of the Great Bend Lowland, the High Plains thrust a longnarrow tongue far to the eastward. Beyond this is a deeply erodedbelt, 10 to 20 miles wide and extending nearly 100 miles along thesouthern edge of the State. Variously known as the Red Hills or theCimarron Breaks, this zone owes its ruggedness to the erosive actionof south-flowing tributaries of the Medicine Lodge and Cimarron 12 BUREAU OF AMERICAN ETHNOLOGY t Boll. 174Rivers on soft Permian formations. Red shales and sandstonesoverlain by gypsum beds have been carved into terraced canyons, boldheadlands, isolated buttes, cones, and pinnacles. Always a colorfuland picturesque landscape, it becomes one of striking beauty whenseen in the light of early morning.Along with the greater topographic diversity of this central Kansassection, as compared with the High Plains, there are climatic andbiotic differences that have influenced man's activities. Originally,this was a mixed bluestem and short-grass region. On the drier up-lands and thinner soils, buffalo and grama grasses predominated ; thebottoms and slopes were occupied by little bluestem and other prairiegrasses, decreasing toward the west as growing conditions became lessfavorable. Western wheatgrass which, like the short grasses, cureswithout cutting into an excellent winter forage, grew in the heavierbottomland soils. North of the Arkansas, the principal streams werefringed with hardwoods, and strips of timber evidently grew alongmost of their perennial tributaries as well. Unlike the scattering andspotty tree growth in the High Plains valleys, these belts of woodlandwere more or less continuous, and they became heavier toward the east,that is downstream. They consisted of cottonwood, willow, ash, elm,and bur oak, and may have included some walnut also. Here, too,there were chokecherries, wild currants, gooseberries, plums, andgrapes. Farther south, the Little Arkansas appears to have been aboutthe western limit of fairly heavy tree growth, although patches ofwoodland and scattering trees were undoubtedly to be found alongCow Creek, Walnut Creek, and on Pawnee River. The low grass-grown banks of the Arkansas above the Little Arkansas seem to havebeen largely treeless except for occasional cottonwoods ; and this wasprobably tiiie of most of the region west and south of the main stream.Native game included virtually all the forms enumerated above forthe High Plains ; but the increased cover along the watercourses andthe more deeply grassed prairies undoubtedly supported greater num-bers of elk, deer, black bear, beaver, otter, and wildcat. Tree squirrels,porcupine, and cougar can probably be added to this list. In thetimber there was a plentiful and varied bird life, including a host ofsongbirds as well as quail and wild turkey. Prairie chicken andgrouse were plentiful throughout this section. In marshy areas, onthe sandhill lakes and ponds, and along the larger streams, variouswater and shore birds had their habitats, and ducks and geese couldbe taken in season. For peoples wholly or primarily on a hunting andgathering subsistence economy, this region could furnish an abundanceof food; and to a thin and scattered population, afoot and relyinglargely on noncommunal methods, the timbered valleys and brokenterrain would probably have afforded better himting than the open Wbdbl] an introduction to KANSAS ARCHEOLOGY 13High Plains country. Moreover, there is here an increased andsurer yearly precipitation, averaging from 22 to 30 inches or more;and this, added to fertile valley bottoms, relatively plentiful wood forfuel and building, water, and convenient flood-free terraces well suitedfor village locations, made the region an attractive one for native, asfor the later white, farming peoples.Eastward again, the Osage Plains are characterized by a successionof irregular east-facing limestone escarpments, 50 to more than 200feet high, which trend northeast-southwest across the State. Theformations composing these escarpments, mainly of Pennsylvanianage, dip slightly toward the west and are separated from one anotherby less resistant strata of shale on which have been developed flatto gently rolling plains. Westernmost of the successive scarped up-lands, and the most prominent, is that known as the Flint Hills.Here the protecting strata are Permian limestones, many of whichcontain an abundance of chert, and the gently rolling upland surfaceis often littered with fragments and nodules of "flint." The irregulareastern edge of the Flint Hills upland generally follows the meridianof 96?30' west longitude; its maximum elevation approximates 1,500feet south of Cottonwood River, dropping to about 1,300 feet northof Kansas Eiver, where the hills lose much of their distinctive andconspicuous character. On the west, the upland surface merges withthe flat plains along the Arkansas River.The Osage Plains are the most generously watered portion of theState, with annual precipitation averaging generally in excess of 30inches. Along their northern margin, drainage is by a series of shortperennial creeks into Kansas River, which traverses the Flint Hillsand then skirts the region to the east in a well-defined bluff-borderedand terraced valley up to 200 feet deep. Farther south, numerousspring-fed creeks give rise to several permanent rivers. The Maraisdes Cygnes (Osage) courses generally eastward, roughly parallelingthe lower Kansas, to join the Missouri beyond the border of the State.The Neosho and its principal affluent, the Cottonwood, head in theFlint Hills upland and flow southeast to join the Arkansas in easternOklahoma. Paralleling the Neosho to the west is the Verdigris, withElk and Fall Rivers as its main tributaries, and uniting with theArkansas a few hundred feet above the Neosho. The Walnut rises onthe west, or back, slope of the Flint Hills and empties into the Arkan-sas a few miles above its point of exit from the State. All of thesestreams run in wide, flat-floored valleys with high bordering bluffs,terraces, and alluvial bottoms; where they cut through the escarp-ments, the channels are often deep, narrow, and lined with rockyledges or cliffs. All are subject to major floods that often overspreadthe valley bottoms; and the bottom-land soils are typically heavy, 14 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174though fairly productive. In dry years the Neosho and Verdigrisparticularly carry little water during the summer.In their original state, the Osage Plains were primarily a tall grassprairie, with big bluestem dominating the rolling plains betweenstreams and much of the valley bottoms as well. The immediatestream banks and adjacent valley floors were heavily timbered withoak, black walnut, elm, linden, sycamore, locust, hickory, pecan, andother hardwoods. Smaller forms that undoubtedly entered into thenative economy of the Indians included Osage orange or bois d'arc,persimmon, papaw, elderberry, serviceberry, chokecherry, and wildgrape. The forested belts and nearby prairies provided shelter andfood for an abundant mammalian fauna, chief among which were elk,white-tailed and mule deer, black bear, cougar, wildcat, timber wolf,gray and red fox, raccoon, opossum, the gray, fox, and flyingsquirrels, beaver, otter, muskrat, and cottontail rabbit. On theprairies were bison, coyote, antelope, jackrabbit, badger, and manysmaller mammals. Among the numerous birds, the plentiful wildturkey was doubtless of primary importance to man; but prairiechicken, ruffed grouse, and quail were also useful, and the passengerpigeon and Carolina parakeet were present. The larger streams,usually running clear and unsilted, yielded an abundance of ediblefish and shellfish.With the advent of the white man and clearing of much of theforest and bluestem cover, the larger mammals have vanished, leavingonly the smaller fur-bearers. The fertile bottom lands and prairies,combined with plentiful rainfall, today produce heavy yields of cornand other crops. The shallower upland soils to the west, as in theFlint Hills district, are much less suitable for crop production ; butthe grass cover of this famous "Bluestem Region," superficially rem-iniscent in a dry year of the High Plains, annually fattens thousandsof Texas and New Mexico cattle for midwestern stockyards.In the Dissected Till Plains north of Kansas River, whose coursefrom Manhattan eastward represents the approximate southern limitof the Kansan ice sheet, the older scarped terrain of the Osage Plainshas been largely obscured by a mantle of glacial drift and loess.The gently rounded and smoothed upland surface merges northwardinto the loess plains of Nebraska and Iowa. In Kansas, the driftis relatively thin, and extends only to a line just south of, and roughlyparalleling, the Kansas River. However, it contains many bouldersof transported materials, some of which are otherwise absent fromthe State. These include much pink to purplish Sioux quartzite fromsouthwestern Mmnesota and northwestern Iowa, as well as granites,diorites, diabases, gabbros, and greenstones. As we shall see, some ofthese were evidently drawn upon by the Indians for manufacture of WBDBL] AN INTRODUCTION TO KANSAS ARCHEOLOGY 15 axes, hammers, and other heavy-duty implements. The loess mantlevaries in thickness from 10 feet on the west to nearly 100 feet alongthe Missouri River bluffs. The Missouri flows in a relatively narrowalluvium-filled trough 2 to 4 miles wide, with well-wooded bot-toms bordered by prominent loess-capped limestone and shale bluffs.Before its stabilization by means of dikes, revetments, and piling, thechamiel shifted frequently, leaving oxbow lakes, sloughs, and marshesalong the old courses. Creeks and ravines joining it from the westare usually short, deeply incised, and heavily wooded ; they are numer-ous and have resulted in a very rough broken strip usually extendingsome distance back from the main valley. The Blue River, at ornear the western edge of the till plains, likewise has a deep, narrow,flat-floored valley ; and this, like the well-defined valleys of the Ver-million, Delaware, and Wolf Rivers and the innumerable smallercreeks, was formerly well timbered. The loessial soils throughoutthis section are among the finest in the State for corn and other crops,which flourish on uplands and alluvial bottom lands alike. Nativefloral and faunal assemblages closely paralleled those in the OsagePlains to the south. These two sections are today the most heavilypopulated portion of Kansas.With regard to climate, Kansas generally is characterized by warmsummers with abundant sunshine, cold dry winters, strong windmovement, and wide variations in temperature and precipitation(Flora, 1948). The mean annual temperature for the State is 55? F.,with extreme thermometer readings of 121? and ?40?. The dailyrange, which is especially marked in the western part, may be as muchas 40? to 50?. The heat of the summer days is alleviated by lowrelative humidity and, in western Kansas, by pleasantly cool nights.During the summer, shade temperatures sometimes exceed 100? forseveral consecutive days with night readings then seldom droppingbelow 75? or 80? ; and these periods are often accompanied by hotwinds whose drying effects cause great discomfort to man and animalsand heavy loss to crops, especially to corn in the tasseling stage.Winds vary from 8 to 16 miles per hour, generally being strongest inthe afternoons and the highest velocities occurring in the west ; theyblow mostly from the south. Average annual precipitation varieswidely from slightly over 40 inches in the southeast corner to lessthan 16 inches along the western border. This comes mostly as rain,varying from widespread cyclonic storms to localized thundershowersthat may be of such great violence and short duration as to leavelittle moisture in the groimd on which they fall. Snowfall is notgreat, and it seldom covers the ground for more than a few days at atime. Owing to the dry winters, soil moisture is generally ratherlimited, so that crops and other vegetation must depend mainly on4?4172?59 3 16 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174spring and summer rainfall. In this respect, the climate is favorable,since 70 to 75 percent of the annual precipitation falls from April toSeptember, when it is most urgently needed. A frost-free growingseason averaging from 150 days in the northwest to nearly 200 daysin the southeast is ample for maturing of most field and gardencrops.Since much of the archeological evidence with which we shall beconcerned in later pages relates to people relying in part on corn-growing, certain aspects of Kansas geography deserve further con-sideration. From what has already been said, it should be obviousthat the region is one of varying agricultural possibilities (Kans. St.Bd. Agr., 1929, pp. 11-20, 153-157). In the eastern stream valleys,a sufficiency of rain, good soils with subsurface moisture throughoutmuch or most of the year, and favorable conditions generally offer asetting comparable to that in neighboring Corn-Belt States of thecentral Mississippi valley. Westward, however, as rainfall dimin-ishes, evaporation increases, and the soils become drier, the GreatPlains environment is progressively less favorable. Here, althoughthe soils are still of great fertility, the climatic factor becomes a criti-cal one for the corngrower, and doubly so where he lacks the tech-nological advantages of the highly specialized methods and crops ofthe present-day farmer. The western limit of successful corn-growing corresponds approximately with the line of 20-inch annualprecipitation?or, perhaps more accurately, with the line of 8-inchJune-to-August rainfall (Jenkins, 191:1) . Normally, these lines lie notfar apart in western Kansas, near or just beyond the 100th meridian.Over the years, however, they fluctuate widely and unpredictably(fig. 3). In 1915, the wettest year on record in Kansas, the 20-inchisohyet lay in eastern Colorado; the western tier of counties inKansas received from 23 to 32 inches of rain while some of the east-em counties had up to 60 inches. In 1936, driest year on record,precipitation along the Kansas-Colorado line dropped to 10 inchesand few counties west of the Flint Hills received as much as 20 inches(Flora, 1948, pp. 31-32). Stated in another way, wet years mayproduce over portions of Kansas a climate normally characteristic ofIowa and western Illinois ; in dry years, parts of the region approachthe aridity of a desert (Thornthwaite, 1941). The western half,thus, lies in a region where the all-important rainfall varies widelyboth above and below a barely adequate average ; and in more thanhalf the years for which records are available, the deviation fromthe average has been below rather than above. Added to this are theexcessively high temperatures and hot searing winds that usuallyaccompany droughts here; they may be even more damaging to thecom crop than the decline in precipitation. That these climatic Wbdbl] .^N INTRODUCTION TO KANSAS ARCHEOLOGY 17 Figure 3.?^The Central Plains, showing historic tribal locations as of approximately1825; shifting rainfall averages; and major grassland divisions. Stippling, tall grassprairies; unstippled, short grass plains. (Courtesy x\merican Anthropologist.) vagaries, periiaps further complicated by dust storms and insect in-festations, probably adversely affected the native populations, as weknow they did the early white settlers in western Kansas, will bepointed out after we have considered the archeological data (see alsoWedel, 1953 a) . For the State as a whole, it should be remembered that Kansas, likeNebraska, lies well within the former range of the great bison herds ; and the presence of this once abundant food resource was, of course, afactor of primary importance in the aboriginal utilization of its grassyplains and broad river valleys. From the earliest Spanish explora-tions through the 18th century, as doubtless long before, these animalsranged generally throughout the length and breadth of the State.Whatever their actual numbers and habits, to the Indians residing inand near the area, they must have appeared as a well-nigh inexhaust-ible food supply. By the 19th century, however, as game east of theMissouri vanished before the food and fur hunter, the tribes thereresident turned increasingly to the trans-Missouri plains on theirhunts. Still other tribes assigned to reservations west of the Missouriadded to the slaughter after 1825, so that the receding herds were tobe found mainly west of the longitude of the great bend of the Arkan- 18 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174 sas, in the short-grass belt. In 1873, Mudge observed (Allen, 1876,p. 147) that?The buffalo ranged to the eastern border of Kansas as recently as 1835. Aboutthat time the United States authorities removed the Delaware, Pottawattamie,Kaws, and other tribes of Indians to "Reservations" in the eastern part of whatis now Kansas. These Indians soon drove the buffalo as far west as the BlueRiver (one hundred miles west of the Missouri River), which was as far as thereservations extended. The buffalo held that range till 1854, when Kansas wasmade a Territory and whites began to settle here. For fifteen years from thattime the buffalo receded, on an average, about ten miles a year. For threeyears past they have been hunted in summer for their hides for tanning ; this isexterminating them very rapidly. Now they are not found in northern Kansaseast of 100? of longitude ; in Southern Kansas as far easterly as longitude 98?,the western boundary of Kansas being 102?. In a few years I think they willnot range north of the Arkansas River.With construction of the Union Pacific Railroad along the overlandroute up the Platte valley in 1868, the bison were split into what cameto be known as the northern and southern herds. To the documenta-tion given by Strong (1935, pp. 36-37) regarding the last decades ofthe bison in the Nebraska-Kansas region, we may add the followingfrom Dodge (1877, p. 131) : In 1870 the original great buffalo range had become permanently dividedinto two ranges. The Southern buffalo ranged from Northern Texas to aboutlat. 41?30'. The Northern ranged from about lat. 43?, through what is known asthe Powder River country, into the British possessions. Of the numbers andposition of the Northern Buffalo but little is known.We will see what has come to the Southern.Their range was as described above, but their most prized feeding ground wasthat section of country between the South Platte and Arkansas rivers,watered by the Republican, Smoky, Walmit [Walnut], Pawnee, and otherparallel or tributary streams, and generally known as the Republican country.Hundreds of thousands went south from here each winter, but hundreds ofthousands remained. It was the chosen home of the buffalo.The split between the two great herds was quickly widened, withconstruction of the Kansas Pacific and the Atchison, Topeka, andSanta Fe Railroads across western Kansas. The railroads laid thegroundwork for the commercial hunter and his bloody work; andwithin a decade of Dodge's writing, these hunters had practically ex-terminated the southern herd, except for a few thousand head whichfound temporary refuge in the Indian country south of the Cimarron.The story of this slaughter has been adequately set forth by others(Allen, 1876, esp. pp. 177-180; Dodge, 1877; Hornaday, 1889; Roe,1951) . Its effects on the Indians who for so long had drawn much oftheir sustenance from the bison are vividly recorded in the reports ofthe Commissioners of Indian Affairs during this period, and in othercontemporary documents. WBDBL] AN INTRODUCTION TO KANSAS ARCHEOLOGY 19THE HISTORICAL BACKGROUNDHigh on the list of objectives set in formulation of the field pro-gram here being reported was the identification and definition,wherever possible, of the material remains left by the principal historictribes of the Kansas region. Here, as elsewhere in the Plains area,the older Indian occupation and the more recent one of his whitesuccessor overlapped through some centuries of time, and the archeolo-gical record merges into that of written history and ethnography.This being so, an obvious avenue of attack was what has come to beknown as the direct-historic approach. As I have elsewhere indicated(Wedel, 1938 c, p. 1), this approach involves two basic steps: (1) theisolation and definition of the archeological criteria characterizingspecific tribal groups, whereby some or many sites could theoreticallybe removed from the category of unknowns, and (2) comparison ofthe materials so identified with earlier remains and the establishmentthus of a time and culture sequence extending from the known historicto the unknown prehistoric. For some years, of course, this method hasbeen applied with conspicuous success in Nebraska and Plams pre-history, as also elsewhere ; but ample opportunity remains for furtherfruitful application throughout much of the central United States.For various reasons, achievement in Kansas fell considerably shortof anticipation ; and the final results of the three-season program herepresented give a picture based only in part on historic sites archeology.Since I am convinced, nevertheless, of the basic soundness of theprocedure wherever it is applicable, and believe it deserving of furthertrial in Kansas, it seems worth while to review in some detail thehistory of white exploration as it relates to the native peoples formerlyresident here. Others, perhaps, will find it feasible to carry on wherethe National Museum cannot.Documentary information on the Indians of the Kansas regioncomes from a variety of sources, mainly Spanish, French, and Ameri-can. These sources involve narratives of explorations, memoirs ofvarious sorts, letters and reports, and maps. Spanish sources are earli-est, beginning with reports of explorations before the middle of the16th century and continuing intermittently for nearly 200 years. Theyrelate usually, and often only incidentally, to the western parts of theregion, nearest the Spanish theater of operations on the upper RioGrande. Beginning in the last half of the 17th century and more thana hundred years after the initial Spanish explorations, the French ap-peared on the scene?first, as missionaries and explorers from theGreat Lakes area, and later as traders among the Indians of the lowerMississippi and Missouri valleys striving to locate nonexistent minesand to establish commerce with the Spanish in New Mexico. French-men continued to carry on in the regions even after the transfer of 20 BUREAU OF AMERICAN ETHNOLOGY [Bdli-. 174French political control in Louisiana to the Spanish in the 1760's.French records, as well as many of those left us by the Spanish, areoften very brief and tantalizingly inadequate; and their interpreta-tion in terms of modern geography is seldom easy. With Americanacquisition of the Louisiana territory in 1803 and the launching of aseries of official exploring ventures and of Indian agency reports, muchfuller and more precise information on the Indians becomes available.The earliest accounts of Indians in the region of present Kansasappear in the narratives of the Coronado expedition (Hammond andRey, 1940; Winship, 1896), and thus date from just before the middleof the 16th century. These narratives have been variously interpretedfor many years, and have given rise to a voluminous literature ; butrecently adduced historical and archeological evidence (see Jones,1929, 1937 ; Wedel, 1942 ; Bolton, 1949, and references cited therein)permits fairly precise conclusions as to the route followed and thedestination finally reached by the Spanish. Since these accounts arebased largely on eyewitness observations, rather than on hearsay, andbecause they afford us a glimpse of the country traversed and of thenative ways of life completely uninfluenced by previous white contact,direct or indirect, they are of first importance.Leaving Pecos in the spring of 1541 and guided by two captivePlains Indians whose homes were east and northeast of that town,Coronado led his army eastward into the Texas Panhandle country.En route, he met wandering bands of tipi-using, bison-hunting Indianswhom he called "Querechos" and "Teyas," evidently Plains Apache.At the eastern edge of the Llano Estacado, at or near Palo Duro Can-yon, the army was sent back to the Rio Grande pueblos while Coronadoand a picked detacliment of 30 horsemen (and six foot soldiers?)turned northward. Marching "by the needle," this party traveled, saysCoronado, "forty-two days . . . living all this while solely on theflesh of the bulls and cows . . . going many days without water, andcooking the food with cow dung, because there is not any kind of woodin all these plains, away from the gullies and rivers, which are veryfew . . .".The Coronado party probably entered present Kansas in the vicinityof Liberal, shortly after the middle of June. On Wednesday, June29, "the feast day of Saints Peter and Paul," Coronado reached the "river Quivira"?the Arkansas near the present town of Ford justabove the beginning of the great northward bend of the stream. Herethe Spaniards turned northeast and downstream. In 3 days, be-tween present Kinsley and Earned, they met a small hunting party ? the first Indians mentioned since Coronado's departure from thebarrancas where he left the army. Three or four days later, afterrounding the northernmost point of the bend and leaving the river, Wbdbl] an introduction to KANSAS ARCHEOLOGY 21they found the first of the settlements of Quivira. Here, in what isnow Kice and McPherson Counties, they spent nearly a month, visitingthe various rancherias?"not more than 25 villages of straw houses,"says Coronado?scattered north and east of the great bend betweenthe Arkansas and Smoky Hill Rivers. Unlike the Querechos andTeyas, wrote Coronado, the natives of Quivira "have the advantage inthe houses they build and in planting corn." The Spaniards werefavorably impressed with the land generally, which, by comparisonwith the High Plains terrain through which they had journeyed solong, seemed well-watered and productive ; but the villages of grasshouses, the corn, beans, and melons growing nearby, and the Indiansgenerally were a great disappointment. In mid-August, having satis-fied themselves concerning the lack of treasure in the region and withguides provided by the natives, the Spaniards retraced their courseto the Arkansas River crossing and then headed by a more direct andshorter route back to the Canadian River and on to Pecos. In thefollowing year, apparently. Father Juan de Padilla returned toQuivira to bring the gospel to its inhabitants, only to lose his lifeafter some months. Thereafter, a half century was to elapse beforewhite men again viewed the settlements of Quivira.Concerning the next Spanish visit to the region we know almostnothing. This was an unauthorized expedition under Francisco Leyvade Bonilla and Antonio Gutierrez de Humana. It proceeded, in 1593or 1594, from San Ildefonso on the Rio Grande northeastward intothe Plains. On a stream identified by Bolton (1916, p. 201) as theArkansas, the Spaniards found a large rancheria where the houseswere of grass and the natives grew bountiful crops. Thence theymarched northward an undetermined distance for 12 days to anotherstream, which Bolton suggests may have been the Platte; I suspectit was the Smoky Hill or Kansas. Here quarrels between the leadersresulted in the murder of Bonilla ; and subsequently, the entire com-mand, with the exception of its Indian guide, was destroyed by theIndians of the region.A few years later, in the summer of 1601, Don Juan de Oiiate leda large expedition consisting of soldiers, missionaries, and others east-ward from the Rio Grande into the buffalo plains (Bolton, 1916).Oiiate left San Gabriel, between Santa Clara pueblo and the mouth ofthe Chama River, on June 23. He crossed the Pecos and GallinasRivers and then descended the Canadian, where he met several bandsof Apaches Vaqueros. Leaving the river because of sand dunes 110leagues from San Gabriel, he moved north and east along an un-certain route. Finally, at more than 200 leagues (ca. 500 miles) fromSan Gabriel, he encountered a large rancheria of Escanjaques, a no-madic bison-hunting people who dwelt in skin lodges, practiced tattoo- 22 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174ing or body painting, and were led by chiefs who had little authority.The Escanjaques guided the Spaniards to a large settlement ("granpoblacion") about 12 leagues farther on where, they alleged, membersof the Humaila party had been slain.Fording a large east-flowing river, the Spaniards found a greatrancheria consisting of groups of grass lodges scattered along thebanks of a small river joining the larger one from the north. Amongthe lodges were patches of maize, beans, and calabashes. The nativeshad metates and made clay pottery, and the men were streaked withpaint (or tattooed ? ) on the face. Other settlements of the same peopleoccurred farther up the stream. Ofiate nowhere gives these people aname ; but in the testimony given in Mexico City in 1602 by membersof the expedition, it is stated that the Indians of San Gabriel, that is,the pueblo Indians, "... name the Indians of the poblacion the Ju-manes because all the Rayados (painted or tattooed Indians) are calledthus..." (Scholes andMera, 1940, p. 274). Outbreak of hostilitiesbetween the grass-house people and the Escanjaques accompanyingOiiate led him to discontinue his exploration and turn back, the ex-pedition arriving at San Gabriel on November 24. Soon thereafter,it appears, a delegation of Indians from the grass house settlementsappeared before Onate to request his aid in their struggles with theEscanjaques (Thomas, 1935, pp. 8-9).Bolton (1916, p. 260) suggests that the gran poblacion visited byOfiate was either on Cow Creek or on the Little Arkansas ; and Scholes(Scholes and Mera, 1940, p. 274) comments that it "was undoubtedlya Quivira settlement, located in the same region as the Quiviras whomCoronado visited in 1541. The distance recorded for Coronado's re-turn trip in 1541 compares with the estimated number of leaguestraveled by Oiiate 60 years later. Moreover, the descriptions of housetypes and other phases of material culture are essentially the same ineach case." I have elsewhere (Wedel, 1942, pp. 18-20) suggested thatthe correct location is perhaps at the confluence of the Walnut Riverwith the Arkansas in present Cowley County, Kans. This location,which I still regard as a possible one, has its difficulties, however, sincethere seems to be no arrangement of streams to the southwest such as isdescribed by Oiiate. If his gran poblacion was on the lower LittleArkansas, perhaps somewhere in the vicinity of present Wichita, theapproach would have been across the Chikaskia and Ninnescah Riversand several smaller streams. In this case, the Escanjaque rancheriamay have been somewhere in present Harper or Sumner Counties. Inany case, it would seem fairly improbable that the earlier Humanaparty in its 12 days northward march got far beyond the Smoky Hillriver. The Escanjaques were almost certainly living somewhere insouth-central Kansas or northern Oklahoma, along some stream notfar west of, and tributary to, the Arkansas. Wbdbl] an rNTRODUCTION TO KANSAS ARCHEOLOGY 23In the half a century following Onate's return fi'om the Arkansas,there appears to have been little or no direct contact between theSpanish on the Kio Grande, busy with their exploitation and proselyt-izing of the Pueblos, and the Indians of the present Kansas region.In that time, however, the Spanish learned more about the variousApache tribes or bands by whom they were surrounded. Here we areinterested only in the Apaches Vaqueros to the east of the New Mexi-can settlements. These Apaches ranged, according to Benavides,more than 150 leagues along the boundary of New Mexico, and ex-tended more than 100 leagues toward the east. Like the Querechosseen by Coronado, the Vaqueros lived solely on the bison, dwelling inskin tents, apparently using the travois, and trading the products of thechase to the Pueblos or to the Indians of Quivira. Despite these tradecontacts and the practice of wintering near the settled Indians, it isobvious that the Apaches were not regarded as friendly. The docu-ments tell us, in fact, that they raided the Pueblos as much as theytraded with them. There is no mention of corn-growing among theeastern Apache, that is, the Vaquero, although the Apache de Navahoto the north and northwest were described as good farmers.^Apache raids on the Pueblos and Spaniards of the Rio Grande be-came increasingly serious after the mid-lTth century. Stimulatedto these attacks by drought and food shortages (Scholes, 1930) andencouraged by Spanish dissension and inefficiency, the Apache evi-dently contributed to the growing unrest that culminated in thePueblo Revolt of 1680. In the general disquiet preceding that event,apparently some time after 1664, several families of Taos fled to thebuffalo plains and established themselves in a fortified spot "whichsince on this account they call El Cuartelejo" (Thomas, 1935, p. 53).Here they remained, among the Apache, until Juan de Archuleta, atthe governor's order, brought them back to the Rio Grande, sometimebefore 1680. Few details of this expedition are available, and thereis nothing from which the location of the "fortified spot" can be cer-tainly determined. It is of interest, however, to note that the fugi-tives were found in possession of "kettles and other pieces of copperand tin . . ." which they said they had obtained "from the Quivirapueblos, to which they made a journey from El Cuartelejo." TheSpaniards were informed also that from El Cuartelejo it was possibleto travel by way of Quivira to the Pawnees, with whom the Frenchwere said to be trading. * Ayer, 1916. The information and specific designations by Benavides, who served inNew Mexico from 1622 into the 1630'8, were used by Nicholas Sanson on his map "Amer-Ique Septentrlonale," dated 1650. The terminology and general delineation were copiedby other map makers. They may be noted on Gulllaume Delisle's maps as late as 1703(Carte du Mexlque et du Mississippi), although the Rio del Norte and adjacent towns andtribes become more condensed and are pulled farther to the west. 24 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174Following subjugation of the Pueblos after the revolt, the Spanishagain found themselves under the necessity of sending out militaryforces in pursuit of fugitive Pueblo Indians who had sought refugein the Plains. Among these were a group of Picuries said to havebeen living among the Apache since 1696. Seeking them, Ulibarriin 1706 led a party of soldiers, colonials, and Indians northwardfrom Santa Fe via Taos to the Arkansas. En route, he noted sev-eral tribes or bands of Apache practicing maize agriculture; and,as a portent of things to come, he learned also of the fear these peoplehad of an attack by the Ute and Comanche, the latter tribe in thisyear making its first documented appearance in history. Crossingthe Arkansas somewhere east of present Pueblo, Colo., the expeditiontraveled eastward parallel to, and a few miles north of, the stream.The Cuartelejo Apache were met ". . . approximately in presentOtero or Kiowa County, Colorado . . . probably [near] the junctionof Mustang and Adobe creeks . . ." (Thomas, 1935, p. 264). Hereand from neighboring rancherias, the Spanish gathered up 62Picuries, but apparently left behind some other Pueblo Indians whowere away with Apache hunting parties. Further news was ac-quired about attacks by the Pawnee, reportedly aided by some French,on the Apache rancherias to the north.Thirteen years later, in 1719, Governor Valverde led another forcenorthward over much the same route in search of the Comanche andUte who were now harrassing the settlements of New Mexico. Againthere is mention of corn-growing and irrigation among the Apachesnorth of the Rio Grande, as well as statements that the Ute and Co-manche, evidently well provided with horses (Thomas, 1935, p. 127),were creating serious disturbances among these peaceful peoples.Valverde never caught up with his quarry, but he met again theApaches of El Cuartelejo. They are said to have "numbered morethan 200 tents, and more than 300 Indians under arms. Togetherwith the crowd of women and children there were probably more than1000 persons . . ." Visiting their camp, the Spaniards noted "thedogs, on which were loaded the poles for tents and other utensils theyused." As with Ulibarri, there is no mention of horses among thesepeople; but they appear to have been growers of maize, beans, andpumpkins. Moreover, among them there was a Paloma Apache from "the most remote borderlands of the Apaches . . ." who bore a gun-shot wound inflicted when the French, allied with the Pawnee andJumanos [Wichita?], attacked his people from ambush while theywere planting corn (Thomas, 1935, p. 132) . The Paloma Apache hadrelinquished their lands to the north as a result of these attacks, andwere moving nearer the southerly Apache groups. Whdel] an ESTRODUCTION to KANSAS ARCHEOLOGY 25The Cuartelejo Apache are mentioned again in the following year,in connection with the ill-fated Villazur expedition. Dispatched byValverde to determine the location of the French, Villazur stopped atEl Cuartelejo on his way to the Pawnee country and apparently rein-forced his party with some of the friendly Apache. Fifty leaguesbeyond, the command was dealt a crushing defeat by the Pawnee,supposedly aided by the French, and three-fourths of the Spanishwere destroyed. This seems to have ended serious exploration by theSpanish toward the northeast, though they continued to get wordfrom the Apache concerning events at El Cuartelejo. In 1727, agroup of Paloma and Escalchufines Apache reported that six French-men had come to El Cuartelejo, five of them going out with theApache in search of the Comanches. Other Frenchmen were reportedsettled on the Rio de Chinali not far away. According to anotherreport, some Frenclimen had gone "with a great force of Apaches ofthe nations Palomas, Cuartelejos, and Sierra Blancas to look for theComanches (a people widely scattered because of the numerousnessof their nation)." The Comanche, wrote Bustamente in 1727, "arein El Almagre [Colorado Front Range, according to Thomas, 1935,p. 283] or a little farther away."The historical data reviewed in the past few pages are of interestas indicating the nature of the early postcontact Indian occupation ofthe short-grass plains between the Rio Grande and the Arkansas ? and perhaps as far north as the Republican or South Platte. Theysuggest, in addition to roving bison himters, the presence of semihor-ticultural peoples in the stream valleys of southeastern Colorado andprobably as far east as western Kansas. At least one historian, fol-lowing his documents closely and discounting the archeological evi-dence, places El Cuartelejo in eastern Colorado, where the distancesand directions given in the narratives seem indeed to locate this point.Curiously enough, no archeological evidence has ever been adduced insupport of this identification. This seems the more remarkable inview of the fact that Pueblo Indians on several occasions between1664 and 1706 fled into the Plains and established themselves amongthe Apache at El Cuartelejo; that for at least 10 years there weremore than 60 of them here (Thomas, 1935, pp. 77-78) ; and that theseIndians seemingly erected small huts (pueblos?) apparently distinc-tive enough to merit comment from the Spaniards who came to rescuethem (Thomas, 1935, pp. 68 ; 262 n. 6 ; 264 n. 23 ; 268 n. 60). In thisconnection, it should be noted that for nearly 60 years archeologistsand historians have known of the existence in western Kansas of anindisputably puebloan structure, with irrigation works and associatedremains; and, historians to the contrary (Thomas, 1935, p. 268 n. 60)notwithstanding, this structure belongs to the very period under dis- 26 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174 cussion. We shall return to this mteresting problem in another place ; but the indubitable presence of early 18th-century puebloan remainsin western Kansas is adequate justification for our somewhat detailedreview of contemporary Spanish observations in the region approach-ing it from the west.Of the Frenchmen who, the Apaches insisted, were allied with thePawnee against them during the closing decades of the I7th centuryand shortly thereafter, there appears to be little or no direct docu-mentation in the contemporary French accounts. It seems fairlylikely, nevertheless, that adventurers and traders from the Missis-sippi Valley or the Great Lakes, perhaps traveling together formutual protection, must have been among the Plains tribes some timebefore the explorers whose travels were recorded in documents nowextant.Just when this westward penetration began and how extensive itwas there is at present no way of knowing. Such published Frenchdocuments as are available concerning the Missouri Valley and thePlains in the late decades of the I7th century indicate only a vagueand generalized knowledge of the region?little more, in fact, thanthe awareness that there were numerous and varied tribal groupsthere. This is apparent both in the terse lists of tribes and in themaps that resulted, directly or indirectly, from these explorations.Most of the latter are so ambiguous and confused that it is difficultor impossible to reconcile them with present-day maps of the area.Probably our earliest French information on the Indians of theMissouri River region came from the explorations of Marquette andJolliet, who passed the mouth of the Missouri late in June, 1673, andnamed the stream "la riviere Pekitanoui." No journals survive; butmaps compiled by these men, and others based on their findings, givesome crude indication of tribal distributions on the lower reaches ofthe Missouri. Their information was probably derived mostly fromnatives of the Illinois country. The "Pekittanoui" of the Marquettemap of 1673-74 is also shown, though without name, on the "Jolliet"map of 1674 and the Randin map of 1674-81 (Tucker, 1942). Mar-quette shows the Osage, Missouri, Kansa, and Paniassa, in that order,westward from the head of his veiy short "R. Pekittanoui." On theother two maps, what is evidently the same stream is shown longerand larger ; and south of it, going upstream, are successively the Mis-souri, Kansa (or Canssa), Osage, Pani, and another unrecognizedname.These or similar tribal names, plus some additions, appear on mapsand in documents based on the discoveries of La Salle, who reachedthe mouth of the Missouri 9 years after Marquette and Jolliet, inFebruary 1682. The La Salle journals and his map are now lost ; but Wbdbl] an introduction TO KANSAS ARCHEOLOGY 27from letters and other contemporary documents, and from mapsevidently based on missing originals, it is clear that La Salle hadgathered a good deal of geographic information beyond that possessedby his predecessors. It is thought that he derived some, if not all, ofhis inforaiation on distant Missouri River tribes from a Pawnee slave.The Franquelin map of 1684,'' showing a confused and impossibledrainage pattern for the lower Missouri and its western tributaries,includes the following tribes in this general region: Missourits(Missouri), Zages (Osage), Paneassa (Wichita), Paneake, Cansa(Kansa), les Pammaha, and the Panetoca. Far to the west are theGatacka (Kiowa Apache) and Manrhaut (Kiowa?), indicated ashaving horses. The Minet map of 1685 (Tucker, 1942, pi. 7), show-ing the same strange network of streams on the lower Missouri, hasthe Missouri, Paneake, Pana, Pani, and Pentoca. Farther west, be-yond the Missouri Eiver system, is the "Rio del Norte" and the "apaches de vaqueros" of the Sanson 1650 map. So far as I am aware,there are no contemporary descriptions of any of these Central Plainspeoples or their ways of life and cultural characteristics. Sketchyand incomplete as the Marquette, Jolliet, and La Salle materialsadmittedly are, they still suggest tribal alinements not radically differ-ent from those of later times. It must be acknowledged, however,that convincing identification of some of the names appearing on thesemaps has not yet been made.After the turn of the century, with the founding of a Frenchcolony at the mouth of the Mississippi and resultant intensificationof activity in the Mississippi VaUey, the picture steadily comes intosharper focus. For a decade or so there are only passing referencesto the tribes of the lower Missouri (Villiers, 1925; Nasatir, 1952,vol. 1, pp. 6-12) . These, named frequently in the order of their oc-currence along the stream, include the Missouri, Osage, Kansa, Oto,Iowa, and Pawnee (Panimahas, Panis, Paniassas, Panimana, Pani-boucha) in the writings of Father Marest (1700), M. de Montigny(1699), Henri de Tonti (1700), Sieur d'lberville (1702), and FatherBergier (1702) . The Panioussas (Wichita) are located on the AcansaRiver by Diron Dartaguiette, writing in 1712; but even earlier, in1700, Tonti (Delanglez, 1939) wrote of the Arkansas River that "TheMentons and the Paniassas are located on this river." On a sketchmap using Tonti's information, the "Paniassa" are on an affluent ofa river intended to be the Arkansas and their village is indicated asfive "nights" from the Mentons, who in turn are five nights from one * According to Delanglez (1943, p. 60), "For all that pertains to the course of theMississippi, its tributaries, and the Indian tribes, this map is a copy of La Salle's largemap, Some La Salle Journeya, 34, which was re-drawn by Franquelin and Minet." Seealso HamUton, 1934. 28 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174of the Arkansas Indian villages. Delanglez equates five nights with40 leagues or 108 miles (Delanglez, 1939, p. 229) . It is noted by Father Marest that the Pawnee are trafficking withthe Spanish and getting horses from them. This information isduplicated for the Paniassas by Dartaguiette. Iberville commentedthat the Canadians were now traveling on the Mississippi and Mis-souri in bands of seven or eight.That these and other activities were sharpening geographical knowl-edge of the region is suggested in G. Delisle's maps for the years1700-1703, inclusive. The Kio del Norte and Apaches Vaqueros arewithdrawn more to the southwest, and the western tributary systemof the Mississippi expands with much detail. On the Delisle 1703map, "Carte du Mexique et de la Floride," the Osage are shown on abranch of the "K. des Osages" and the "Cansa" are located up theMissouri on another westerly tributary termed "Metchigamiki."West of the Cansa, on the lower reaches of another upriver tributaryof the Missouri are the Apana, and at its head the Paniassa. Southof the Missouri, well up the "R. des Acansa," two tributaries come infrom the north ; on the lower are les Paniassa, on the upper les Panis.Significant additions to contemporary knowledge of the regioncame from the activities of Bourgmond, who by 1714 was living in aMissouri Indian village above Grand River. In that year, Bourg-mond prepared a detailed statement of distances by river from themouth of the Missouri to the River of the Pawnees, i. e., the PlatteRiver of Nebraska.^ Three years later, he wrote a description ofLouisiana, summarizing the location and characteristics of a numberof tribes residing along the stream as far north as the Arikara. Thefirst-hand information gathered by Bourgmond undoubtedly con-tributed materially to the greatly improved Delisle map of 1718 forthe region. On this map (Tucker, 1942) appears, apparently for thefirst time, the Kansas River so designated?"Grande Riviere desCansez"?and in approximately its correct position with reference tothe bend of the Missouri River above present Kansas City. On asouthwesterly tributary of this stream are "les Cansez," and on itsheadwaters are the Padouca. The Padouca are also shown on theupper reaches of the "Riv. des Akansas." Some distance down the "Akansas," and on a tributary flowing southward in about the longi- ' In this connection it Is Interesting to compare certain of Boiirpmond's fignires here with correspondingmileages as reckoned by the Corps of Engineers along the Missouri River thalweg of 1890, thus:Mouth of Missouri River to? Wbdel] an introduction TO KANSAS ARCHEOLOGY 29tude of the upper Osage river are "les Paniassa 4 villages" ; and thePaniassa are shown again farther down at the junction of anothernortherly tributary with the "Akansas." On the west (right) bankof the Missouri, just below a small westerly tributary termed "PetiteRiv. des Cansez" and entering between the Kansas River and theRiver of the Panis, or Platte, is another village of "les Cansez."Bourgmond returned to the Missouri in 1723 to establish FortOrleans on its north bank near present Malta Bend, Mo. In the sameyear, the engineer La Renaudiere briefly described the "Grand Villagedes Quans," consisting of 150 lodges standing near the Missouri on asmall stream 30 leagues north of the mouth of Kansas River. Fromthis village, presumably the one shown some distance above KansasRiver on Montigny's map of circa 1728 (Nasatir, 1952, vol. 1, opp. p.14) , Bourgmond in the following year traveled to the Padouca to makepeace between them and the Missouri River tribes.The exact location of the Kansa village in Bourgmond's time hasbeen established beyond reasonable doubt (Remsburg, 1911, 1919).The Montigny map, cited above, shows it on a small westerly tributaryof the Missouri, some distance above Kansas River; and La Re-naudiere's statement that it was on a small stream 30 leagues above the "Quans" (Kansas) River is not far in excess of the 60 miles, by water,from Kansas River to Independence Creek near present Doniphan,Kans. When Lewis and Clark passed the second old Kansa villagesite, going upstream from Kansas River, in 1804, and named the streamjust below it Independence Creek, they referred to Bourgmond'searlier visit to the Indian town which had stood on this spot. Thisis undoubtedly the earliest certainly identifiable Indian village ineastern Kansas.From this Kansa village, Bourgmond left on October 8, 1724, on hissecond attempt of that summer to treat with the Padouca. Hemarched in a west-southwest direction, crossing grassy uplands andtimbered streams during the first 3 days; and on October 11, a halfday's march beyond 20 leagues of travel, he crossed the "Great Riverof the Canzes ... at a ford. ... It enters the Missoury about 20leagues from where we crossed it . . . ." After the crossing, he con-tinued west-southwest and southwest for approximately 48 leaguesmore, mentioning a number of streams and commenting on the abun-dance of bison, deer, stags (elk) , turkey, etc. On the 18th, he met thePadouca, with whom he spent the next 3 days. On October 22, he setout on the return?not to the Kansa village, but to the Missouri somedistance below. Distances and directions for each day of the returnmarch are given, but almost no other details. For the first 2 days, hemarched east-northeast for 15 leagues, there being nothing to indicatethat he crossed any streams of consequence. Then he turned east for 30 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174another 26 leagues, at which point he was across the "Canzes" Riverand camped on its north side. Continuing east for 24 leagues more,Bourgmond says on October 31 that "we are half a league from theCanzes River." At noon or in the early afternoon of November 1,for which day no distance is stated, he reached the Missouri, embarkingnext day by canoe for Fort Orleans.Accepting the site of present Doniphan as Bourgmond's point ofdeparture on this trip, his probable route to the Padouca and back tothe Missouri can be determined with fair accuracy. There is, of course,some uncertainty in the fact that we can never be positive as to howmuch latitude must be allowed for such terms as "westsouthwest," "eastnortheast," etc., nor can we know the probable margin of errorin the distances estimated in the day-to-day marches. Both directionsand distances were unquestionably affected by irregularities of terrainand by other factors. Despite these and other difficulties, we mayprofitably attempt the reconstruction of his line of march. I havetaken the value of the league on land, as used by Bourgmond, to bebetween 2.5 and 3 miles (cf . p. 28) . There seems little reason to doubt that Bourgmond crossed theKansas River, outward bound, in what is now northern ShawneeCounty, between present Topeka and Rossville. His mention of aford suggests, though it does not necessarily prove, that from the Kansavillage he was making for a crossing point well known to the Kansa,Missouri, Padouca, and other Indians accompanying him?a point,perhaps, where better footing could be found than elsewhere along thestream. Such a crossing place, sometimes known as Rocky Ford andapparently frequently utilized during the 19th century, was situatedabout 18 miles above Topeka, near present Rossville and the junctureof Cross Creek with the Kansas River. This is approximately 60miles, airline, from Doniphan ; Bourgmond's distance to the ford, itwill be recalled, is about 24 leagues, covered in 3i/4 days' march. Ithink it quite likely that this is where Bourgmond forded the KansasRiver (see also Villiers, 1925, p. 111).Continuing his march, Bourgmond evidently followed along thehigher land back from the river and its short southerly tributaries,some of which he had to cross. The "hauteur de terres" mentioned onOctober 18 was probably the Flint Hills upland in northern MorrisCounty, between the Neosho headwaters on the south and the shorttributaries of the Kansas on the north. The final 2 days of marching(12 leagues, or 30-35 miles) seem to have been toward the west. Thiswould put the final point reached?the Padouca camp or village?inor very near the Smoky Hill valley in the vicinity of Lindsborg orSalina. I see no real evidence that Bourgmond crossed the SmokyHill, which flows north between Lindsborg and Salina, and thus would Wbdel] an introduction TO KANSAS ARCHEOLOGY 31seem to have lain directly athwart his line of march. The allusion toa brackish stream on whose banks was a recently abandoned camp ofthe Padouca, and beyond which at 3 or 4 leagues he finally met thePadouca themselves, may be to one of the saline streams west andnorth of Salina; or it may refer to one of the creeks just east ofSalina, such as Gypsum Creek, whose waters have a mineral content.On the return march, Bourgmond traveled east-northeast for 2days, or 15 leagues (ca. 40 miles). If the Padouca camp was in theSmoky Hill valley, as suggested above, such a march would havetaken him to the vicinity of Junction City. Thence, marching over-land in an easterly direction, he again crossed the Kansas River,probably at or near the ford utilized on the outward journey; andfrom here proceeded down the Kansas valley to the Missouri. Sincehe notes that on the last full day of this march he was half a leaguefrom the Kansas River, it seems possible that much of the returnmarch was somewhat north of the stream, perhaps in the bluffs oruplands back from the river. The descent into the valley of theKansas was then made only as they finally neared the union of theKansas and Missouri Rivers.This identification of Bourgmond's probable route to the Padouca,while not without its difficulties, seems to me a rather more satisfactoryinterpretation than most of those that have been attempted by earlierstudents. Hyde (1951, p. 42) locates the Padouca grand village "evidently in the district where the town of Salina now stands" ; hismap indicates it as on the Saline, somewhere in present LincolnCounty and some 40 or 50 miles northwest of Salina. This I finddifficult to reconcile with the repeated statements in the narrativesthat the Bourgmond party traveled mostly westsouthwest and south-west to reach the village; and it is even more difficult to see how 2days march eastnortheast from this point, thence east, would lead tothe Kansas River crossing. Likewise, it seems to me, Villiers'(1925, p. 112, and map 1, opp. p. 32) placement of the Padouca villagein Rice County runs afoul of the directions and distances given for thereturn march to the Missouri. From Rice County, Bourgmond wouldhave been crossing and recrossing the Smoky Hill and/or the Kansasseveral times, and of this I see no indication in the narrative. Con-nelley (1918, pp. 454^55) calculated that the Padouca were in Russellor Ellis County; but since this seemed to him surprisingly far eastfor a Padouca settlement, as indeed it does to me, he apparently re-jected Bourgmond's distances and decided that the village was actuallymuch farther west, in Trego or Gove County.Whatever reliance one wishes to place in Bourgmond's estimates ofhis daily mileage, I do not see how the Padouca can have been muchfarther west than present Salina and still be reached by the Bourg-4S4172??9 4, 32 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174mond party in the 10 days required for the trip each way between thePadouca and the Missouri River. The trips were close to 70 or 72leagues each way; at 2.6 miles per league (cf. Villiers, 1925, p. Ill),this would have meant 175 miles, or more, and an average day's marchof about 17 or 18 miles. This daily mileage compares favorably withPike's day by day marches from the Osage to the Pawnee in 1806,partly over the same terrain traversed by Bourgmond (Pike, 1810, pp.132-143) . I doubt that the Bourgmond party would have maintainedan appreciably higher daily average over a 10-day period than Pikedid.Bourgmond's observations, as just reviewed, are of unusual import-ance for our purposes because they seem to have been rather carefullyrecorded and thus provide more data than do various other French doc-uments dating from the first half of the 18th century. In the seconddecade of this century, the French became increasingly interested inthe Missouri River area?an interest stimulated by the desire to locatevaluable metals and to find a route to New Mexico via the Missouriand its tributaries. The only exploration of which there is much in-formation is that of Claude Charles Du Tisne (Nasatir, 1952, vol. 1,p. 18), who on a second attempt to contact the Padouca, ascended theMissouri and Osage Rivers in 1719 to the Osage villages. Here heobtained horses and went overland toward the southwest, 40 leagues in4 days, to two large villages of "Panichouachas," apparently some-where in southeastern Kansas or in northeastern Oklahoma (p. 533).These people were said by Charles Legac, writing in New Orleans in1721 and using Du Tisne information, to have 300 villages farther westwhose inhabitants traded with the Spanish and warred with the "Palo-guas" (Padouca). They refused Du Tisne permission to go on tothe Padouca, who were their bitter enemies. Legac also mentions inhis Memoir the Causes who descend the Missouri River, and the Pan-iouassea to the north and northwest of the Red River.Other memoirs in this period provide interesting scraps of informa-tion. Hubert, writing from New Orleans (Nasatir, 1952, vol. 1, p. 12) , also mentions the villages of the "paniouassea," places them on the "R.des Acancas," and says they war with the "Padoquas" who trade withthe Spanish. In the area farther north, he places the Kansas River 102leagues up the Missouri and says that 120 leagues up the Kansas itforks, with one branch "running west," one northwest. From one ofthese forks it was only 25 leagues to the upper reaches of the River ofthe "Acancea," where there is a village of the Padoca. He repeats thefact of trade between the Padouca and Spanish whereby bujffalo prod-ucts were exchanged for knives and axes.The manuscript map of Le Maire, dated 1716 and entitled "CarteNouvelle de la Louisiane et pais circonvoisins," reflects little advance Wedbl] an ESTTRODUCTION to KANSAS ARCHEOLOGY 33in knowledge of the Missouri beyond that depicted by Delisle in 1703.It shows the "R. des Cansez" below the Platte, and the Osages on thehead of a river lower down. On the northern tributaries of the "Riv.des Akansas" are the Panis and Panioussa, a delineation distinctlyreminiscent of the Delisle 1703 map cited above. In a memoir of 1718,Le Maire names the rivers of the Pani and Padouca (i. e., the Platteand Kansas) as the largest tributaries of the Missouri; and in a dis-cussion of the Spanish trade, he observes that "The Spaniards who goto the Missouri fear to be robbed by the Apaches or Padoucas who arethe Arabs of these parts . . ." (Villiers, 1925, pp. 65-66).On the Missouri River itself, the French had apparently been longcontemplating establishment of a stronger post than that which stoodfor a time at Fort Orleans, abandoned or destroyed in 1728 or soonafter. Such a post was eventually established, though the date of itsfounding seems obscure. Nasatir (1952, vol. 1, p. 28) implies that "anew fort, Cavagnolle, was built near Kansas City" before the Malletbrothers made their trip from the Missouri across Nebraska and Kan-sas to Santa Fe, in 1739. At any rate, there seems to have been a fortsomewhere on the Missouri above Kansas River during the I740's andI750's ; and at times, if not continuously, there was apparently a KansaIndian village in its neighborhood. If, as seems probable, Fort Cav-agnolle is to be identified with the "remains of a fortification erectedby the French" seen by Long (James, 1823, vol. 1, p. 110) on the westbank of the Missouri some 35 miles above Kansas River, it would ap-pear that the fort stood just above present Fort Leavenworth and afew miles below Cow Island. At Fort Cavagnolle, Chapuis stoppedin 1751 before his departure for the Pawnee, the Comanche, and finallySanta Fe. The post was still functioning in 1757, when Kerlerecwrote that "Fifty leagues further up [above the Missouri and LittleOsages] are the Kansas, where Fort Cavagnolle is located, which con-sists of un entourage de Pieux which encloses some bad cabins and huts.The officer there commands seven to eight garrisoned soldiers and sometraders. These Kansas were very numerous, but the wars that theyhave had with the Pawnees and small-pox have extremely weakenedthem. There remain today only 250 to 300 men. They are very at-tached to the French . . ." (Nasatir, 1952, vol. 1, p. 52).French domination, politically at least, came to an end in 1763, whenLouisiana was divided between the English and the Spanish. ThoughFrenchmen had evidently been beating up and down the Missouri Rivercountry for a long time and doubtless knew its native inhabitants well,much of the information they had acquired was slow getting into thewritten records. The maps dating from near the close of the Frenchperiod show no decided improvement over such earlier ones as theDelisle map of 1718. The Dupratz map, dated 1757, again shows the 34 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174 "K. of the Cansez," with a "Cansez" village some distance up thestream ; but the "Cansez Gr. Village" is indicated on the west bank ofthe Missouri some distance above the Kansas River. West of the Osage,who are shown on a southerly tributary of the lower Kansas and on theOsage River, are the "Panis blanc," near the head of White [laterGrand or Neosho] River. The "Country of the Padoucas" is shownwest of the headwaters of the Kansas and on the upper Arkansas, westof what strongly suggests the great bend of the latter stream. The "Padoucas Gr. Village" is near the head of one of the westerly streamsforming the Kansas, with additional villages on other streams in theregion. The "Country of the Panis" is north of the Kansas River,along a badly misplaced east-flowing Missouri River. South of theArkansas is the "Country of the Kanoatinos" [Wichita or ?]. Else-where, Du Pratz wrote that "The most numerous of all those nations[west of the Mississippi] are the Padoucas . . ." (Du Pratz, 1774, p.321).The data thus left us by the French from circa 1673 to 1763, as herereviewed, introduce us to a series of new tribal designations. TheFrench approach to the region was generally from the east, that is,from the Mississippi Valley, with the documents indicating some in-filtration along the westerly tributaries of the lower Mississippi, suchas the Arkansas, by the second half of the l7th century, and more alongthe lower Missouri and its branches by the opening of the 18th century.By the mid-18th century, Fort Cavagnolle had been established nearpresent Leavenworth, Kans. All of these operations brought theFrench into direct contact with the semisedentary or village Indiantribes of the eastern Plains.For our purpose, the important point in this French activity is theindicated presence of a variety of native peoples following, as theSpaniards earlier noted, two ways of life. From the very beginning ofFrench operations, there were a series of settled semihorticulturalIndians in the eastern Plains; and these included in and around thearea of our interest such groups as the Osage, Kansa, Pawnee, andprobably the Wichita (Paniouassa, etc.). Farther west, up theArkansas, Kansas-Smoky Hill, and other streams to north and south,were other peoples, hostile to the settled tribes and the French, andmore directly involved with the Spanish in New Mexico. To theFrench, these were known by the general term Padouca or some variantthereof ; and despite some confusion, the general implication seems tobe that these were the people whom the Spanish knew as the PlainsApache.In the Spanish period intervening between French and Americancontrol of Louisiana, the information available on the Kansas regioncontinues to be fragmentary and vague. There are, however, two sum- WEDEL] AN INTRODUCTION TO KANSAS ARCHEOLOGY 35maries of tribes in the area which are of interest. In 1777, LieutenantGovernor Cruzat included in a report to his superiors a tabulation oftribes receiving presents at St. Louis, and therein enumerated thefollowing either certainly or probably in the region under considera-tion (Houck, 1909, vol. 1, pp. 142-145) :Tribe of the Cances. This tribe is composed of 350 warriors . . . They are150 leagues from this village [i. e., St. Louis], and are located on the banksof the Misury river itself, at a distance of some 50 leagues from the tribe of theMisuris . . . this tribe is hostile to the tribes of the said Misury river, namedthe Panis and La Republica . . . From the work of the hunt in which they areengaged, there results the profits of the trade which are made in the furs ; forevery year that trade produces 180 or 200 packs . . .Tribe of La Republica. This tribe is composed of 350 or 400 warriors . . .They are located some 220 leagues from this village, and about 110 from theMisury river on the shores of the Cances river, and about 40 or 50 leagues fromthe village of the tribe of that name by land . . . This tribe is hostile to thetribes of the Cances and the Big Osages . . .Tribe of the Panis ... of 5 or 600 warriors . . , 230 leagues from this village,about 15 leagues from the Hotos [Oto] tribe, on a small stream branching ofEfrom the Plata river . . . hostile to the Canzes and the Sioux tribe . . .Tribe of the Big Osages . . . composed of 800 warriors . . . They are located180 leagues from this village by water, and about 110 overland, on the banks ofa river emptying into the Misury of about 140 leagues in length. This tribe ishostile to the tribes of La Republica, the Hotos, the Alkanzos, the Panis, thePiquies [Wichita], and the tribes living on the Misisipy in the English district . . . every year this tribe produces 500 or 550 packs of deerskins . . .Eight years later, Governor General Miro at New Orleans compileda description of Louisiana, dated December 12, 1785, which suppliesadditional details (Kinnaird, 1946, pt. 2, pp. 159-167) . Here we aretold that the "Rio de Cans, or Kansas ... is 108 leagues from themouth of the Missouri, on its right bank. At high water one may go upthe river to the village of the Republica or Pawnees whom the Indianscall Paniguacci, or 'eyes of the partridge.' " Other Pawnees, in-cluding the "Wolf Indians, or Panimahas" lived on the "Chato orPlatte River," and on its tributary, "the Rio de Papas, or Wolf" thatis, the Loup. Concerning the Indians themselves, Miro says: "TheKansas have their villages 140 leagues from the mouth of the Missourion a very high bank two aranzadas from the shore of the river men-tioned above [the Missouri or Kansas?] . . . have about 200 men whocan bear arms . . . make war on the Pawnees in order to obtain horses.Their hunting lands are found extending up the Kansas River [sic]as far as the Nemaha River." The "Indians of the Republica, orPawnees, called Paniguaccy or ojos de perdiz, live on the Kansas River130 leagues from its mouth . . ." The Pados or Comanches, Mirowrote : were formerly the most numerous nation on the continent, but the wars whichother nations have made upon them have destroyed them, so that today they 36 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174form only four small groups. They wander from place to place constantly, andthis saves them from the fury of the other nations who constantly make warupon them. They number about 350 men who are very skillful with the arrowand in running.The Laytanes, or wandering Apaches, who, like the Pados, are better knownin those provinces than in this, live on the borders of New Mexico. They areconsidered the best warriors on the banks of the Missouri, and dominate all theneighboring nations. Although divided into various bands or parties, they livein perfect friendship. . . .Forty leagues from the mouth of the Missouri on the right bank, is the riverof the Great Osages, quite a large stream, which flows from the west, one-quarterto the northwest. It is possible to go up this river to the village of the GreatOsages situated 120 leagues from the Missouri ; but high water is necessary forthis. . . . The village of the Great Osages is situated 120 leagues in the interioron the river bearing their name. . . . The Great Osages are the most numeroustribe of the Missouri, at least of those with whom we trade. . . .Finally, speaking of the village Indians of the lower Missouri, Miroobserves : "It is necessary to state in passing, that the wealth of the Indians of the Missouriconsists in having large droves of horses which they take from the Laytanes, orApaches . . . ."A few years later, the Frenchman Pedro Vial was dispatchedfrom Sante Fe by the Governor of New Mexico, Don Fernando de laConcha, to "open direct commmiication with our settlements of LosYlinneses, which are located on the shores of the Missouri Eiver,"that is, at Spanish St. Louis (Hulbert, 1933, pp. 48-54). Vial andtwo companions left Santa Fe in May, 1792, and traveled about 100leagues in an easterly direction, much of the time along the CanadianRiver. Then they turned northeast "in search of the Napeste River,which we call in French the Arkansas River." Thirty-seven leaguesfrom the Rio Colorado, or Canadian, the party reached the Arkansas.Following this stream, "which flowed east northeast . . . ," they pres-ently came upon a hunting camp of Indians who "told me that theywere Cances." This was evidently in the region of the great bendof the Arkansas in central Kansas. These Indians stripped thewhites, robbed them of their horses, and otherwise ill-used them ; butafter some six weeks, in the middle of August, the Vial party "leftwith the above-mentioned Indians on their return to their villagegoing in a northeasterly direction, and we have journeyed in ten daysabout 50 leagues going through level plains." On August 25, saysVial, "we reached their village, which is located on the River of theKances. That river flows into the river called Misoury." Here theystayed until September, when they left with three French traders forSt. Louis, finally reaching their destination on October 6, 1792.Vial appears to have made a second trip over the same generalroute, for in 1795 Zenon Trudeau, Lieutenant Governor of "Western Wbdbl] an introduction TO KANSAS ARCHEOLOGY 37Illinois," reported to Carondelet that "Pedro Vial, who two years agowas commissioned by the governor of Santa Fe to come to these estab-lishments of Illinois, arrived this year (with four inozos who accom-panied him) as far as the Panis nation (commonly called republic)which has its village on the bank of the Kansas [Eepublican?]Kiver. . . . The mentioned Vial said he traveled from Santa Fe tothe Panis in 8 days . . ." (Nasatir, 1952, vol. 1, p. 329).In the final decade of the 18th century, prodded by reports ofBritish activity on the Upper Missouri, the Spanish Government de-veloped new interest in exploratory and trading expeditions to theUpper Missouri. In order to meet the British competition there, toopen new areas for fur trading, and to find a route to the PacificOcean, several parties of traders departed from St. Louis, mainly onbehalf of the newly organized Commercial Company for Discoveryof the Nations of the Upper Missouri. Since their sights werefocused on the upper reaches of the river, few seem to have takenmuch notice of the peoples inhabiting the lower valley. An exceptionis Jean Baptiste Trudeau, whose journal of his 1794-95 trip makesbrief mention of tribes in what is now northeastern Kansas. Leav-ing St. Louis on June 7, 1794, Trudeau reached the Kansas River onJuly 12, an estimated 110 leagues from the mouth of the Missouri.He says (Nasatir, 1952, vol. 1, p. 261) that ? this river in the spring is navigable for more than 100 leagues from its mouth.It abounds in beaver, otter, and other wild animals. 80 leagues from itsmouth is the village of the Kansas ; good hunters and good warriors.10 leagues further up issues [illegible] which the Republican Pawneeinhabit. . . . There are two items of especial interest in these documents fromthe latter part of the 18th century. One is the reported presenceof the Republican band of Pawnee in the Kansas River drainage;the other is the shift of the Kansa Indians from their old habitaton the Missouri to a new one on the Kansas River. The time ofthis Kansa shift is nowhere clearly indicated, so far as I am aware ; but the documents cited would seem to bracket the move within aperiod of a decade or so. Cruzat, for example, has them on theMissouri in 1777, and Miro's report of 1785 apparently confirms this.By 1792, according to Vial, they were certainly on the Kansas, whereTrudeau locates them in 1794. On the other hand, the Bellin mapof 1743 and the Delisle map of ca. 1750 (PauUin, 1932, pi. 23B, C)both show the Kansa on the Kansas River but not on the Missouri.Henry L. Ellsworth, in a letter to Indian Commissioner Herringaccompanying the former's report of a treaty with the Oto and Mis-souri in September, 1833, mentions this shift, but without dating it,in the following words : 38 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174 ". . . the evidence is satisfactory that the Otoes attacked theKansas at their old village on the Missouri near Independencecreek?drove them from their village and took possession. The Kan-sas never afterward occupied that ground but pitched their tents 60or 80 miles distant on the Kansas River . . ." (Nat. Arch., Off. Ind.Aff., Rec. Gr. 75, Treaty File) . With consmiunation of the Louisiana Purchase in 1803 and transferof this vast, but vaguely defined, territory to the United States, thestage was finally set for systematic explorations. From these ex-plorations, which were official functions of the United States Gov-ernment, and from the additional observations of Government factors,military men, scientists, artists, and other travelers in the region,definitive information on the Indians and other features of the Mis-souri Valley was finally forthcoming. There are still a great manydifficulties in the way of identifying precise tribal and village loca-tions even from such meticulously kept and detailed journals as thoseof Lewis and Clark, Pike, Long, and their contemporaries and suc-cessors. On the whole, however, it is usually possible by this timeto correlate locations and districts with modern knowledge of the areainvolved. For the most part, unfortunately, the tribes in our im-mediate area were now giving up many of their native culture traits,and they were destined to lose many more by the middle of the 19thcentury. It is, of course, in the sites of this period and somewhatearlier?rouglily between 1775 and 1830?that the links between thefully historic tribes on one hand and the partially historic, or pro-tohistoric, tribes on the other will have to be sought.First and most ambitious of the American exploring projects toget under way in the new territory was the Lewis and Clark expedi-tion. It reached the Kansas River on June 26, 1804, and remaineduntil the morning of June 29. Here Clark noted that ? This river receives its name from a Nation which dwells at this time onits banks & [has] 2 villages one about 20 leagues & the other 40 leaguesup . . . they formerly lived on the South banks of the Missourie 24 Leaguesabove this river in a open & butiful plain, and were verry noumerous at thetime the french first Settled the Illinois . . . [Thwaites, 1904-5, vol. 1, p. 60.]Elsewhere (Thwaites, 1904-5, vol. 6, p. 84), in a summary of theIndian tribes met and reported, Lewis and Clark say of the Kansa, ". . . their village is 80 leagues [cf. Trudeau, in Nasatir, 1952, vol.1, p. 261] up the Kanzas River . . ."; and again, in a table of geo-graphic features of the Kansas River, they list "Bluewater river andthe present village of the Kanzes just below . . ." Their "Bluewaterriver" is, of course, the Blue River of today, and the last item citedwould place the Kansa village a short distance east of present Man-hattan, Kans. Wbdbl] an introduction TO KANSAS ARCHEOLOGY 39Continuing up the Missouri, Lewis and Clark camped on July 1st ". . . on the lower point of one of the two large & 2 small IsdsCalled Isles des Parques or field Isl'ds . . . one of the french handsSays 'that the french kept their cattle & horses on those Islands atthe time they had in this quarter a fort & trading establish-ments.' " This was 291^ miles above Kansas River, about op-posite present Leavenworth. Next day, 11 miles farther up, they"camped after dark on the S. S. above the ["Bear Medesin," nowKickapoo] Island & opposit the 1st old village of the Kanzes, whichwas Situated in a valley, between two points of high land, and im-mediately on the river bank, back of the village and on a riseingground at about one mile The French had a garrison for some timeand made use of water out of a Spring rimning into Turkey [nowPlum] Creek. . . ." Next day, July 3, they passed Cow Island ; andJuly 4 they "came to and camped in the lower edge of a Plain wherethe 2nd old Kanzas village formerly Stood, above the mouth of aCreek 30 yds wide this Creek we call Creek Independence . . ." Onthe following day, they "proceeded on near the bank where the oldvillage stood for two miles . . . The orrigan of this old village isuncertain M. de. Bourgmond, a French officer who Com*^ a fort nearthe Town of the Missouris in about the year 1724 and in July of theSame year he visited this village . . ." Both of these ancient vil-lages are indicated on the expedition map, as is the inhabited Kansavillage at the mouth of the Blue.Two years later, Capt. Zebulon M. Pike traveled from the Osagevillages on the river bearing their name to the upper valley of theArkansas by way of the Republican River Pawnee (Pike, 1810).His party left the Osage towns in present Vernon County, Mo.(Chapman, 1946, p. 16), early in September, 1806, and enteredKansas in present Bourbon County. Thence, marching west bynorth. Pike crossed the Grand (Neosho) River and several headwatercreeks of the Verdigris. September 11 to 12, he crossed the FlintHills, camping successively on the South Fork of Cottonwood Riverand on the Cottonwood above present Cottonwood Falls. Approach-ing the latter point, the party "Passed very ruff flint hills. My feetblistered and sore. I stood on a hill, and in one view below me sawbuffalo, elk, deer, cabrie [antelope], and panthers. Encamped onthe main branch of Grand river, which had very steep banks andwas deep. . . . The Indians [Osages accompanying Pike] allegedit was the Kans hunting ground, therefore they would destroy allthe game they possibly could . . ." (Pike, 1810, p. 136). Despitethe apparent abundance of game, no Indians were seen ; but on Sep-tember 15, while he was traversing the headwater creeks of the Cot-tonwood in present Marion County, Pike ". . . passed a very large 40 BUREAU OF AMERICAN ETHNOLOGY [Bull. 1T4Kans encampment, evacuated, which had been occupied last sum-mer . . ." (ibid., p. 137). Crossing the Smoky Hill above presentSalina, the expedition turned northward across the Saline, Solomon,and lesser streams to reach on September 25 the village of the Repub-lican Pawnee at what is today known as the Hill Site on the southbank of the Republican in Webster County, Nebr.**Leaving on October 7 the village of the Pawnee, which had beenvisited a few weeks previously by a large force of Spanish under Mal-gares. Pike marched south by west to strike the Arkansas somewhereon the upper leg of its great bend. This stream he ascended to themountains. Curiously enough, no Indians were encountered after heleft the Pawnee town, although he was told that ". . . the Tetaus[Comanche] had recently killed six Pawnees . . ." Old campfires,probably left by Indian hunting parties who were not identified, werenoted at several places along the route up the river.Pike's "Chart of the Internal Part of Louisiana" (Pike, 1810), sup-plies additional information. The "Kanses Village" is shown on thenorth side of "Kanses River," immediately above (sic) the "BlueEarth River" [Blue River]. On the west bank of the Missouri, aboveKansas River and below the "Nodowa R" [Nodaway], are two siteslabeled "Old Vill Kans," both of which were reported by Lewis andClark and appear again in later journals and maps of the region. The "Kanses crossing place" from the Smoky Hill crosses the Arkansas justbelow a "Strong Saline" [Rattlesnake Creek?]; and the "Kanseshunting Ground" lies south of this on the north side of the "NegrackaRiver" [here evidently the Ninnescah River]. The "Crossing placeof the L. Osage" is not far below the mouth of Pike's Negracka. The"Pawnee hunting Ground" is also south of what is evidently meant tobe the great bend of the Arkansas, some distance west of the Kansahunting grounds. The Osage villages, as already indicated, were out-side present Kansas, on the Osage River in Vernon County, Mo. ; butan Osage camp is shown on an unnamed tributary of the Arkansas,between that stream and the Vermillion [i. e.. Verdigris]. For whatis now western Kansas, no tribal locations are given ; but Pike's generalobservations on the "internal parts of Louisiana" contain the followingpertinent comment (Pike, 1810, app. to pt. 2, p. 17) : The Tetaus or Camanche, as the Spaniards term them, Padoucas by the Paw-nees, are a powerful nation, which are entirely erratic, without the least speciesof cultivation, and subsisting solely by the chase. But their wanderings are con-fined to the frontiers of New Mexico on the west : the nations on the Lower Redriver on the S, the Pawnees and Osage on the R, and the Utahs, Kyaways, andvarious unknown nations on the N. This nation although entirely in our terri- ?My reasons for believing that the Hill site, not the Kansas Monument site southwestof Republic City, Kans., Is the scene of Pike's visit with the Pawnee have been set forthin Wedel, 1936, pp. 35-36. Wbdkl] an introduction to KANSAS ARCHEOLOGY 41 tories, is claimed exclusively by the Spaniards, and may be said to be decidedlyin their interest . . .Five years after Pike's trip, in the summer of 1811, Maj. George C.Sibley, factor at Fort Osage, made a 2-month excursion to the Indiancountry. According to Sibley, the Konsee (Kansa) town wassituated ? immediately on the north bank of the Konsee River, about 100 miles by Itscourse above the junction with the Missouri ; in a beautiful prairie of moderateextent, which is nearly encircled by the river; one of its Northern branches(commonly called the Republican fork, which falls in a few hundred pacesabove the village) and a small creek that flows into the north branch. On thenorth and southwest it is overhung by a chain of high prairie hills which give avery pleasing elfect to the whole scene.The village consisted of about 128 lodges, covered with skins, bark,and mats; patches of corn, beans, and pumpkins, were scattered invarious directions about it; and the adjacent valley bottom was cov-ered with horses and mules (Sibley, 1927, p. 198). The territoryclaimed by the tribe, he says, was "all that which is intersected by thewaters of their beautiful river. It affords as yet abundance of game,and is supposed to be rich in fine furs."From the Kansa town, Sibley marched "North 40? West about 120miles to the Pawnee Towns." These were two in number, situated "on the north bank of the north branch (commonly called Otto fork)of the Platte, about 200 miles above its mouth." The Pawnee Re-publican town, which probably stood at what is today known as theHorse Creek site, 9 miles southwest of Fullerton, Nebr., on the LoupRiver (Wedel, 1936, p. 36), was inhabited "by three tribes of thePawnees, two of which formerly dwelt on the north branch of theKonsee River, about 50 miles in a direct course, above the Konseevillage. The successive incursions of the Konsees obliged them toabandon their old towns about two years ago . . ." This village con-sisted of about 170 earth lodges, but was expected to be about twicethat size when the newly arrived people completed construction oftheir habitations. The Pawnee hunted and raided over a very ex-tensive territory, said Sibley, and waged "an unceasing warfare againstthe Itans (or Hie-tans) from whom they plunder an incredible numberof horses and mules, and many of these in turn fall into the hands ofthe Osages and Konsees, Ottos and Missouris, by theft and pur-chase. . . ."From the Pawnee town, Sibley marched south by east an estimated175 miles to a Little Osage hunting camp "on the bank of the Ar-kansas." En route, between the Kansas and Arkansas Rivers, hestopped briefly at a Kansa hunting camp, about 5 hours' journey fromthat of the Little Osage. Crossing the Arkansas with the LittleOsage, Sibley then rode 30 miles southeast to the Great Osage hunt- 42 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174ing camp ; and next day another 20 miles in the same direction to thehunting camp of the Arkansas Osages. Sibley noted that "the Osages(or as they call themselves Wa-shash-ees) are separated into threedistinct tribes. The Cha-neers who live on the Arkansas, the Bar-har-che (or Great Osages), and the Eu-jet-ta (or Little Osages) whodwell on the Osage River . . ." Rejoining the Little Osage band,Sibley traveled with them to the Great Saline, on the Salt Fork ofthe Arkansas, and then to the Rock Saline, farther west on the Cim-arron.Sibley's estimates of distances traveled are unfortunately not veryaccurate. The Kansa village, if situated as he says at the mouth ofthe Republican, was much more than 100 miles by the river fromits mouth. By water, the Republican is more than 170 miles fromthe Missouri, the Blue about 150. If "by its course" meant merelyalong the valley of the Kansas, not necessarily following the manyturns of that stream, the Blue enters about 112 miles from the Missouri,and the Republican at 136 miles. The Blue River would thus becloser to his estimated distance from the mouth of the Kansas, andthe terrain as he describes it would also seem to fit this location some-what better. Still, Sibley was no newcomer to the Missouri Riverregion, and he should have known what stream he was on when hereached the Kansa village. Regardless of the exact location of theKansa, Sibley could not have reached the Pawnee towns on the Loupby 120 miles' march. Airline from the mouth of either the Blue orthe Republican to Nance County, Nebr., where the Pawnee townsstood, the distance is nearly 150 miles; marching distance would bestill greater.Thus, Sibley's 1811 account indicates that the Kansa were on thenorth bank of the Kansas River, either at the mouth of the Blue orat the mouth of the Republican; the Republican Pawnee were inpresent Nebraska, having recently moved from the Republican Riverto the north bank of the Loup ; the Kansa hunting camp was in east-central Kansas, just north of the Arkansas, perhaps on Cow Creekor the Little Arkansas; and the Little Osage hunting camp wasprobably on the east bank of the Arkansas in what is now southernKansas. The Great Osage and Arkansas Osage may have been northof the Kansas-Oklahoma line, too; but the data are too vague toindicate certainly where their camps were situated.The narratives of Maj. Stephen H. Long's exploring expedition tothe Rocky Mountains in 1819-20 give us our next glimpse of theIndians of the Kansas region. Traveling up the Missouri River onthe steamboat Weste?^ Engineer, Long arrived at Fort Osage (nearpresent Sibley, Mo.) on August 1, 1819; and on the 6th, he dispatchedan overland party of 13, in charge of Professor Say, to visit the Wbdel] an introduction to KANSAS ARCHEOLOGY 43Kansa Indians while the steamboat carried the rest of the partyupstream. Say reached the Kansa village of 120 earth lodges onAugust 20, and spent several days there ; his eyewitness accounts areperhaps the best contemporary statement we have regarding thistribe. On the 24th, Say's party left the Kansa, traveling north "along the margin of Blue Earth creek [Blue River], a stream ofthe width of 25 yards, and greatest depth of three feet, which dis-charges into the [Kansas] river a mile or two above the Konzavillage . . ." (James, 1823, vol. 1, p. 131). Seven or eight miles fromthe village, they were set upon and robbed by a war party of Republi-can Pawnee, 140 in number, whose ". . . nation was at war withthe Konzas . . ." (ibid., p. 135). Say returned to the Kansa vil-lage, resupplied himself, and then marched overland northeast to Isleau Vache, arriving there after departure of the Western Engineer,which his men finally overtook at the mouth of Wolf River. Twoformer village sites of the Kansa had been noted, meanwhile, by theriver party as it proceeded upstream from Fort Osage. Some 24miles, or slightly more, above the mouth of Kansas River, the narra-tive states (James, 1823, vol. 1, p. 110) that "The site of an old villageof the Konzas, and the remains of a fortification erected by theFrench, were pointed out a few miles below Isle au Vache" [CowIsland is 7 miles above Leavenworth, in southeastern AtchisonCounty, Kans.]. On the 25th, the boat left Isle au Vache and theparty ". . . proceeded in the course of the day about 23 miles, andencamped at night near the entrance of a small stream called Inde-pendence Creek. A little above, and on the south [right] side of theriver, is the site of an old Konza town, called formerly the villageof the Twenty Four . . ." (James, 1823, vol. 1, p. 113).^ Onlythe second of these two sites, designated "Old Konza Vill.," is shownon the map that accompanies the James edition of Long; it is justabove "Independence R.," at present Doniphan, Kans. The inhabited "Konzas Vill." visited by Say is shown just below the mouth of "Blueearth R." on the north bank of Kansas River east of present Man-hattan.A year later, on his return from the Rocky Mountains in the sum-mer of 1820, Major Long descended the Arkansas to a point a fewmiles above the mouth of the Purgatoire River. Here he dividedhis men; half the party, under Captain Bell, thence descended theArkansas on its north bank, while the remainder under Major Long,marched south and east to Red River. Bell and his party of 11 mencamped the night of July 25 about opposite, and slightly below, themouth of the Purgatoire. In the afternoon of the 26th, probably ' Hyde, 1951, p. 48, suggests that this village was so designated because of Bourgmond'scouncils here in 1724, 44 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174 west of present Big Sandy Creek and near or at the upper end ofwhat was later termed the Big Timber, Bell came upon a large tipicamp ? consisting of Kiawas, Kaskaias or Bad-hearts [Kiowa Apache], Shiennes (some-times written Chayenne), and Arrapahoes. . . . These nations have been forthe three past years, wandering on the head waters and tributaries of Redriver, having returned to the Arkansa, only the day which preceded our firstinterview with them, on their way to the mountains, at the sources of thePlatte river. They have no permanent town, but constantly rove, as necessityurges them, in pursuit of the herds of bisons, in the vicinity of the sources ofthe Platte, Arkansa, and Red rivers. . . . Their martial weapons are bowsand arrows, lances, war clubs, tomahawks, scalping knives, and shields . . .[James, 1823, vol. 2, pp. 174-187].Two days later, at noon on the 30th, Bell's detachment encounteredan Arapaho war party, including one Kiowa, in present HamiltonCounty, Kans. On August 1, near present Lakin, Kearney County,they met "... a Shienne war party on their return from an expedi-tion against the Pawnee Loups. They had killed one squaw . . . Theparty was armed with spears, bows and arrows, war-clubs, tomahawks,scalping knives, &c . . . " (James, 1823, vol. 2, p. 197). No meetingswith Indians or their traces are reported during the next 10 days.On August 11, Bell's party camped opposite the mouth of RattlesnakeCreek, in southwestern Rice County; and shortly after noon of thenext day, still above Cow Creek, they met "... an letan or Comanche(a band of Snake Indians) war party, 35 in number, of whom 5 weresquaws. They had marched to attack the Osages . . .," but weresurprised and soundly beaten by a small Oto war party. Bell notedfurther that "They were armed with the bow and arrow, lanceand shield, and 13 guns, but by far the greatest number carried lances ..." (James, 1823, vol. 2, pp. 207-211). Leaving this party aftersome difficulty. Bell continued down the Arkansas; and on the 15th,near its confluence with the Little Arkansas, he saw "... the appear-ance of an Indian village [that] proved to be a large hunting camp,which had probably been occupied during the preceding season . . .much bark covered the boweries, and a few pumpkins, watermelons,and some maize ..." were observed (James, 1823, vol. 2, pp. 214-215) . Next day they camped opposite "... the entrance of a large creek, ofthe width of 90 or 100 yards, and of considerable depth; it seems tobe well wooded, and its course is nearly parallel to the river for agreat distance, before it discharges into it. This stream is called theRed Fork; its waters are turbid, opake, and red . . ." On the map,this is called the "Negi^acka or Red Fork," and the journal leaves noroom for doubt that Bell was camped opposite the mouth of thepresent Ninnescah River. This does not parallel the Arkansas, butthe Cowskin just above does ; and it is possible that from Bell's camp Wbdbl] an introduction TO KANSAS ARCHEOLOGY 45the timber along the Cowskin looked like a continuation of the valleyof the Ninnescah. Two days later, on August 18, he crossed the "Stinking Fork," now Wahiut River, near present Arkansas City,and camped 8 miles farther on the east side of a "small creek whichwe call Little Verdigris [now Grouse Creek]." Along this creekthey also saw Indian lodges?bark-covered and some 60 or 70 innumber?as well as cornfields, trails, etc. The lodges were unoccu-pied, but were presumed to belong to the Osage. On the followingday, the 19th, the party crossed the present Kansas-Oklahoma Stateline and continued down the river.Maps resulting from the Long expedition provide some furtherinformation. That which accompanies the James edition, in additionto the Kansa locations already alluded to, shows the Osage villages ? Great and Little?on a large southerly tributary of Osage River.This, I presume, is either the Little Osage or the Marmaton, whichwould place the villages in present Vernon County, Mo., where Pikevisited them in 1806. Clermo's Band of Osage is shown on the eastside of what is evidently the Verdigris River, south of parallel 37? N.The High Plains about the heads of the Solomon, Saline, Smoky Hill,Arkansas, and southward are labeled "Great American Desert" ; andbetween the "Solomon's Fork" and the "Republican Fork" is thelegend : "The Great Desert is frequented by roving bands of Indianswho have no fixed places of residence but roam from place to placein quest of game." The large original map prepared for Long butnever published (U. S. 62, Cartographic Records Branch, NationalArchives) shows a second "Old Kanzes Vil[lage]" on the west bankof the Missouri below Cow Island, roughly midway between KansasRiver and Independence River; but adds nothing regarding theIndians in central and western Kansas.The report of the Reverend Jedidiah Morse to the Secretary ofWar on the condition of the Indians in 1820 includes short summariesof the tribes in our area. Quoting from a letter, dated Fort Osage,1st October, 1820, from G. C. Sibley, factor, to Thomas L. McKenney,Morse states that the Kansa, numbering about 800 souls, resided inone village some 300 miles up the Kansas River, and that they huntedthroughout its drainage and along the Missouri to the Nodoway.The Great Osages of the Osage River, estimated at 1,200 persons, livedin one village on the Osage River, 78 measured miles due south ofFort Osage, hunting in the Osage, Gasconade, and Neeozho [Neosho]River drainages. The Great Osages of the Neeozho, about 400 strong,had a village on the Neosho, "about 130 or 140 miles southwest of FortOsage"; they hunted with the tribe of the Osage river, from whom "they separated six or eight years ago." The Little Osage were inthree villages, totaled about 1,000 souls, and resided "on the Neeozho 46 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174Kiver, from 120 to 140 miles southwest of this place." They hmitedwith the other Osage, and "frequently on the head waters of theKansas, some of which interlock with those of the Neeozho" (Morse,1822, pp. 203-204).Elsewhere in the same report, Morse (1822, p. 237) gives the popula-tion of the Kansa at 1750, apparently on the authority of Maj. Benj.O'Fallon, Indian agent at Council Bluffs, and says that they "live onthe northwest side of Kanzas river, at the mouth of the Grand Saline . . ." ; their hunting grounds were on the upper branches of the river,from which the game was fast disappearing. The accompanying mapshows the Kansa village on the north bank of the Kansas River, justabove the "Earth River" whose upper course carries the name "GrandSaline." The Republican Fork is shown much farther west, and the "Earth River-Grand Saline" is without much doubt the Blue River.An "Old Kansas Vill" is shown on the right bank of the Missouri somedistance above the Kansas River.Farther west, according to Morse, were the Arrapahays (Arapaho),whose country extended "from the head waters of the Kanzas, south,to the Rio del Norte. They are a war-like people, and often makingpredatory and murderous excursions on their eastern and southernneighbors" (Morse, 1822, p. 253). Here, too, were the "ComauchIndians" (Comanche) who were "the largest and most war-like nationin this country. . . . During the winter, they occupy the country on thesources of the Brassos, and Colerado. They spend their summers onthe sources of the Arkansaw and Missouri, among the eastern spurs ofthe Rocky Mountains. They carry on with traders from Red River, anextensive traffic in horses and mules, which they catch in the plains,or capture from the Spaniards" (ibid., p. 259). Other groupsscattered over the western territory included the Castahana, Cataka,and Dotami, "supposed to be remnants of the Great Padouca nation,now under that name extinct, who occupied the country between theupper parts of the Platte and Kanzas river" (ibid., p. 366).Insofar as recorded native tribal locations and movements mightbe expected to have archeological connotations, our chronological sur-vey of Kansas ethnohistory here approaches the point of diminishingreturns. There are several reasons for this. For one thing, thenative peoples along the lower Missouri and its tributaries, includingthose in Kansas, had been in increasingly close contact with whitetraders for more than a century, and the disintegration of native tech-nologies, crafts, and industries was unquestionably far advanced by1825. Most of these Indians were making little or no pottery at thisdate; iron, brass, glass, crockery, steel traps, and machine-made tex-tiles were rapidly superseding the older items of bone, stone, horn,shell, and other locally available materials. It is unlikely therefore WEDKL] AN ESTTRODUCTION TO KANSAS ARCHEOLOGY 47that archeology could greatly amplify the data already available, orthat yet to be extracted, from contemporary ethnographic and historicdocuments, published and unpublished; and I doubt that such con-firmatory data as archeological investigations could perhaps add onthis time level would greatly facilitate extension of tribal and culturalidentifications back into earlier time levels. From here on, in short,the Indians of Kansas become the proper subject of the ethnographer,historian, and archivist, rather than of the archeologist.Equally important is the fact that the year 1825 brings us to thethreshold of the reservation period and, thus, to a series of whollyartificial tribal alinements. In the States east of the Mississippi,there was a growing public and official insistence on final expulsionof the Indian tribal remnants still residing there; consummation ofthe Louisiana Purchase, with acquisition of a vast and sparsely popu-lated new territory provided an obvious solution to the problem oflocating the Indians who were to be moved. As a prelude to transferof these eastern Indians to the trans-Missouri region, it was necessaryfirst to extinguish the primary land titles of the tribes residing there,thus restricting and defining their holdings and making available landsto be assigned to the newcomers. In line with this policy, the Govern-ment entered into a treaty with the Kansa and Osage in 1825 wherebythe territory claimed by each was drastically reduced. By this trans-action, the Osage became for a brief time a Kansas tribe, which theyhad been only in part previously. Into the lands given up by theKansa and Osage came, within the next 15 or 20 years, a whole host ofIndians formerly residing, mostly but not altogether, east of theMississippi. Reservations were assigned to such diverse groups as thefollowing : Oto, Missouri, Iowa, Sac, Fox, Kickapoo, Delaware, Shaw-nee, Chippewa, Ottawa, Peoria, Kaskaskia, Weas, Piankeshaw, Pota-watomi, Miami, Wyandot, Cherokee, and New York Iroquois. Ex-cepting those of the Kansa and Osage, the reservations assigned thesetribes lay generally in the eastern third of the State, east of a linerunning north-south through the mouth of the Republican River.Some of the eastern reservation Indians, at least, were occasionallyrepresented in hunting parties in central and western Kansas, wherethey came into conflict with the Pawnee, Cheyenne, Arapaho, Kiowa,and Comanche. For obvious reasons, none of these introduced groupsis of interest directly or indirectly, in the prehistory of the Kansasregion. HISTORIC INDIAN TRIBES IN KANSASThe documentary record, as we have just reviewed it, suggests thatthe historic tribes who have been resident in what is now Kansas longenough and early enough to be of possible archeological significance484172?59 5 48 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174number not more than five or six. This does not seem like a veryimpressive figure, considering tlie plethora of tribal and band nameswith which Spanish and French observers fill their narratives andcartographers besprinkle their maps during the 16th, I7th, and 18thcenturies. Some of these names can be, and have been, reconciled withone another ; oftentimes, but not always convincingly, they have beenfurther identified with one or another of the recognizable designationsfor tribes well known in later historic times. Granting the pos-sibility that minor tribal units may have been slighted or overlooked,it nevertheless appears that the actual number of tribes that can besaid to have laid valid claim to territory in Kansas is not much, if any,in excess of the figure suggested. It also seems clear that these tribes,besides representing several distinct linguistic stocks, were distributedin rather close accord with the several natural regions that may berecognized in the State.Western Kansas, geographically better suited to hunting than tohorticulture, was successively occupied in historic times by Athapas-can, Shoshonean, and Algonkian tribes. When the historical recordopens in the 16th century, Plains Apache peoples appear to have beenpresent here as elsewhere throughout the High Plains; and thesegroups were present until the first quarter of the 18th century. At thattime the Shoshonean-speaking Comanche spread eastward from theRocky Mountains, driving the Apache southward and taking overthe High Plains. In the 19th century, the Cheyenne and Arapaho werethe principal tribes between the Arkansas and Platte Rivers, with theComanche, Kiowa, and Kiowa Apache mainly south of the Arkansas.Farther east, along the border of the High Plains, the first Spanishexplorers found the people of Quivira, a semisedentary group whowere almost certainly the Wichita or a closely related Caddoan group.As the Paniassa, they seem to have been in southern Kansas ornorthern Oklahoma into the 18th century. Still farther east, certainlyby the last quarter of the 18th century, were the Kansa, and southeastof them, near the Missouri-Kansas line were the Osage, closely relatedtribes of Dhegiha Siouan affiliations. Both remained in or near thisregion throughout the 19th century, moving steadily southward andwestward to their final residence on reservations in Oklahoma. Latein the 18th century, there was at least one Pawnee village in northernKansas, on the Republican River, and much of northern Kansas re-mained the hunting ground of the Pawnee.SIOUAN TRIBES IN KANSASThe two Siouan tribes, Kansa and Osage, who merit considerationin connection with Kansas Indian history and prehistory were bothof the Dhegiha branch. Their earlier history, prior to the first French Whdbl] an introduction TO KANSAS ARCHEOLOGY 49 notices concerning them, is vague and uncertain. According to often-cited migration legends collected by Dorsey (1886, pp. 215-218), theDhegiha Siouans long ago dwelt east of the Mississippi Eiver, whencethey pushed westward. As one nation, they were known to the Illinoisas the Arkansa or Alkansa, and lived near the Ohio. At the mouth ofthis stream, the Quapaw split off and moved southward; the othertribes, including the Omaha, Ponca, Osage, and Kansa ascended theMississippi, stopped for a time at the mouth of the Missouri, and thenmoved up that stream. At the mouth of the Osage, the Omaha andPonca crossed the Missouri ; the Osage ascended the stream which ac-quired their name, dividing into the Great and Little Osage at an un-named tributary and finally settling down. The Kansa continued upthe Missouri on its south side until they reached Kansas River, wherethey tarried for a time. Resuming their wanderings, they ascended theeast side of the Missouri to the present northern boundary of the Stateof Kansas, where the hostile Cheyenne compelled them to retrace theircourse. They then established themselves at the mouth of the Kansas,where the "Big Knives" came with gifts and induced them to gofarther west.Beyond the vague and non-committal "ages ago," there is nothingin this legend to indicate how much time may have been involved inthe migrations set forth. Neither is there at present any archeologicalconfirmation that the Ohio Valley was the homeland of the Siouans,as the traditions claim. The available linguistic and ethnologic evi-dence shows, however, that the Kansa and Osage tribes were closelyrelated, and their separation need not have taken place very long be-fore their appearance in history. The habitat of the Osage since thebegmning of records concerning them w^as in the area to which thetraditions bring them and where they resided until approximately1820. As for the Kansa, there is at the moment no independentconfirmation of an early stay at the Missouri-Kansas River junc-tion; but what the legends have to say regarding their residence onthe Missouri above the Kansas, but not farther north than the Kan-sas-Nebraska boundary, prior to their settlement on the KansasRiver, is in line with their documentary history since at least 1723.That they ever dwelt on the east side of the Missouri has not yetbeen demonstrated on historical or archeological grounds. It maybe significant, however, that Chouteau in 1816 described their earlierhunting grounds as lying both east and west of that stream, whereasin historic times they seem to have hunted mostly west of the Mis-souri (Chouteau, 1816, and map in Tucker, 1942, pi. 41). This, plusthe fact that there are reported to be sites of late prehistoric or 50 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174protohistoric age and of Oneota affinities on the east bank of theMissouri below Nodoway River, may hint at former residence of theKansa, or of one of their near relatives, in that locality.THE BLA.NSAOn documentai-y grounds, the Kansa seem to have been resident inthe State about as long as any of the historic groups; and they canjustifiably be termed primarily a Kansas tribe. I have elsewhere pub-lished a sketch of the Kansa (Wedel, 1946), and here we need onlyreview briefly their recorded movements (fig. 4) . The earliest refer-ence to them, so far as I have been able to determine, is by the Frenchin the second half of the 18th century?specifically, on maps resulting Figure 4.?Principal known Kansa village sites in Kansas during the 18th and 19th cen-turies. Numbered sites include: /, Doniphan site, the "grand village des quans" visitedby Bourgmond, 1724; 2, Salt Creek site, near old Fort Cavagnolle, 1757; 3, the "oldKanzas village" reported by Lewis and Clark, 1804, exact location unknown; 4, BlueRiver site, 14P024, near Manhattan, occupied ca. 1800-1830; 5, Fool Chief's village,1830-46; 6, Hard Chief's village, 1830-46; 7, American Chief's village, 1830-46; 8, HardChief's village, 1847-73; 9, Fool Chief's village, 1847-73; 10, Big John village, 1847-73.Lettered sites include A, Leary site, and B, Fanning site, both Oneota and antedatingthe Kansa sites shown by number. Wmuel] an introduction TO KANSAS ARCHEOLOGY 51from the Marquette and Jolliet expedition of 1673. If the Spanishhad prior knowledge of the tribe, I have found no convincing evi-dence of the fact. Earlier allusions by the Spanish to what was sup-posed to be the Kansa seem to me to be either inconclusive or highlyimprobable. The Guas, for example, to whom Castaiieda says FatherJuan de Padilla was trying to extend his missionary efforts fromQuivira when he was martyred in 1542 or soon after (Hammond andKey, 1940, p. 263), have been identified with the Kansa (Hodge,1907-10, pt. 2, p. 1058) ; but in the absence of any details, beyond thename and the assertion that the Guas were enemies of the people ofQuivira, I fail to see how a convincing correlation is possible here.Again, the Escansaques met by Oiiate in 1601, apparently somewherein present southern Kansas, have been cited as "the earliest recordednotice of the Kansa" (Hodge, 1907-10, pt. 1, p. 653) ; but the briefdescription Oiiate gives of this nomadic, bison-hunting, tent-dwellingpeople sounds much more like one of the Plains Apache groups whomthe contemporary Spanish were calling "Apaches Vaqueros." I havegrave doubts, too, that the still later Spanish references to the Can-ceres, allies of the French, as given in Martinez' declaration of No-vember 23, 1720, "clearly indicate," as Thomas says (1935, pp. 171,277) , the location of the Kansa in the early 18th century. Such a lo-cation would put the Kansa in southeastern Colorado or southwesternKansas at just about the time Bourgmond and La Renaudiere clearlysay that the Kansa town was on the Missouri River in the northeasternpart of Kansas, and more than 400 miles from the Sierra Blanca. Itseems much more probable that the Canceres were also Plains Apache(Hodge, 1907-10, pt. 2, p. 1036) . The earliest documented Kansa village evidently stood at the junc-tion of Independence Creek and the Missouri, on the present site ofDoniphan, Kansas (Remsburg, 1919), about 60 miles by water abovethe Kansas River. Here the Kansa are shown on the Delisle map of1718 (Tucker, 1942, pi. 15), were reported by La Renaudiere in 1723(Margry, 1886, pt. 6, p. 393) , and were visited the next year by Bourg-mond (ibid., pp. 398-449). Whether they were already at this loca-tion half a century earlier, in Marquette's time, when they first emergeinto recorded history, we have no way of knowing, nor have I foundany record as to when the spot was abandoned. In the 1740's and1750's, the Kansa were farther downstream near the mouth of SaltCreek, a few miles north of present Leavenworth, and about 35 milesabove Kansas River (Remsburg, 1919) ; and here for a time stood aFrench post. Fort Cavagnolle. Whether this village was continuouslyinhabited and over what period is again obscure, but the records hintat occasional absence of the Indians from the vicinity of the post(Nasatir, 1952, vol. 1, p. 48) . From here, sometime after 1757 or 1777, 52 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174the Kansa moved west and south to the Kansas River ; so far as I candetermine, they did not again reside in a permanent village on theMissouri. Their oldest identified village site on Kansas River is thatat the mouth of the Blue River, about 2 miles east of the present Man-hattan. In this locality the Collot map of 1796 (Collot, 1924) and theDuLac map of 1802 (DuLac, 1807) place the tribe; Trudeau, writingin 1794, has the Kansa 80 leagues up the Kansas, which would placethem far above the Blue (Nasatir, 1952, v. 1, p. 261). If sites earlierthan this at the mouth of the Blue actually exist on Kansas River, theyhave apparently not been located or identified. According to Lewisand Clark (Thwaites, 1904-5, vol. 1, p. 60) , the Kansa had two villagesin 1804," . . . one about 20 leagues and the other 40 leagues. . ." upthe Kansas. Elsewhere, however, in a summary table of various fea-tures of the Kansas River system (Thwaites, 1904-05, vol. 6, p. 36),they note "The old Kanzas Village" on the north bank 9 miles above "Heart River" (Soldier Creek) and 40 miles from the Missouri; and "Bluewater river and the present village of the Kanzas just below" aresaid to be 80 miles up the Kansas. The latter reference is clearly tothe Blue River location ; the old village is supposed to have stood some-where in the vicinity of present Silver Lake, between Soldier Creekand the Red Vermillion. Strangely enough, the Lewis and Clark mapdoes not show the old village below the occupied one ; but it does indi-cate an "Old Konza Vill" on a narrow neck of land between the Re-publican and Kansas Rivers immediately above their junction. Thiswould appear to be on the spot now occupied by Junction City. So faras I am aware, these two old villages have never been located.The Blue River village is supposed to have been the principal settle-ment of the Kansa for some 30 years or more, but it may not have beenthe only one. According to McGee (1897, p. 193), a Kansa village of1500 people occupying 30 lodges stood at the mouth of Saline River,where "After the cession of Louisiana to the United States, a treatywas made between the Indians and the government"; but I can findno independent confirmation of this location. Again, Sibley in 1811(Sibley, 1922) locates the Kansa in 128 lodges just below the mouth ofthe Republican "about 100 miles by its course above the junction withthe Missouri." The Republican joins the Kansas 135 miles by landabove its mouth ; and Sibley's description of the village surroundingswould fit the Blue River site as well as, if not better than, the junctionof the Republican and Kansas. A few years later, Morse (1822, p.237) has the Kansa at the mouth of the "Grand Saline River," thoughhis map shows them immediately above the "Earth River" whose uppercourse is labeled Grand Saline. Here there is manifest confusionabout the tributaries of the Kansas River (see also Melish, 1818, andBradbury, 1817) , but the location given on the Morse map can hardly Wedkl] an introduction to KANSAS ARCHEOLOGY 53have been anything else than that at the mouth of the Blue River. Iknow of no contemporary map for the period from 1800 to 1830 whichlocates the Kansa above the mouth of the Blue, variously known asBlue Earth, Blue Water, etc.Following their treaty with the United States in 1825, the Kansabegan a drift eastward, and the Blue River location seems to havebeen finally abandoned by about 1830. In 1834 Townsend (1839, pp.32-34) found them mostly in two villages on both sides of KansasRiver a few miles west of present Topeka, with a third and smallervillage of 30 lodges some distance upstream on the north bank, butbelow the Vermillion River. Four years later, in 1838, Richard Cum-mins, Indian agent, reported the Kansa still in three villages?thetwo lower ones apparently in the vicinity of present Rossville, andthe third some 30 miles upriver (Comm. Ind. Aff., for 1838, p. 478).Whether these were the same locations as those given by Townsend,I do not know. According to Morehouse (1908, p. 346) and Adams(1904), the principal Kansa towns from 1830 to 1846 were a fewmiles above present Topeka (see also Chouteau in Anon., 1881, p.387 fn.). The largest, under Fool Chief, stood on the north bankof Kansas River, about 6 miles above Soldier Creek and just northof Menoken (Cole, 1904, p. 483). Two miles south of the river, onthe west side of Mission Creek, was American Chief's village of 20lodges; 2 miles away, on high land nearer the river and about 11/^miles west of Mission Creek, was Hard Chief's village of 100 lodges.Below Fool Chief's village, where Father De Smet visited in 1841,were the halfbreed allotments; and below them, near present Wil-liamstown, was the first tribal agency, with blacksmith shop andfarmer attached. In this locality was opened, under Methodistauspices, the first mission among the Kansa.?From this locality, following the treaty of January 14, 1846, theKansa moved southwest a few miles to their new reservation on theNeosho River. Here, in the vicinity of Council Grove, they dweltin three villages from 1847 to 1873. Hard Chief's village, largestof the three, was on Cahola Creek south of present Dunlap. FoolChief's village stood in the Neosho valley near the same place. Thethird village was to the northwest, near Big Jolin Creek southeastof Council Grove and near the agency. From these villages of bark-,mat-, and skin-covered lodges, according to Morehouse, the Kansacontinued their seasonal bison hunts into the Plains to the west.Eventually, this resulted in a well-defined trail, which began near themouth of Big John Creek southeast of Council Grove. Thence thistrace, known as the Kaw Trail, ran slightly south of west to cross * Swanton, 1952, p. 293, lists the names of more than 20 Kansa villages, including twoon the Big Blue, one at its mouth, and 12 or more at various points along Kansas River. 54 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174Diamond Creek within a few yards of the Diamond Springs railwaystation. Entering Marion County near the old Bethel post office onthe head of Middle Creek, not far from present Lincolnville, it con-tinued westward through Marion and McPherson Counties to theforks of Cow Creek, a few miles south of present Lyons. The routegenerally paralleled that of the Santa Fe Trail, at a distance of 3 to6 miles to the south. From the hunting camp here in central RiceCounty, some of the Indians returned with their kill to the villageson the Neosho ; others, preferring the buffalograss as forage for theirhorses, spent the winter on the spot.Greatly reduced in numbers by disease and neglect, and under con-tinual pressure from white settlers, the Kansa finally gave up theirlands on the Neosho. In 1873, just 200 years after their first recordednotice, the remnants of the Kansa finally left the Neosho and settledon a portion of the Osage Reservation in northern Oklahoma. Todaywhat remains of this tribe resides in a small tract near the ArkansasRiver in Kay County, Okla. THE OSAGEThe Osage emerge into written history at the same time as theKansa, that is, with the Marquette and JoUiet explorations of 1673.As with the Kansa, it is impossible to locate them precisely at thistime, but they were apparently west of the Missouri Indians andsouth of the Missouri River. For more than a hundred years thelocation of their villages is given by various observers and com-mentators in such general terms that it is not possible to indicatethem accurately on modern maps. We do not know, of course, howpermanent the villages were, but it seems likely that throughout muchor most of this time they were on the Osage River, well above itsmouth, and probably within relatively few miles of the eastern Kan-sas line.^ It is not improbable, therefore, that they left some as yetimdiscovered or unrecognized evidence of their presence, either ashunters or as temporary residents, at the extreme edge of the areawith which this paper is primarily concerned ; and it is a historical ? There Is, of course, wide variation In the estimates given by early writers for thedistance from various points (e. g., St. Louis or the mouth of the Missouri River) to otherpoints up the Missouri ; and, in the present problem, for the distances from the mouth ofthe Osage River to the Osage villages. The following actual figures on the 1890 thalweg,or channel line, distances from the mouth of the Missouri may be helpful : to the mouthof Osage River, 137.6 miles ; mouth of Grand River, 260.7 miles ; mouth of Kansas River,392 miles ; mouth of Big Nemaha, 534.5 miles ; mouth of Platte River, 636.9 miles (letterof 1 December 1953 from L. Q. Fell, Chief, Engineering Division, Kansas City District,Corps of Engineers, U. S. Army, to Wedel).As measured by map gage on the U. S. Geological Survey quadrangle sheets, the dis-tance from the mouth of Osage River upstream to the Little Osage, near which the Osagewere living In 1806, is very nearly 280 miles. At 2.6 miles per league, therefore, theOsage River enters the Missouri 53 leagues above the mouth of the latter ; and the Osagevillages In present Vernon County, Mo., were approximately 100 leagues up the OsageSlTcr. Wbdbl] an introduction to KANSAS ARCHEOLOGY 55fact that they were resident in Kansas for at least 50 years priorto their final removal to the Indian Territory. Moreover, their resi-dence just outside the State but in proximity to other tribes who wereliving in eastern and southern Kansas and their long-continued war-fare against these tribes, makes them a factor that must beconsidered.Although the Marquette and Jolliet maps do not include the OsageRiver, it is possible that already at this time the tribe was living onthat stream. In 1687, according to Hennepin, the Osage occupied 17villages on a river bearing their name, which discharged into the Mis-souri. Tonti in 1692 observed only that they were near the Missourisand Otos in some prairies 150 leagues up the Missouri River. In 1700Tonti wrote that on the banks of the River of the Ozages there were300 lodges of Osage, located 15 day's journey by canoe from the mouthof the Missouri ; from the Osage to the Kansa was a 3-day journey. In1717 Bourgmond mentions "the Osage River because of the tribe whichlives there and bears the same name." Soon after, in 1719, Du Tisnemade the first recorded visit by a Frenchman to the tribe ; he locatesone village 80 leagues above the mouth of Osage River, with another30 leagues distant from the Grand [Osage] village and 1^/2 leaguessouthwest of the Missouri village, the latter evidently on the MissouriRiver in present Saline County, Mo. In the same year, La Harpe lo-cates the Missouri town 80 leagues up the Missouri, with an Osagevillage a league distant and the Grand [Osage] village 30 leaguesaway. In 1769 Rui reported that the River of the Big Osages "goesto the tribe called by the same name, which is some 70 leagues from themouth." Du Pratz in 1757 wrote that the Osage were situated upon asmall river of the same name ; and Cruzat 20 years later reported themon a tributary of the Missouri, distant from St. Louis 180 leagues bywater and 110 overland (Houck, 1909, vol. 1, pp. 142-145).According to Miro, writmg in 1785, the village of the Grand Osagewas then 120 leagues up the river of that name, i. e., the Osage ; and theLittle Osage, of whom a part had removed to the upper waters of theArkansas River, are somewhat ambiguously placed about 80 leaguesfrom the junction (of the Osage?) with the Missouri, about 1 leaguefrom the Missouri village. The implication here would seem to be thatthe Little Osage, or a part of them, were still, or again, on the Mis-souri River near the remnants of the Missouri, but that a portion ofthem had moved southward onto the Arkansas or one of its tributaries.This situation is also reflected in Trudeau's journal of 1794, whichimplies that the Little Osage had recently fled to the Osage River forsafety (Nasatir, 1952, vol. 1, p. 261),Lewis and Clark (Thwaites, 1904-5, vol. 6, pp. 83-84) have theOsage villages 80 leagues up the Osage River, on which stream they 56 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174say the larger part of the Great Osage tribe had resided since theyfirst became known to the inhabitants of Louisiana. Some 3 yearsprior to their visit, about half the tribe had moved to the Arkansasdrainage ; and the Little Osage, who had lived for a time on the south-west side of the IVIissouri Elver near Grand River, had been compelledto seek refuge with the Great Osage. The Osage claimed lands withina boundary running along the Niangua River, thence southward to theArkansas 100 miles below the three forks (Arkansas-Neosho-Verdi-gris?), up the Arkansas and a southerly fork (Cimarron River?) to apoint some distance above the Great Saline (Great Salt Plain, AlfalfaCounty, Okla.), and nearly to Kansas River, thus including a largearea in the southeast part of the present State of Kansas. TheNicholas King map of 1806, based on the Lewis and Clark observations(Tucker, 1942, pi. 31B) , has the Little Osage in the fork between SwanPond River (Marais des Cygnes, or Osage) and Long River (LittleOsage River) ; and the Osage village, of 1,200 warriors, is shown somedistance downstream below Cookesfield River (now Sac River?) andThe Fork (Pomme de Terre River ? ) , which would seem to locate themin present St. Clair County below Osceola.Li 1806 Pike visited the Osage, and his map shows the Great andLittle bands on the upper waters of Osage River, apparently near thejunction of the Osage and Little Osage Rivers. According to hisjournal (Pike, 1810, pp. 116-126), this locality was approximately360 miles by water up the Osage River ; the true water distance is about280 miles. The third band or village of Osage was to the southwest,in the Arkansas drainage. Sibley's observations 5 years later suggestthat the Osage were then distributed in about the same fashion (Sibley,1922).By 1820 the drift from the upper Osage River and out of the Stateof Missouri was well under way. The Osage were then scattered inthree separate localities. According to James (1823, vol. 2, pp. 244,251), on the authority of Sibley, the Great Osage were still near theheadwaters of Osage River; Sibley (Morse, 1822, p. 203) puts them78 measured miles due south of Fort Osage, which would be in presentVernon County, Mo. The Little Osage, along with some of the GreatOsage, were on the Neosho River, evidently in what is now Neosho andLabette Counties, Kansas. The Osage of the Oaks (Chaneers orArkansas Osage) were on the lower Verdigris River, apparentlysome 55 miles from its mouth, in present Oklahoma. According toSibley (James, 1823, vol. 2, p. 248), this latter band had been inducedto move to the Arkansas by Chouteau, prior to the cession of Louisianaterritory to the United States, when Chouteau found himself barredfrom the Osage trade by the monopoly granted Manuel Lisa by theSpanish authorities. Wbdbl] an introduction TO KANSAS ARCHEOLOGY 57The exact locations of the Osage villages on the river bearing theirname, as reported by French, Spanish, and other observers andwriters during the iTth and 18th centuries, have not been conclusivelyestablished; and the generally vague and incomplete statements inthe contemporary documents make this matter of present day identi-fication a formidable problem indeed. At the same time. Chapmanand his colleagues (Berry, Chapman, and Mack, 19M; Chapman,1946, 1952) have made a good case for identification of certain sitesin Vernon and Saline Counties, Mo., with late 18th- and early 19th-century Osage villages. Earlier Osage sites, if not in the samelocalities, apparently still await discovery or identification.In 1825 the Osage ceded the lands they claimed in and west ofMissouri, in Arkansas and Indian territories, and south of KansasEiver. They reserved a strip 50 miles wide and some 75 miles long,beginning 25 miles west of the Missouri line and running west alongthe southern boundary of Kansas (Eoyce, 1899, p. Y08). Apparently,they had already moved entirely out of Missouri by 1822 (Berry,Chapman, and Mack, 1944, p. 3), congregating on the Neosho Riverin southeastern Kansas. The exact location of the new villages seemsuncertain; but they are thought to have been scattered along theNeosho from present Oswego, Labette County, upstream to approxi-mately Erie, Neosho County (McDermott, 1940, p. 127, n. 28; Abel,1904, p. 77 fn.). In this locality, on the east bank of the Neosho atpresent St. Paul, a Catholic mission and school were established inor about 1846. Also in this general locality, Tixier visited the Osagein 1840 (McDermott, 1940), found them living in bark- and mat-covered lodges, and accompanied them on their summer buffalo huntto the treeless plains west of the Arkansas and on to the Great Salineon the Salt Fork of the Arkansas. On this hunt, the Osage traveledfrom the Neosho to the Verdigris ; from this stream the route is notdetailed but very likely was close to the Osage trail described byMead for a somewhat later period. This is said to have left theVerdigris near the mouth of Fall River in southern Wilson County,running northwest to the Walnut near present Eldorado, and on tocross the Arkansas a few miles above present Wichita.After some 40 years' residence on the Neosho, the Osage in 1865relinquished to the United States a 30-mile block at the east end oftheir reservation, where the school, mission, and agency establislmientsstood, and moved 40 or 50 miles west to the Verdigris drainage.Here six or seven new villages sprang up ; they consisted of skin- ormat-covered lodges, surrounded by extensive cornfields tilled withhoes by the women and children. Their stay here was a short one,however. White settlers, taking advantage of the absence of theOsage on their periodic buffalo hunts, steadily encroached on the 58 BUREAU OF AMERICAN ETHNOLOGY [Bcli,. 174lands set aside for the Indians and stubbornly resisted eviction. Thebison, moreover, were disappearing from the Western Plains beforethe increasing inroads of professional hunters and Indians equippedwith firearms, so that the traditional tribal hunts were less and lesssuccessful. Yielding to the growing pressure from all sides, theOsage finally gave up their lands in Kansas, and in 1872 moved afew miles south to a new reservation in the Indian Territory. Here,in what is now Kay County, Okla., their descendants reside today.CADDOAN TRIBES IN KANSASThe problem of Caddoan peoples in Kansas is a diflEicult and com-plex one. Unlike the Kansa and Osage just considered, whose namesare usually readily recognizable from their earliest mention in thehistorical literature, the Caddoans involved appear under a puzzlingvariety of names. Moreover, it seems possible that some of the vagueand still unexplained terms used by the early French and Spanishmay also refer, at least in part, to Caddoan peoples. The liistoricdistribution of Caddoan tribes from Texas to North Dakota suggeststhe possibility that some of the northern representatives of the stock,such as the ancestral Pawnee, may have passed through what is nowKansas at some remote time. Thus, the reconstruction of their his-tory involves not only the locating and identification of their varioushistoric or documented habitats, but also the eventual isolation andrecognition through archeological means of the older sites markingtheir passage from south to north along the eastern margin of theGreat Plains, including eastern Kansas.THE PAWNEESo far as there is documentation in terms of certainly identifiablehistoric tribal groups, the Pawnee and Wichita are the only tribesof Caddoan stock whose former residence in Kansas can be actuallydemonstrated. Of these, the Pawnee were primarily a Nebraska tribeand only secondarily a Kansas tribe. Their main villages throughoutthe historic period, to judge from historical accounts, maps, andarcheological evidence, were mostly in east-central Nebraska?usuallyon the lower Loup River, on the Platte above and below its juncturewith the Loup, and less frequently on the Blue River in southernNebraska and on the Republican (Strong, 1935, pp. 15-16; Wedel,1936, 1938; Hyde, 1951, p. 130). Throughout much of the historicperiod, the hunting grounds of the Grand, Republican, and TappagePawnee lay mainly south of Platte River, including much of north-central Kansas north and west of the great bend of the Arkansas;and, according to Pike and others, they also hunted south of thatstream. Of the former Tappage Pawnee village sites said by Dunbar Whdel] an introduction TO KANSAS ARCHEOLOGY 59(1880, p. 260) to have been pointed out on the Smoky Hill in westernKansas, which may or may not be the same as the Tappage Pawneesites said by Connelley (1918, p. 442) to have been inhabited on theSmoky Hill possibly as late as 1830, no evidence has yet come to light.Actually, only two Pawnee village sites have been identified to datein the Kansas River drainage, and these are both on the Republican.One, known as the Hill site, stands on the right bank of the Republi-can in Webster County, Nebr., about 7 miles below Red Cloud; theother, designated in the archeological literature as the Kansas Monu-ment site, is some 30 miles downstream, also on the right bank, westand south of the town of Republic, Kans. The upper of these two isvery likely the site of the Republican Pawnee village visited in 1806by Pike ; and it seems very probable that the lower is likewise assign-able to that group, possibly at a slightly earlier period.The Republican Pawnee, as a separately recognized tribal or villageunit, apparently make their first appearance in recorded history in1775 when Piernas, lieutenant governor of Spanish Illinois, listed the "Republic" among the nations "with which we are accustomed to tradein the dependency of the Missouri River . . ." (Kinnaird, 1949, pt.1, p. 228). Two years later, in 1777, Cruzat (Houck, 1909, vol. 1,pp. 142-145) included them in his list of tribes receiving presents atSt. Louis. They are again mentioned in Miro's report of 1785 toRengel (Nasatir, 1952, vol. 1, p. 126), where they are termed the"Panis Republic, called Paniguacey or Eyes of the Partridge [Ojosde Perdiz]," and were said to "live on the River Cances about 130leagues from its mouth." Taken literally, this would place them onthe Smoky Hill far out in central Kansas, perhaps in Ellsworthor Russell County, away from the stream which bears their name andin a locality where no historic documents or maps indicate a Pawneevillage at this period. It seems much more likely that Miro, lacldnga firsthand acquaintance with the region he was describing, simplyfailed to distmguish between the Kansas River and its major tribu-tary, the Republican. The Kansas Monument site is some 315 milesby water above the mouth of Kansas River, and at 2.6 miles perleague this is not far from the figure Miro gives ; by land, followingthe Kansas and Republican River valleys, the distance is approxi-mately 225 miles.The Pawnee village on the Republican is also mentioned in thejournal of Trudeau in 1794 (Nasatir, 1952, vol. 1, p. 261). Here,despite an illegible word or passage, there seems to be a clear impli-cation that the Republican Pawnee were living on a branch of KansasRiver. At about the same time, CoUot (1924, vol. 1, p. 279) gives asomewhat confused description of the location of these people; buthis accompanying map has the "Republican N" (i. e., Nation) on the 60 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174left bank of a northerly tributary of the Kansas River next abovethe Blue Water (Blue River) . Later, in 1804, Lewis and Clark wrotethat the Panias Republicans were an offshoot of the Grand Pawnee,having separated from the parent band "about ten years since" andestablished themselves "on a large northwardly branch of the Kansas,to which they have given name. They afterwards subdivided, andlived in different parts of the country, on the waters of Kansas river ; but being harassed by their turbulent neighbors, the Kansas, theyrejoined the Panias proper last spring."As is evident from the foregoing review, Lewis and Clark were inerror as to the history of the Republican Pawnee?at any rate, as re-gards the date of their separation and the time of establishment oftheir village or villages on the Republican River. The RepublicanPawnee were certainly present as early as 1777 on the stream which hassince borne their name ; and since Pike in 1806, 2 years after Lewis andClark wrote, visited a Republican Pawnee town on the Republican,not all of the tribe had left the stream by 1806. Sibley ( 1927, pp. 200-203), in fact, indicates in 1811, that the Pawnee town he visited thatyear on the Loup included two bands which had resided until 2 yearspreviously on the Republican. Archeological investigations suggestthat the Hill site and the Kansas Monument site, both safely ascribedto the Pawnee, were probably not far apart in time. Accepting theHill site as the location of the Republican Pawnee in Pike's time, thatis, after ca. 1800, it seems likely the Kansas Monument site dates froma slightly earlier period. Its abandonment may well have been due toheavy pressure from the perennially hostile Kansa living on the Mis-souri or, more probably, on the Kansas somewhere near the mouth ofthe Blue. Whether the inhabitants of that village moved upriver tothe Hill site, or whether the latter represents another community thatheld out longer, is not yet clear. Final solution of this problem willhave to be made on the basis of fuller archeological evidence from bothsites, as yet not presented in full detail. Meanwhile, I think it can besafely said that the only certainly identified historic Pawnee villagesite located to date within the present limits of Kansas is the KansasMonument site in Republic County, and that this site is without doubtattributable to the Republican Pawnee of the late 18th century, thatis, after ca. 1775. If other earlier, contemporaneous, or later Pawneesites of post-Coronado times exist elsewhere in the State, they remainunrecognized, undiscovered, or unreported.THE WICHITAUndoubtedly long resident in the State, the Wichita have frequentlybeen identified by historians and ethnologists with the natives ofQuivira, who were met by Coronado in 1541 and by Oiiate in 1601. Wbdbl] an introduction TO KANSAS ARCHEOLOGY 61The location of Quivira itself has been debated for years ; but its identi-fication with the great bend region of the Arkansas River in centraland southern Kansas now seems established beyond successful contra-diction (Jones, 1929, 1937; Wedel, 1942; Bolton, 1949, pp. 282-304).This, then, places the Wichita in the heart of Kansas in the mid-16thcentury. How long they had been there when the first Spaniards ar-rived, we have as yet no good basis for judging, nor is it clear how longafter Oiiate's visit they continued their residence in the region.Identification of the natives of Quivira with the Wichita has beenbased primarily on the close similarity between the mode of life andmaterial culture reported by Coronado and Ofiate, on the one hand,and what is known of the Wichita tribes, on the other. At least twoof the accounts left by members of the Coronado expedition?that ofCoronado himself and that by Jaramillo?are those of eyewitnessesand thus of prime importance. The anonymous Relacion del Suceso,which Bolton (1949, p. 283) is inclined to attribute to Alvarado, alsogives indication of being a firsthand narrative. The lengthy accountby Castaiieda is at secondhand insofar as it relates to Quivira, sincethe writer did not accompany Coronado and his select band to theArkansas valley.According to these several accounts (Hammond and Rey, 1940),Quivira included not more than 25 towns; Jaramillo says 6 or 7,Castaneda only that the country is "well settled." Some of the townswere said to include 200 houses, clustered together. The houses wereof various sizes, but mostly round and of straw. On the outside, "they had a sort of chapel or sentry box with a doorway where theIndians were seen either seated or lying down. . . ." Other struc-tures, grass roofed but without walls and likened to pile-dwellings,suggest arbors or ramadas. Coronado commented on the stature ofthe Indians and the comeliness of the women ; and by comparing thelatter with Moorish women in appearance, he may be implying thatthey were tattooed. These natives, besides hunting buffalo, cultivatedmaize, beans, and squash ; they had no cotton or domestic fowls, noris there mention of dogs. Bread was baked under the ashes. Exceptfor a copper breastplate and "some copper jingle bells," there was nometal. Beyond Quivira, but not visited by the Spaniards, were othersettlements including, according to Jaramillo, "Arahey, with the samecustoms, settlements, and size as the former [i. e., Quivira]." TheEelacion del Suceso says the Spaniards were told of "two other largepueblos one called Taraque and the other Arae. The Taraques havestraw houses ; the Araes some of straw and the rest of hides." Arahey,or Harahey, has often been interpreted as an allusion to the Pawnee,or possibly to the Arikara (see also Lesser and Weltfish, 1932). 62 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174Onate's observations 60 years later give parallel and additional in-formation (Bolton, 1916, pp. 259-264). The "gran poblacion" hevisited in Quivira in 1601 is said to have consisted of 1,200 housesscattered on both sides of a stream flowing from the north into a largereast-flowing river. Circular in form, the houses were constructed ofpoles covered with grass. The Spaniards were impressed by theabundant crops of maize, and noted also beans and calabashes. Theywere informed that large numbers of people like these lived fartherup the river. Testimony given by members of the expedition inMexico City in 1602 (Scholes and Mera, 1940, p. 274) included theinformation that ? the huts were grouped in barrios of 30 or 40, the huts being about 30 to 40 pacesapart, and . . . the barrios were separated by two or three hundred paces.Surrounding each hut was a small cultivated plot where maize, beans, andcalabashes were raised . . .The Indians, it was further reported, wore clothing of deer andbuffalo skins, had no textiles, used metates, made pottery of brownclay and small dishes from calabashes, and apparently possessed a fewsmall dogs.All this briefly describes a way of life essentially like that followedby the various semisedentary Wichita tribes visited farther south onthe Brazos, Trinity, and Red Rivers by De Mezieres in the 18th cen-tury (Bolton, 1914, vol. 1, pp. 285-296; and vol. 2, pp. 192-204) ; andalso one still more or less characteristic in the late 19th century(Dorsey, 1904, pp. 4-6). Unfortunately, nothing resembling thename Wichita itself appears in any of the documents left by the Cor-onado and Onate expeditions; but a few of the local Indian wordstranscribed by the Spanish chroniclers seem to offer some linguisticsupport for identification of the people of Quivira with a Wichitagroup or groups (Lesser and Weltfish, 1932, esp. pp. 10-15). Thus,according to Bolton (1949, p. 293), there is eyewitness testimony "thatthe last settlement reached by Coronado was called Tabas, which is butanother spelling of Towash [cf. Tawehash], or Taovayas, the well-known name of a Wichita group at a later date." He also suggests thatthe word "z!ewc?rea," which Jaramillo uses in describing "the remotestregion of Quivira" (Hammond and Rey, 1940, p. 304), is a "mistran-scription of taovaias?that is, Taovaias or Towash." Hyde (1951,p. 19) , on the other hand, is inclined to regard teucarea as synonymouswith Touacara or Toucara, which later became Tawakoni, the name ofanother tribe closely related to the Wichita of history. He goes on toargue from this and other evidence that the Wichita proper, whom heapparently regards as originally an earth-lodge-using people, can-not be identified with the grasshouse natives of Quivira.In the present state of our knowledge, it seems to me a fruitlevSSquibble over the point whether the 16th century Quivirans can accu- Wbdel] an introduction TO KANSAS ARCHEOLOGY 63 rately be called Tawakoni, Towash, or Wichita. If there is anyvalidity in Bolton's linguistic arguments, it appears that the first twoof these terms are the only southern Caddoan tribal names that havemuch claim to identity with anything linguistic that appears in theCoronado documents. Both terms can be recognized in the lists ofWichita band names given by Mooney (Hodge, 1907-10, pt. 2, p. 947),who includes the Wichita proper as questionable, and by Lesser andWeltfish (1932, p. 10) , where nothing resembling Wichita appears. Itis possible, I suppose, that there was no Wichita band, as such, in the16th century, or that it was a relatively small and unimportant groupat the time which became prominent or dominant only in later years,or that it was synonymous with one of the two groups named above. Inany event, the Tawakoni, Tawehash and/or Wichita (see Schmitt andSchmitt, 1952, p. vii) appear to have been closely related linguisticallyand also closely associated in later historic times, whence they cameto be collectively known and designated as the Wichita or the Wichitaconfederacy. In the present discussion, particularly so far as it relatesto events prior to the mid-18th century, I have chosen to continue thisbroadly inclusive, or "loose," meaning of the term Wichita.For a century or more following Oiiate's visit to Quivira in 1601,references certainly to the Wichita are distressingly meager. So faras I have been able to determine, the name "Wichita" or anythingresembling it very closely, does not appear in the available publisheddocuments until La Harpe's mention of the Ousita in 1719 (Margry,1886, pt. 6, p. 289). There are, however, a number of other namesand appellations by which the Wichita and cognate tribes were knownat various times, and some of these offer clues to the whereabouts andmovements of these people. According to Mooney (op. cit.), the termWichita itself is of uncertain origin and etymology (cf. Lesser andWeltfish, 1932; Haas, 1942; Schmitt and Schmitt, 1952, p. vii) ; buthe observes further that the Wichita were "known to the Siouan tribesas Black Pawnee (Paniwasaba, whence 'Paniouassa,' etc.), to theearly French traders as Pani Pique^ 'Tattooed Pawnee,' to the Kiowaand Comanche by names meaning 'Tattooed Faces,' and are designatedin the sign language by a sign conveying the same meaning . . . fromtheir abundant tattooing they were designated preeminently as the 'tattooed people' in the sign language." This practice, too, stronglyimplied in Coronado's report on the natives of Quivira and explicitlymentioned in the documents relating to Oiiate's observations, resultedin application by the Spaniards of the term Jumano or Jumanes tothe Wichita. As Scholes has shown (Scholes and Mera, 1940, pp.271-276), this name was much used in early New Mexico for "widelyseparated tribes and settlements of different linguistic and cultural af-filiations"; it seems, in fact, to have been "a general term for all4,84172?59 6 64 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174indios rayados^'' that is, Indians whose body decoration included pat-tern tattooing, dyeing, or painting. In the 18th century, according toBolton (see Hodge, 1907-10, pt. 2, p. 705), the Tawehash were speci-fically referred to by the Spaniards of New Mexico as the Jumanos.Evidently, then, any interpretation or evaluation of early documen-tary references to Jumano or Jumanes in the Plains east or northeastof Santa Fe must take into consideration the strong possibility thatWichita peoples are involved.There appear to be at least two Spanish allusions to Quivira duringthe 17th century which suggest that the country so known was stillin the Kansas region or immediately to the south. According to theBenavides Memorial of 1630 (Ayer, 1916, pp 58-64), the Xumana na-tion lived more than 112 leagues east of Santa Fe, beyond the ApacheVaquero; 30 or 40 leagues farther, in the same latitude as Santa Fe,was the Kingdom of Quivira. Here, a distinction seems to have beenmade between the Xumanas and Quivira ; but if the distances and di-rections given are anywhere near correct, they suggest that the latter"kingdom" was understood to be somewhere in the Canadian or Cimar-ron valleys of northern Oklahoma. Hodge's suggestion (Hodge,1907-10, pt. 1, p. 636) that the main body of the Jumano in 1629 seems "to have resided 300 miles east of Santa Fe, probably on the Arkansas,within present Kansas," where he says they appear to have been alsoin 1632, runs afoul of the fact that the Arkansas cannot be reachedeastward from Santa Fe in less than 500 miles. His subsequent viewthat these Jumano may have been in the El Cuartelejo region of west-ern Kansas (Hodge, 1910, p. 259) has since been shown by Bolton(1911, p. 69) to be untenable, since there is good documentary evidencethat these people were then on the Nueces River in Texas.Later in the century, after 1664 but before the Pueblo Revolt in1680, several families of Taos Indians fled eastward to the buffaloplains and fortified themselves in a place thereafter termed El Cuarte-lejo. They were brought back to the Rio Grande by Archuleta(Thomas, 1935, p. 53) , who "found in the possession of these rebelliousTaos kettles and other pieces of copper and tin. Having asked themwhere they had acquired these, they answered from the Quivirapueblos, to which they had made a journey from El Cuartelejo."Moreover, by way of Quivira, "one goes to the Pawnees," with whomthe French were reported to be then trading. Whether reference hereis to the Nebraska Pawnee or to the Paniouassa farther south, I donot know; in either case, location of the Quiviran peoples in theArkansas drainage of central Kansas would fulfill the geographicrelationships implied.French maps of the latter 17th century are apparently confirmatoryof this supposition. Thus, as we have elsewhere indicated, the Mar- Wbdbl] an introduction to KANSAS ARCHEOLOGY 65quette and JoUiet maps of the lower Missouri River region, show thePaneassa west and southwest of the Osage, Missouri, and Kansa.Somewhat later, the Franquelin map of 1684 shows the Paneassa in10 villages west and south of the Missouris and Zages (Osage), somedistance north and east of the Arkansas Eiver. Unfortunately, thereappear to be no other documentary materials which would enable amore precise evaluation of these map locations.By the first quarter of the 18th century the Wichita tribes, or aconsiderable portion of them, had evidently moved southward and werecoming into the French sphere of influence. Their first recorded meet-ings with the French seem to have been in 1719. In that year La Harpe(Margry, 1886, pt. 6, p. 289) made an alliance with eight or nine bandsresiding in a string of villages on a southwesterly tributary of theArkansas, which Bolton (1914, vol. 1, map) evidently takes to be theCanadian near its confluence with the North Canadian. In a footnoteto La Hai-pe's account, Beaurain (Margry, 1886, pt. 6, p. 289) saysthat these nine villages were allied with the Paniouassa living 40leagues to the north. One is tempted to speculate that these Paniouassaof Beaurain may have been the Panioussas or Panis village visited inthe same year by Du Tisne (Margry, 1886, pt. 6. p. 314), who reachedthem overland from the Missouri Eiver at 4 days distance (La Harpe,op. cit., p. 311, says 40 leagues) southwest of the Osage villages on theupper Osage.The Beauvilliers map of 1720 (orig. in Bibl. du Dept. de la Marine,Paris. C. 4040 (7) ; photostat in Library of Congress) , which relatesdirectly to the French explorations of this period, shows the ArkansasRiver with two large forks. The "9 Nations . . ." with whom LaHarpe treated are shown on or near a short tributary of the southfork, which is designated the "Atcantca R." On the north fork,labeled "R. decouverte en 1720," and which, if not actually to be iden-tified with the Arkansas itself, could be the Neosho or Grand, are thePanis with 290 warriors. Between the forks, but northwest of thetribes just noted, are shown "Villages Ascanis et Ousita." It ismanifestly impossible to ascertain accurately, from either the narra-tives or the map, just where the various tribes were located; but itseems very probable that the nine nations met by La Harpe were, asBolton suggests, on the Canadian in east-central Oklahoma (seeWright, 1951, p. 247, for a different view). The other tribes mustthen have been to the north and northwest in the Arkansas drainage,but whether in northern Oklahoma or southern Kansas is impossibleto say. It has been customary to identify Du Tisne's Panis villageswith some locality near present Vinita, Okla. ; but there is a growingfeeling among archeologists that two early, and manifestly importantcontact village sites situated on the Arkansas just south of the Kansas 66 BUREAU OF AMERICAN ETHNOLOGY [Bltll. 174line in Kay County, Okla., may actually be the correct location(Bell and Baerreis, 1951, p. 91; see also Steen, 1953, pp. 177-178.).That Wichita tribes were still residing in the western and northernportions of the Arkansas drainage in the first half of the 18th centuryis also suggested by bits of information from the Spanish in NewMexico. In 1706, for example, Ulibarri reported that the PlainsApache northeast of Santa Fe were under attack by the Pawnee alliedwith the French, and also by a combination of Comanche and Jumano.The latter, Thomas (1940, p. 6) thinks, were the Taovayas along KedEiver, who on later occasions were known to have joined forces withthe Comanche against the Lipan. Since Ulibarri's report gives noindication as to location of these Jumano, or Wichita, if such theywere, it seems possible that they were actually located farther norththan Thomas suggests. A few years later, in 1719, Valverde reportedthe French in alliance with the Pawnee and Jumano against theApache. StUl later, but prior to 1750, a Spanish force under Busta-mente in pursuit of the Comanche, marched down the Arkansas tothe vicinity of the Jumano villages (Bolton, 1914, vol. 1, p. 48;Thomas, 1940, p. 17). Again it is not apparent just where thesewere situated. The Delisle map of 1718 has the Paniassa on the leftbank of the Arkansas, near two short tributaries entering from thenorth, and south and southwest of the Osage. The Popple map of1733, where tribal names are generously used, also locates the Paniassaon the Arkansas, with another location on the upper Red River.Excepting the Red River location, one wonders whether these repre-sentations, like that on the Beauvilliers map of 1720, are perhaps to becorrelated with the Neosho-Arkansas River locality. Archeologicalas well as further docmnentary data may throw additional light onthis interesting problem.By the latter half of the 18th century, most or all of the Wichitatribes seem to have been on the Red River and southward on the upperBrazos, Trinity, and other river valleys of north-central Texas. Herethey were visited in 1772 by De Mezieres (Bolton, 1914, vol. 1, pp. 284r-302) , whose roster of tribes met includes the Quitseys (Kechi) , Iscanis(Waco), Tuacana (Tawakoni) , Ouedsitas (Wichitas), and Taouaiazes(Tawehash). As Nortefios, or Indians of the North, the Tuacana,Quitsey, and Toauyaces among others are again mentioned by DeMezieres (Bolton, 1914, vol. 2, pp. 172-175) in 1778 as prospectiveallies against the Eastern Apache or Lipanes who, along with theOsage, were their implacable enemies. Writing from San Antonioearly in the same year, Bonilla (Bolton, 1914, vol. 2, p. 164) statedthat some of the Indians of the North "have been known in this prov-ince since the latter part of the past century, when the first entry ordiscovery was made." A promising beginning on the archeology of Wbdbl] an introduction TO KANSAS ARCHEOLOGY 67the Wichita in Texas, concerned specifically with locations identifiablewith certain of Bolton's historical determinations of the 1770's, hasalready been made (Krieger, 1946, pp. 161-164).The documented hostility of the Apache on the west and the Osageon the east toward the Wichita in the 18th century suggests an import-ant factor in the steady southward withdrawal of the latter. Withregard to the Osage, it also raises some doubts concerning anotherreference supposed to pertain to Wichita peoples north of theArkansas in the latter 18th century. In a report on the Wichita in1877, the Indian agent (Williams, 1877) stated that "The Wichitas,Wacoes, and Tawaconies are virtually one people, speaking the samelanguage, the names of Wacoes and Tawaconies being given to thedescendants of two bands of Wichitas, who about 100 years ago, leftthe main tribe on the Neosho River, in Kansas, one taking up a resi-dence on the Arkansas River, near the present town of Wichita, andthe other pushing on to Texas. . ." This may be another hint thatthe Neosho drainage was perhaps an important part of the Wichitahabitat as late as the 18th century ; but in view of the bitter warfarewaged against the Wichita all through that century by the Osage,the Neosho River in Kansas would seem to have been a most undesir-able locality for permanent Wichita towns as late as 1777.During the closing decades of the 18th century and on into the 19th,the Wichita tribes continued to reside in the upper Red and BrazosRiver valleys, far south of their old Kansas habitat. They clung tomany of the old ways, including the dual subsistence economy by whichtheir interests were divided between hunting and horticulture; andthey were on amicable terms with the Comanche. In 1805, the Wichitaproper and the Tawehash were reported to be living in two villages onRed River about 800 miles above Natchitoches. Subsequent moves tookthem to the North Fork of Red River west of the Wichita Mountainsin present Kiowa County, Okla., where they were visited in 1834 by theUnited States dragoons under Dodge and accompanied by Catlin ; tothe east end of the Wichita Mountains in the vicinity of present FortSill ; to Rush Creek in the Washita drainage, where Marcy found themin 1852 ; and ultimately, in 1859, to lands assigned them farther up theWashita in present Grady and Caddo Counties. Here they werejoined by the Tawakoni, Waco, and other Indians who had remainedon the Brazos in Texas. Their first agency was on Leeper Creek northof the Washita ; Fort Cobb was shortly established about 4 miles to thesouthwest (see Wright, 1951, pp. 258-259; Sclimitt and Schmitt, 1952,pp. vii-ix).In 1863, finding themselves at odds with their proslavery Indian andwhite neighbors, the Wichitas left their homes along the Washita andfled north to Free-State Kansas Territory. Here, according to Mead 68 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174(1904, pp. 172-175) , they "built their town of grass houses at the junc-tion of the two rivers, Big and Little Arkansas, or 'Neshutsa,' and 'Neshutsa Shinka' of the Osages, in whose territory it was located,which became known all over the plains as 'the Wichita town,' and ontheir village site has arisen the third [sic] largest city in the state,Wichita." A temporary camp was established in the summer of 1863in dense timber at the mouth of the Little Arkansas, "just across fromthe present Murdock avenue bridge." By the next summer, "grasshouses were built in groups along the Little [Arkansas] river for amile, on the east bank" ; and the women had cleared ground and plantedgardens along the stream. Here they grew corn, pumpkins, melons,and Mexican beans, in plots protected by fences made of small uprightpoles. Mead, who established a trading post among the newcomers,has left brief but interesting comments concerning their customs, in-cluding an excellent description of the lodges. Buffalo were stillplentiful to the west, and with horses given them by the Comanche,the Wichita drew on the herds as well as on their gardens. Interest-ingly enough, he observed too that "many of the older women wereartistically tattooed in pink and blue zigzag circles and lines, as wastheir ancient custom." Apparently, each of the several bands livedmore or less separately, though the context here is not clear.This final sojourn of the Wichita tribes in Kansas was destined tobe a short one. In 1867, cholera reached them by way of Americantroops (Mead, 1904, p. 176), and the Indians suffered heavily. Inaccord with orders from Washington calling for their return to theWashita, the Wichita moved southward along the Chisholm Trail.Decimated by disease, winter weather, and inadequate food supplies,they finally reached their destination in the vicinity of present Ana-darko, Okla. Here, as the "Wichita and Affiliated Tribes," they havesince resided.ATHAPASCAN AND SHOSHONEAN PEOPLES OF WESTERN KANSASWest of the regions in Kansas for which Caddoan and Siouan oc-cupations can be historically demonstrated was formerly the habitatof various wandering peoples of alien affiliations, notably the Apacheand Comanche. The former, of Athapascan linguistic affinities, wereclearly the earlier inhabitants. Their representatives were met bythe first Spanish exploring expeditions eastward from the upper EioGrande ; they were present in numbers throughout the 17th century,and evidently dominated much of the western Plains from the upperKio Grande northward at least to the upper Platte valley until wellinto the 18th century. The Shoshonean-speaking Comanche first ap-pear on the pages of history about the year 1700. Adapting them-selves readily from a mountain-based to a nomadic Plains life and Wbdul] an introduction TO KANSAS ARCHEOLOGY 69taking over the horse, they steadily expanded their range at the ex-pense of the Apache. By the middle of the 18th century, they hadswept the latter from the Plains north of the Canadian. The Co-manche likewise drifted steadily southward through the High Plains,so that by 1800 or soon after, they were sharing the upper Arkansasvalley and territory to the north with the Kiowa, Arapaho, andCheyenne, and still later with Dakota Siouan and other tribes.THE PLAINS APACHEUnder this term are included a number of peoples variously desig-nated by the 16th-, I7th-, and 18th-century Spaniards, and sometimescollectively known as the Eastern, or Prairie, Apache. Harringtonpoints out that these were the eastermnost of all the Apaches andsuggests their general designation as Lipanans in contrast to theApacheans, that is, Navaho and Apache proper (Harrington, 1940,pp. 509-512). Consistent with this view, he renames the KiowaApache the Kiowa Lipanan, possibly to be correlated with the oldPaloma Apache; and he goes on to consider briefly other early his-toric Plains Apache mider such terms as Cuartelejo Lipanan,Jicarilla Lipanan, Querecho Lipanan, Teya Lipanan, and the Lipanproper. This, in effect, substitutes the term Lipanan where Hodge(1907-10, pt. 1, p. 631) used the word Jicarilla in his earlier discussionof the Plains Apache. Wright (1951, p. 180) regards the Lipanas originally an offshoot of the Jicarilla and says they have beenidentified in reservation days with the Mescalero Apache in NewMexico.The first allusions to the Plains Apache are found in the narrativesof the Coronado expedition (Hammond and Key, 1940). Fourteendays east of Pecos, near the Canadian River and the present Texas-New Mexico line, the Spaniards found encampments of roamingpeople they called Querechos. The several eyewitness accounts con-cerning these natives furnish a concise but clear picture of a pre-horse Indian economy in the southern plains. The Querechos, likethe Teyas met soon after, lived "like the Arabs," says Castaneda;according to the Relacion Postrera, the "maintenance or sustenanceof these Indians comes entirely from the cows, because they neithersow nor reap corn." They followed the herds and in winter carriedtheir stores of dressed hides to the settled Indian towns and barteredthem for maize and blankets, "each company going to those whichare nearest, some to the settlement at Cicuye [Pecos], others towardQuivira, and others to the settlements which are situated in the di-rection of Florida." Clearly indicated in the accounts is the restlesslife of these hunter folk, and there is mention of dog traction and thetravois, the conical skin tipi, the breechcloth and skin clothing, the 70 BUREAU OF AMERICAN ETHNOLOGY [Bdll,174bow and arrow, pemmican, body painting or tattooing, sign language,and the use of articles and implements fashioned from bone, stone,buffalo hair, and sinew. The people were friendly and unafraid.Despite their trade contacts with the Pueblos, the Teyas were saidby the Spaniards to have been responsible for the recent destructionof several pueblo towns between Tiguex and Cicuye, and even to havebesieged the latter town.Later Spanish explorations eastward from the Rio Grande addedsteadily to knowledge of the distribution and habits of these earlyplainsmen. One Jusephe, a deserter from the ill-fated expedition ofBonilla and Humaiia in ca. 1593, was a prisoner among the ApacheVaquero for a year before finally making his escape to Picuris, whereOiiate found him. In 1598, Onate sent Zaldivar toward the eastwhere he met numbers of Vaqueros, some of whom were returningfrom Taos and Picuris where they had traded meat, hides, and otheritems for cotton blankets, maize, pottery, and "some small greenstones" (turquoises?). He described their tents as colored brightred and white, noted their heavy dependence on the buffalo, and wastold that the Jumano were their enemies. In 1599, in a letter to theviceroy, Onate mentioned the "Querechos, o baqueros," who livednear the buffalo herds. Two years later, on his march to Quivira, hemet "Apachi" near the Canadian River; farther on, he met still more "people of the Apache nation, who are the ones who possess theseplains," and spoke of their moving about with the herds. Approach-ing the Arkansas River in south-central Kansas, he presently en-countered another populous tribe, who are elsewhere named by mem-bers of his expedition the Escansaques. They were enemies of theQuivirans living not far to the east, and subsequently turned againstthe Spaniards. Formerly regarded as the Kansa, there can be littledoubt that these people, as described in the Oiiate records, wereactually another nomadic Apache group, very likely encamped inpresent southern Kansas.The Benavides Memorial of 1630 (Ayer, 1916, pp. 39-57) suggeststhat by this time, nearly a century after their first meeting, the Span-ish were deeply impressed by the presumed numbers, distribution, andvalor of the Apache. Four principal groups are recognized anddescribed in the Memorial?the Apache de Navaho, Apache de Xila,Apache de Perillo, and Apache Vaquero (see also the maps of Minet1685, Coronelli 1688, and Delisle 1703, in Tucker, 1942, pis. 7, 9, and13). Of these only the last-named concerns us here. The ApacheVaquero were said to inhabit the territory east of the Rio Grande,ranging "more than 150 leagues along the boundary of New Mexicoand extending more than 100 leagues eastward." Again their heavyreliance on the bison is stressed. Planting is mentioned under the Wbdbl] an introduction to KANSAS ARCHEOLOGY 71description of the Apache generally, and is specified for the Apachede Xila and Apache de Navaho ; but it is not mentioned in the longdiscourse on the Apache Vaquero. In a footnote, Hodge (Ayer, 1916,p. 263) identifies the Apache Vaquero with tribes later known as theJicarilla, Faraone, Lipan, Llanero, and Mescalero.For the next 50 or 60 years, there appears to be little informationconcerning the Apache in the distant plains east and northeast ofSanta Fe. To the diiSculties facing the Spaniards in maintainingtheir long communication lines to Mexico and in holding the restivePueblos under control was added the thorny problem of the nearerApache, who were increasingly raiding the Spanish and pueblo settle-ments alike. Conditions worsened rapidly during the two decadespreceding the Pueblo Eevolt of 1680, partly because of severedroughts and food shortages. Some time during this period, a groupof dissatisfied Taos fled to the buffalo plains of eastern Coloradoor western Kansas from which they were brought back to the RioGrande by Archuleta ; the spot in which they had settled briefly andbuilt houses thereafter was called El Cuartelejo, and this term wasalso applied to the nearby Apache whose slaves the Taos had become(Thomas, 1935, pp. 11, 264, n. 23). Unfortunately, there are no de-tails regarding these Apache, thought by Thomas to have been livingin the Arkansas drainage of eastern Colorado ; and the record is silentas to how closely their manner of living resembled that of the ApacheVaquero of earlier days.Following the reconquest of the pueblo area after the Revolt, theSpanish again came into intimate contact with the Plains Apache;and to their observations from this time on can be added those of theFrench, pressing toward New Mexico from the Mississippi Valley.Before 1700, news of the French in the lower Mississippi Valley wasreaching the Spaniards by way of the Apache; and there are re-ports of Navaho journeys to Quivira and raids by them againstthe Pawnees and French (Thomas, 1935, p. 13). In 1706 Ulibarrimarched northward from Santa Fe via Taos to the Cuartelejo Apachecountry beyond the Arkansas, where he rounded up 62 Picuris In-dians who had been living there since 1696 and returned with themto the Rio Grande. Among the Xicarilla, Penxaye, and other namedApache in northeastern New Mexico and southeastern Colorado,Ulibarri found maize-bean-squash agriculture, as well as unrestover the threat of Comanche raids. In the Cuartelejo country, hevisited or reported four or five rancherias, giving their names, andnoting that they were widely scattered over some 40 leagues of land.More important, he noted also that "at the end of July they hadgathered crops of Indian com, watermelons, pumpkins, and kidneybeans." The natives, though still unconverted to Christianity, wore 72 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174 crosses, medals, and rosaries. The Spaniards saw guns, metal kettles,and other items which had been taken from the French and Pawnee,and observed that from white men to their east, the Apaches "buymany iron things such as hatchets, swordblades, arquebuses, copperthings." There appears to be no mention of horses among theCuartelejo at this time.Thirteen years later, Yalverde in pursuit of Ute and Comanchehostiles, traveled northward and then east over substantially thesame route taken by Ulibarri. In the valleys north and east of Taosand south of the Arkansas he found the Apache tribes growing siz-able crops of maize, beans, and pumpkins; some of them practicedirrigation and lived in stone houses. He also heard of rancheriaslaid waste by the Comanche and Ute, whose war trails his expeditioncrossed. On the banks of the Rio Napestle (Arkansas) , he was metby a host of Cuartelejo Apaches?200 tents, and many dogs who drewtheir baggage. There is no clear indication of horses among theCuartelejo. Valverde also interviewed an Indian with a gunshotwound, received in "the most remote borderlands of the Apaches,"presumably to the north or northeast, when the "French, united withthe Pawnee and Jumano attacked them from ambush while theywere planting com." Under pressure from the French and Pawnee,these northern Apache, whom Thomas (1935, p. 31) identifies asthe Paloma Apache, had given up their old lands and were with-drawing to the Arkansas. Later testimony given by participants inone or both of these Spanish expeditions, when the suitability ofEl Cuartelejo as a frontier outpost was being evaluated, developedthe fact that in that locality, "small returns can be expected, since theIndians, who ranch there seasonally, gather the scanty harvestswhich the place provides, and retire to other spots to pass the win-ters, which are rigorous, and where there is no supply of wood formany leagues . . ." (Thomas, 1935, pp. 157, 161, 173).The Cuartelejo were visited again in the following year, 1720, whenValverde's lieutenant, Villazur, was dispatched northward from SantaFe at the head of a force of about 100 soldiers, settlers, and Indianallies to ascertain the exact location of the French. Proceeding viaEl Cuartelejo, Villazur reached the Platte in present Nebraska aftera march of some 60 days, or approximately 300 leagues, where he anda large part of his command perished at the hands of the Pawneeand their allies, presumably including some Frenchmen. The sur-vivors were rested and fed at El Cuartelejo on their flight to Santa Fe.As elsewhere indicated, the heart of the Cuartelejo and PalomaApache territory in the first quarter of the 18th century, as the con-temporary Spanish knew it, is identified by Thomas with easternColorado, extending northward probably as far as the South Platte WKDEL] AN INTRODUCTION TO KANSAS ARCHEOLOGY 73River. Inferentially, however, it must have included a considerabletract to the east, well into Kansas and southwestern Nebraska, wherethese or related peoples had their habitat beyond the direct reachof Spanish expeditions. Other Apache groups doubtless lived farthersouth, on the westerly tributaries of the lower Arkansas River. Itwas these easternmost Apache with whom the French first came intocontact during their attempts to reach the Spanish territory overlandfrom the lower Mississippi-Missouri area early in the century. Theirwestward passage across the plains was blocked by warlike peopleswhom they called the Padouca. These are shown on numerous Frenchmaps, usually in essentially the same region as is ascribed by the Span-ish to the Plains Apache, and apparently beginning as early as theDelisle 1718 map of Louisiana. A possible variant of the term appearseven earlier on the Franquelin map of 1684, in which the "Riviere desParouke" is shown.^?Not much information is available from the French regarding thePadouca, at any rate as compared with what the Spanish tell us aboutthe Apache; but they are usually described as the implacable enemiesof the semisedentary tribes living along the lower Missouri, the Ar-kansas, and their western tributaries. Bourgmond in 1717 spoke ofthe "Padauccas" as the allies of the Pawnee and Panimaha on thePlatte. At about the same time, Le Maire wrote that "The Spanishwho go toward the Missouri fear to be robbed by the Apaches orPadoucas on horseback, who are the Arabs of this area." Du Tisnevisited the grand village of the Padouca in 1719, noting their war-like nature, use of leather armor for their horses, and of the bow andarrow as weapons. In the following year, Boisbriant wrote of thedilemma in which the French found themselves with respect to thePadouca, stating that if peace were not made with them, the Frenchwould have to give up all hope of "reaching New Mexico by crossingthe vast country roamed over by the Comanches [sic], a name underwhich the Padoukas are known." In 1724 Bourgmond finally madepeace between a Padouca band and the Kansa in east-central Kansas ; his brief description of their mode of life, with limited maize agri-culture and some pottery, is reminiscent of the Spanish accounts con-cerning the Apache of the same period. Except in his mention ofhorses, there is no marked disagreement between this description byBourgmond and the fuller accounts by Valverde and Ulibarri con-cerning the Cuartelejo and their neighbors farther to the west. (SeeChampe, 1949 a, p. 291, for a different view.) "The Homann map cited by Secoy (1951, p. 525) as "one of the earliest definite exam-ples" of the use of the word Padouca is much later than the date of 1687 he assigns it(Lowery, 1912, p. 333) ; and for the Plains area it obviously draws on the Delisle map of1718, which utilizes geographic and ethnic data gathered by Bourgmond between 1714and 1717. 74 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174The Cuartelejo and neighboring Apache peoples northeast of SantaFe figured more or less in Spanish frontier affairs for another decadeor two after the time of Valverde and Bourgmond; but the recordsshow that, under growing pressure from the Comanche and Ute on thewest, theirs was an increasingly mihappy existence here in the upperArkansas and nearby drainages. In 1726 a group of Escalchufinesand Palomas fleeing from the Comanche reached New Mexico and in-formed the Spaniards that among their pursuers were some French.Thomas (1940, p. 14) suggests that these may have been from Bourg-mond's expedition of 1724 or from a later party of Frenchmen who 2or 3 years after were reported to have ascended the Arkansas in searchof the Padouca (Bolton, 1914, vol. 1, p. 58). In 1728 Governor Bus-tamente passed on to the viceroy another Apache report that severalFrenclmaen had established themselves at El Cuartelejo; and, further,that some of these 'Svent with a great force of Apaches of the nationsPalomas, Cuartelejos, and Sierra Blancas to look for the Comanches(a people widely scattered because of the numerousness of their nation)to see if they could force them to leave these regions . . ." (Thomas,1935, pp. 256-258). The Comanche, he noted, were then "in ElAlmagre or a little farther away," that is, somewhere m the area of theFront Range of Colorado (Thomas, 1935, p. 283; see also 1932, mapopp.p.l).In the next 10 years, the expulsion of the Plains Apache from theterritory northeast of Santa Fe by the Comanche appears to have beencompleted, for when the incompletely documented Mallet brothers'expedition in 1739 passed through the Platte and upper Arkansascountry, they encountered only the Laytanes, or Comanche, in theregion (Margry, 1886, pt. 6, p. 455). According to Bolton (1914,vol. 1, pp. 24-25) : Till after the opening of the 18th century the Apache tribes, especially theLipan, regarded as their own the territory from the upper Nueces and MedinaRivers to the upper Red and Colorado, while their range between summer andwinter might cover many hundred miles. . . . About 1700 [the Comanche]reached New Mexico and the Panhandle country. Next they attacked the Apacheand crowded them southward, destroying the extensive Apache settlements ofsouthwestern Kansas, and occupying the northern Apache lands themselves. ... By the middle of the [18th] century the more usual haunts of the Lipanwere the districts about the San Saba River, in west-central Texas, while theupper Colorado, Brazos, and Red Rivers were in the hands of the Comanche.At this time the Carlanes, who early in the century had lived in southwesternKansas, the Chilpaines, Palomas, Pelones, Faraones, and Natages, were all livingsoutheastward from Santa Fe in what are now eastern New Mexico and westernTexas.That part of the old Apacheria with which we are here primarily con-cerned, including western Kansas and adjacent regions, had in otherwords become part of the new Comancheria. Wedbl] an introduction TO KANSAS ARCHEOLOGY 75With respect to Kansas, then, Plains Apache peoples appear to haveheld the western third or more from perhaps the earliest white contacttimes until the first quarter of the 18th century. As Kroeber (1939,p. 37) notes, "they were thus part of the tribes within the old, pre-horse, Plains culture." It is at present only an inference, but perhapsa not unreasonable one, that peoples following the Querecho-Teya-Vaquero type of subsistence economy were present in the Arkansasdrainage and northward in the 16th century and probably earlier, sincethere are no eyewitness accounts for the plains north of the CanadianRiver during that time. When details concerning the northerly re-gions become available, principally after the Pueblo Revolt in 1680,it is a partly hunting, partly horticultural subsistence economy that isrevealed. The contemporary accounts, moreover, suggest strongpuebloan influences on the Plains Apache who were closest to thepueblo towns; farther away, irrigation and stone houses among theApache are not mentioned, and cultivation seems to have been progres-sively less intensive. To what extent puebloan fugitives from the RioGrande, before and after the Revolt of 1680, may have been instru-mental in introduction of maize agriculture to the Plains Apache I amunable to say. What does seem clear is that this partly horticulturalbasis of life, but with cultivation less strongly emphasized than amongthe village Indians to the east or the Pueblos to the southwest, lingeredon until the southward dispersion of the Apache following the timeof Valverde and Bourgmond. Thus, among the archeological mate-rials of the late I7th to early 18th centuries from western Kansas, it isreasonable to expect some evidence of the Plains Apache. Such evi-dence has in fact already been found and reported, in Kansas andNebraska, as the Dismal River culture (Strong, 1935, pp. 212-217;Wedel, 1940, p. 323; Hill and Metcalf, 1942; Champe, 1949 a). Weshall return in another place to a discussion of this problem (p. 590) . THE COMANCHEThe Shoshonean-speaking Comanche, in alliance with the Utesagainst the Spaniards and Pueblos, make their first certain appearancein documentary history shortly after 1700 (Thomas, 1935, pp. 61, 262,n. 8), when Ulibarri reported that the two tribes were about to attackTaos. Their home then appears to have been in the mountain valleysto the northwest, presumably around the headwaters of the ArkansasRiver. Earlier the Comanche must have resided with the closely re-lated Shoshone yet farther to the northwest?perhaps, as Shimkin(1940, p. 20) suggests, in the Snake River drainage of southern Idaho,northern Utah, and western Wyoming. From this ancestral home-land, the Shoshone supposedly spread north and northeast, the Co-manche southeast?first, as a mountain-based hunting and gathering 76 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174people, later as full-fledged wide ranging Plains bison hunters. Un-fortunately, there are no contemporary documents for the Wyoming-Nebraska region from which a possible Comanche spread into thePlains there can be considered.Acquiring horses early in the 18th century, the Comanche promptlyembarked on a career of alternate raiding and trading vis-a-vis theSpaniards and Pueblos and of almost constant warfare against thePlains Apache northeast of Santa Fe. The story of their expulsionof the latter from the plains of the upper Arkansas is vividly sketchedin the Spanish documents presented by Thomas (1935, 1940) . During the second quarter of the century, their destruction of thePlains Apache settlements in eastern Colorado and western Kansaswas completed. By midcentury, the Apache from this region had beenpushed far to the south?the Lipan, according to Bolton (1914, vol. 1,pp. 24-25), into west-central Texas, the Carlana, Paloma, and othersinto west Texas and eastern New Mexico southeastward from SantaFe. The High Plains from the upper Colorado River of Texas north-ward to the Platte in Nebraska, and the Colorado piedmont frontingthe Rockies, were firmly in the hands of the Comanche. It was pre-sumably the Comanche, under the name of Laitanes, whom the Malletbrothers met, probably somewhere on or near the upper Arkansas, in1739. A few years later, the French engineered a treaty between theComanche and Jumano (Wichita) along the Arkansas, both tribesbitter enemies of the Apache ; and by 1748 it was reported that therewere 33 Frenchmen in La Jicarilla, selling guns to the Comanche inthe former Apache stronghold. In addition to dispossessing theApache from their old lands, the Comanche also acquired the namePadouca by wliich the French had known the Apache. They were soknown to the Pawnee, according to Pike (1810, app. pt. 2, p. 17), aswell as to various Siouan and other tribes.The heart of the Comancheria of the latter 18th century and laterevidently lay south of the Cimarron (Wallace and Hoebel, 1952, p.8) , but the Comanche also ranged well to the north of that stream untilthe 19th century. In 1787 the Spaniards in New Mexico almost per-suaded a large Comanche band to settle permanently on the upperArkansas near present Pueblo, Colo., even going so far as to send car-penters, erect houses, and furnish livestock and seed for farming ; butbefore the following spring the attempt ended in failure (Thomas,1929). In 1806, Pike saw unidentified Indian traces which he sus-pected were Comanche on the upper Arkansas ; and he recommendedestablishment of one of two trading posts for them "near the moun-tains on the Arkansas" (Pike, 1810, app. pt. 2, p. 53) . Comanche warand trading parties evidently traveled far to the north even after thattime, where they came into contact with such northern tribes as the Wedbl] an introduction to KANSAS ARCHEOLOGY 77Pawnee, Arikara, and others ; but they no longer held undisputed con-trol of the region north of the Arkansas. The accounts of Long's ex-pedition of 1820 indicate that the Arapaho and Cheyenne had by thattime moved into the region of western Kansas and eastern Colorado,and were occasionally ranging even farther south. From tliis timeon, the Arkansas seems to have been the approximate northern limitof Comanche territory and much of this range north of the upper RedEiver was shared with the Kiowa and Kiowa-Apache. It was thesethree tribes particularly who, in later days, harassed the wagon trainson the Santa Fe trail south of the Arkansas River crossing.That the warlike and aggressive Comanche were influential factorsin the early post-White history of the western Plains is, of course, ob-vious ; and their movements unquestionably are, and will be, stronglyreflected in the later archeological developments in the western Kansasregion. It seems improbable, however, that their stormy passagethrough this region has left remains of their own which will figuresignificantly in the archeological picture. All extant accounts stresstheir highly mobile and nomadic way of life, based almost exclusivelyupon hunting and gathering, with portable habitations of skin andpoles. So far as I am aware, there is no authentic report of maizegrowing, pottery making, or semipermanent architecture among theComanche, or of prolonged residence in any given locality by thesepeople. Their campsites, thus, would be extremely superficial, exceptperhaps in the case of unusually favorable spots frequently used overa period of years. Absence of the trait complexes and practices justenumerated leaves little or nothing on which the archeologist can hopeto lay his spade. The chances seem remote, therefore, that archeologyin Kansas will be able to define a material culture complex that can bedefinitely identified as a Comanche manifestation.THE PADOUCAThe name Padouca figures prominently in the 18th and early 19thcentury literature on the western Plains ; and the identity of the tribe,or tribes, to whom it was applied at various times and by differentwriters has been discussed so often that extended comment here is un-necessary. As Grinnell and others have indicated (Grinnell, 1920;Hyde, 1934, pt. 2, pp. 18-25, and 1951, pp. 278-281 ; Strong, 1935, pp.25-26), there is some confusion in the way the term was used by con-temporary observers and a good deal more in its interpretation by laterstudents. From this confusion, however, two points seem to emerge.For one thing, it is fairly clear that the term Padouca was applied inthe 19th century by Siouan (Mooney, 1896, p. 1043 ; Fletcher and LaFlesche, 1911, p. 101) and other (Michelson, 1921) tribes to the Coman-che. Equally clear, I think, is the fact that the same, or a similar, term 78 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174was generally used by the 18th-century French, on maps and in otherdocuments (see, e. g., Margry, 1886, passim, and Villiers, 1925), inreference to the tribes whom the Spaniards in New Mexico knew as thePlains Apache. Tliat these were all actually Apache-speaking peoplescannot now be proved, of course ; the term may have been used some-times to include groups following the Apache manner of life buthaving other linguistic affiliations. As the evidence long ago assem-bled by Grinnell should have made clear to all of us, however, the mostreasonable interpretation at the moment?and the one I have acceptedin the foregoing pages?seems to be that the name under which thePlains Apache were known to the French and their Indian allies beforecirca 1750 was transferred after that time to the people who displacedthe Apache, namely, the Comanche. In the most recent review of thePadouca problem, Secoy (1951, pp. 525-542) comes to this conclusion,which I think Grinnell almost reached; but Secoy's presentation ismarred by a regrettable misuse of cartographic evidence, by a series ofundocumented assertions regarding sources of geographic and ethno-historic information, and by specious arguments from insufficient,doubtful, or irrelevant premises.OTHER TRIBES IN WESTERN KANSASThere remain for brief consideration here a few other tribes thatranged in historic times through parts of western Kansas. These in-clude especially the Kiowa, Kiowa Apache, Arapaho, and Cheyenne.During their documented sojourn in this region, all were evidentlynonagricultural bison-hunting peoples. With the possible exceptionof the first two, they were generally late arrivals in the area andpresumably left little or no imprint on its archeology.THE KIOWA AND THE KIOWA APACHEThe Kiowa, once classed as an independent linguistic stock, are nowregarded as related to the Tanoan-speaking peoples of the South-west. According to long-held views first propounded by Mooney(1898, pp. 152-168), they migrated in ancient times from the upperMissouri-Yellowstone country of southwestern Montana eastward tothe Black Hills, whence the Dakota claimed to have driven themsouth to the Platte. Under pressure from the Arapaho and Chey-enne, they continued to drift southward across the Arkansas, andfinally established themselves in the region between the upper Redand upper Arkansas Rivers. Here they came into close associationwith the Comanche, and were on friendly terms also with the neigh-boring Wichita tribes. Mooney discusses at some length their rela-tionships with various northern tribes, including the Crow, Hidatsa,Mandan, Arikara, and others, and it is evident that they were wellknown to the Northern Plains peoples. Tabeau, about 1803, men- Wbdkl] an introduction TO KANSAS ARCHEOLOGY 79tions them at the Indian gathering place in the Black Hills (Abel,1939, p. 154) . Their claim to a northern origin has been challenged,however. On linguistic grounds, Kroeber (1939, p. 48) suggests thatthe Kiowa are apparently "a group that anciently broke away fromthe Tanoans of the Rio Grande?somewhat like the Comanche fromthe Shoshone much later on." More recently, Lowie (1953, pp. 1^)has pointed out that there is no valid proof of their prolonged north-ern residence or of close association with the Crow ; traditions of sucha northern stay he is inclined to regard, as Kroeber (1939, p. 80) ap-parently does, as reference to a "temporary northward migration,preceded by southern residence and followed by a return to thesouth." Further, he characterizes them as a distinctly southernPlains tribe, linguistically and by known residence. That they wereperhaps somewhere in the Southern Plains during the period whenMooney would have them moving southward from the upper Mis-souri country would seem to be indicated if the Manrhoat of La Sallein 1682 were actually the Kiowa, as some apparently feel. In hissummary of tribal history in the plains-prairies, Kroeber (1939, p. 86)suggests that three to five centuries ago the Kiowa may have beenamong the southern Athapascans, i. e., Plains Apache, or northward,which seems reasonable. Their early history, however, like that ofsome other wandering Plains tribes, is quite impossible to trace indetail.Closely associated with the Kiowa were the Kiowa Apache(Mooney, 1898, pp. 245-253), who were supposedly with the Kiowabefore they left their alleged early habitat in or near the northernRockies. Their nearest linguistic relatives are the Jicarilla andLipan (Swanton, 1952, p. 296) . Kroeber (1939, pp. 37, 79) sees themas a band of eastern Apache?"mountain tribes, marginally South-western, fronting on the plains and hunting bison"?who, after theygot the horse, "committed themselves definitely to the bison and theplains, and on account of numerical weakness joined themselves tothe Kiowa." The first historical mention of them is thought to be asthe Gattacka of La Salle in 1682. This term, similar to names bywhich the Kiowa Apache were known to the Pawnee, Omaha, andPonca (Swanton, 1952, p. 296), appears on the Franquelin map of1688 (Tucker, 1942, pi. IIA) just west of the Manruth (Kiowa?) inwhat is certainly the Southern Plains. Their earlier background andsubsequent history in the 18th century, like that of the Kiowa, remainobscure. They certainly ranged widely, however, being reported byTabeau at the Black Hills gathering about 1803. Harrington's sug-gestion (1940, p. 510) that the term Paloma Apache was perhaps theSpanish name for the Kiowa Apache is an intriguing one, and is, ofcourse, in line with the previously cited view of Kroeber. It seems484172?59^ 7 80 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174quite possible that both the Kiowa and the Kiowa Apache wererooted, at least in part, in the Padouca or Plains Apache who receivedso much attention from the Spanish in New Mexico during the l7thand 18th centuries. THE ARAPAHO AND CHEYENNEThe Algonquian-speaking Arapaho and Cheyenne were certainlylate in the western Kansas region. The early history of both is ob-scure, that of the Arapaho especially so. Originally horticulturalvillage peoples, both tribes entered the Plains from the westernMinnesota region?first the Arapaho from the Eed River valley, laterthe Cheyenne from somewhat farther south. The Arapaho driftedtoward the southwest and eventually crossed the Missouri, after whichtime they are not known to have practiced corn-growing. TheCheyenne (Grinnell, 1923) dwelt for a time in southeastern NorthDakota, where their 18th century archeological remains on theSheyenne River have been identified (Strong, 1940, pp. 370-376).Continuing toward the west and southwest, they apparently retainedagriculture for a while after crossing the Missouri. In Lewis andClark's time, both tribes were located well to the west of the MissouriRiver?the Arapaho or "Ka-ne-na-vish" on the upper Platte in west-ern Nebraska, the "Chayenne or Sharha" in the Black Hills district atthe headwaters of the Cheyenne River (see Arrowsmith map of 1814,in PauUin, 1932, p. 29). Both are mentioned by Tabeau in thisgeneral region (Abel, 1939, pp. 154-155). During the 1830's andlater came the gradual separation of both tribes into what came to beknown as the northern and southern groups, the latter continuing tomove south through the western Plains. Neither Arapaho norCheyenne were mentioned by Pike on his journey up the Arkansas in1806, and his passing remarks suggest that this was still Comanche,rather than Arapaho and Cheyenne, range at that time. In 1820,however, when members of the Long expedition under Bell descendedthe Arkansas, they met a large band of Arapaho and Cheyenne,along with Kiowa and Kiowa Apache, near the point where Bent'sFort was erected in 1832 ; more of the same Indians were encounteredas the detail marched downstream (James, 1823, vol. 2, pp. 174-198).In 1842 and again in 1843, Fremont reported Arapaho and Cheyenneon the South Platte near the mountains (Fremont, 1845, pp. 18, 23,28-30, 111) ; and in 1844, some 20 miles below Bent's Fort on theArkansas, he met "a very large village of Sioux and Cheyenne In-dians, who, with the Arapahoes, were returning from the crossing ofthe Arkansas, where they had been to meet the Kioways and ComancheIndians. A few days previous they had massacred a party of 15Delawares, whom they had discovered in a fort on the Smoky HiU Wbdbl] an introduction TO KANSAS ARCHEOLOGY 81 river" (Fremont, 1845, p. 288). On Bvt. Maj. Gen. G. K. Warren's1867 map of Nebraska, Dakota, and adjoining territory, a broad arearmming from the Colorado Rockies into the Plains between theSmoky Hill and Republican Rivers as far east as the lower Solomonin eastern Kansas is designated "Shyenne Indians"; to the southis indicated the Arapaho range. Both tribes, but especially theCheyenne, figured in the Indian wars of this section prior to theirplacement on a reservation in the Indian Territory following theMedicine Lodge treaty of 1867. So far as I am aware, no sites at-tributable to the Southern Arapaho and Southern Cheyenne havebeen reported in western Kansas.It should not be necessary to point out that the extensive territoryassigned by Warren to the Cheyenne and Arapaho was by no meansexclusively theirs. Other tribes, such as the Comanche, Kiowa, andApache, also pursued the diminishing bison herds here (Beckwith,1855) ; and as game elsewhere became scarcer and scarcer, still othertribes sent hunting parties into western Kansas. Among these werethe reservation tribes from the eastern part of the State and perhapsoccasional Oglala Sioux from the upper Platte valley in Nebraska.All fought bitterly and unceasingly against the Pawnee on theirseasonal hunting excursions into the Republican-Smoky Hill-Arkansas plains. This, however, belongs to history, rather than toarcheology, and is beyond the scope of the present paper.SUMMARYThe ethnohistorical materials just reviewed suggest that the twotribes with strongest claim to early and prolonged residence in whatis now Kansas are the Wichita and the Kansa. The Wichita, orclosely related peoples who may fairly be included under this name,evidently dominated the Arkansas River valley of central Kansasat the opening of the white contact period in the mid-16th century;and, although moving in general southward, they were probablypresent until some time in the early 18th century. The Kansa, firstrecorded in what appears to be the general Kansas region in 1673,were certainly residing in the State from 1723 until their removal tothe Indian Territory in 1873. How much earlier these two tribeswere in their early historic habitat is a problem for the archeologistrather than the historian. The Plains Apache, either as out-and-outbison hunters (Apache Vaquero) or as half-hearted horticulturists(Paloma, Cuartelejo, and/or related groups) held the western Plains,certainly from the opening of the l7th century?perhaps for sometime before?until into the second quarter of the 18th century. It ispossible, I suppose, that their residence in the region began about asearly as that of the neighboring Wichita and Kansa. Of the Pawnee, 82 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174only the Kepublican band can be documented as historic residents ofa portion of Kansas ; but they were late and only briefly here. If theother bands of this powerful and important Nebraska tribe resided inKansas, there is no historic proof of that fact?or, as seems muchmore probable, their residence must be ascribed to an earlier, probablypre-Coronado, period, in which case they become an archeologicalproblem. As to the other tribes whose names may be linked with theState as former residents, the Comanche, Arapaho, and Cheyennewere demonstrably late or transitory, or both. Essentially, the sameholds for the Osage, for whom residence in the State can bedocumented only after circa 1815.The historic distribution of these tribes, with their varying sub-sistence economies, conforms rather nicely to the several natural re-gions elsewhere outlined for the State. Absolute concordance neednot be expected. Thus, the eastern portion, where normal precipi-tation averages around 32 inches or more annually and valley bottomhorticulture is a reliable food source, was occupied or controlled bysemihorticultural Siouan-speaking tribes?the Kansa in the glaciatednortheastern corner, with the rolling unglaciated Osage Plains to thesouth as back coimtry of the Osage whose villages stood mainly inextreme western Missouri. In the drier central third?the PlainsBorder of Fenneman?with annual rainfall averaging 25 to 30 inches,were semihorticultural Caddoan-speaking peoples?the "Wichita southof the Smoky Hill drainage, the Pawnee in late villages on the Repub-lican near the Kansas-Nebraska State line. Much of the north-central region was, of course, included in the regular seasonal huntingrange of the Nebraska Pawnee. The western third, short-grass coun-try too dry for dependable corn-growing, was Apache, then Co-manche, finally Arapaho and Cheyenne?hunters all, though withsome sporadic creek-valley corn-growing around the beginning of the18th century among the Apache. This distribution of semihorticul-tural Siouans in the east, semihorticultural Caddoans in the middle,and basically nonagricultural Athapascan, Shoshonean, and otherpeoples in the west parallels the historic pattern of tribal distributionsin Nebraska as outlined by Strong (1935, p. 40). When our archeo-logical data have been presented, we shall have occasion to compareagain these two areas with respect to the antecedents of the historicsubsistence economies and the tribal or cultural entities involved.PREVIOUS ARCHEOLOGICAL WORK IN KANSASObservations and passing comments on archeological remains inwhat is now Kansas apparently began with the first American ex-ploring expeditions through the region. As we have already notedelsewhere, Lewis and Clark, Long, and others on their way up the Wbdbl] an introduction TO ICANSAS ARCHEOLOGY 83Missouri regularly reported on two ancient villages of the Kansa ; andClark observed further that "this Town appears to have covd a largeSpace, the Nation must have been numerous at the time they livedhere" (Thwaites, 1904-5, vol. 1, p. 67) . None of these early observ-ers, so far as I can determine, did any digging or collecting withinthe area of our interest.Later, as the American frontier pushed westward and settlementof the trans-Missouri lands began, increasing numbers of personssaw and commented on the ancient remains along the stream valleysand adjacent bluffs of eastern Kansas. Since these antiquities weregenerally insignificant and unspectacular by comparison with theearthworks and other remains farther east, they attracted little noticeand less discussion from either laymen or scholars. Thomas, forexample, in his catalog of prehistoric works east of the Eocky Moun-tains, lists mounds reported in the literature from Coffey, Cowley,Leavenworth, Marion, Riley, Shawnee, and Wyandotte Counties(Thomas, 1891, pp. 88-89) ; but the State is not even mentioned in hislater comprehensive report on mound explorations of the Bureau ofEthnology. The accompanying distribution map, however, indicatesmounds on the lower Kansas, Republican, and Walnut Rivers and onthe upper Cottonwood (Thomas, 1894, pi. 20). Similarly, GarrickMallery's paper on picture writing devotes less than half a page toKansas pictographs (Mallery, 1893, p. 80) ; and Holmes, a decadelater, in his great study of pottery of the Eastern United States, makesno reference whatever to Kansas materials. As late as 1941, Shetrone(1941, p. 340) dismissed the mounds of Kansas in the following words : "A few scattering mounds have been noted in eastern Kansas, mainlyalong the Kansas River." Actually, as will become clear in the fol-lowing pages, our long-standing ignorance of Kansas prehistory isdue not to absence of aboriginal remains, but to the dearth of reli-able information based on systematic research and excavation.The earliest archeological excavations of which I find record per-tain to mound investigations by the Reverend Isaac McCoy in thevicinity of Fort Leavenworth. McCoy was in charge of the boundarysurvey for the Delaware Indian reservation in 1830, in course ofwhich he described Waconda Spring and various other features ofinterest in present northern Kansas. Apparently a good observer,he commented on the relative scarcity of Indian earthworks west ofthe Mississippi ; and, intrigued by a group of eight mounds arrangedin form of a cross about a mile west of Fort Leavenworth, he openedone on October 4, 1830 (Barnes, 1936 b, p. 360). He reported thatthe mounds were composed "of stones and earth, the former placedin a circle." Within the circle imcovered, were found fragmentarybones of adults and children which had been "under the action of 84 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174 fire. It appeared that after the bodies had been subjected to theaction of fire, without being consumed entirely, they were coveredwith earth." McCoy conjectured that "human sacrifice had there beenoffered" (McCoy, 1840, p. 408). As will become apparent presently,McCoy's observations closely parallel, except in his final conjecture,those since made by others, including the writer, at other burialmounds in northeastern Kansas (p. 172).Another mound in the same locality, possibly even in the samegroup, was explored a few years later by "three gentlemen from thegarrison, with a view of removing a doubt as to their being naturalor artificial." Again, only a few small fragments of bone werefound. Lieutenant Carleton, reporting this work, observed that otherhilltops in the vicinity also were crowned by mounds (Carleton, 1943,p. 163).Kansas was admitted to statehood in 1861, and in the next twodecades her population grew from slightly over 100,000 to nearly amillion persons. Inevitably, the vestiges of earlier human occupa-tions of the land came more and more to the notice of the white set-tlers, farmers, and others. By 1866, the State university and an agri-cultural college had been established, and scientists from these insti-tutions were engaging in widespread investigations into the geology,natural history, and other resources of the State. The growing in-terest in these matters led to organization in 1868 of the Kansas Nat-ural History Society which in 1871 became the Kansas Academy ofScience. Scattered through the published reports of its early meet-ings are many notes that bear testimony to the abundance of archeo-logical remains throughout the eastern half of Kansas. Both scien-tists and laymen shared in this early reporting ; and while its qualityvaried a good deal, promising leads may still be found in some of thepapers. Other notices appeared elsewhere in periodicals in and out-side the State.In a meeting of the American Philosophical Society in 1868, Mil-ler discussed certain geologic features along the Union Pacific Rail-way in Kansas. He called attention to Indian and later petroglyphscai-ved in the soft sandstone of Inscription Rock, located 15 milessoutheast of Fort Harker on Smoky Hill River (Miller, 1869, p. 383,pis. 7, 8). Some of the accompanying figures were subsequently in-cluded by Mallery (1893, fig. 44) in his monumental work on Amer-ican Indian picture writing (see also p. 483) . In a paper read before the sixth annual meeting of the KansasAcademy of Science in 1873, Mudge reported three localities wherehe had seen potsherds and other refuse of former Indian occupancy(Mudge, 1896). One of these was about half a mile north of theSanta Fe Trail crossing of Cow Creek, in Rice County (see p. 323) ; Wedel] an introduction TO KANSAS ARCHEOLOGY 85another was near a fine spring in the extreme northeast comer ofRiley County; and the third was near Asher Creek and SolomonRiver, in Cloud County. Of particular interest is the last location,regarded by Mudge as most promising of the three. Here he re-ported much clay "mixed with straw," as well as "fragments of whatappeared to be the ovens in which the pottery had been baked." Iam inclined to suspect that what Mudge actually saw here was grass-impressed wattling clay or daub, probably from the covering ofburned earth lodges. Wliether any of the sherds, grass-impressedclay, or other objects from the site were saved and, if so, their prasentwhereabouts, remains unstated.Five years later, in another paper read to the Academy at its elev-enth meeting, Adams described the results of recent digging by him-self and others in mounds near Fort Leavenworth (Adams, 1906).This work was evidently in mounds near those reported by McCoy(1840, p. 408) and Carleton (1943, p. 163), perhaps even in the samegroup. Adams' findings were, in the main, confirmatory of the ear-lier observations, except that he suggested the former presence ofvaultlike dry-masonry structures within the mound.For the next decade there are several records of mound observa-tions. In 1879 Curtiss investigated a burial mound and several rockcairns in the vicinity of Marion, and a collection of bone, stone, andother objects resulting from this work was deposited in the PeabodyMuseum (Putnam, 1880, pp. 718, 738). Further infoi-mation on an-tiquities of this interesting locality was published a few years later byBillings, a local collector in Marion. He reported several moundgroups, suggested that some of the mounds were house ruins, anddescribed deep ash beds which suggest cache pits. No artifacts aredescribed ; but Billings suggested that the relics found fell into threeclasses: Mound Builders, Cremators, and Modem Indians (Billings,1883).At about the same time, Robinson (1881, p. 446) reported manystone heaps and other antiquities, including pottery, on Wolf Creek,a tributary of the Neosho near Burlington, in Coffey County ; Stock-ton (1883, p. 685) reported a cave with carvings on its walls nearToronto, in Woodson County, but said there were no mounds in thevicinity; and Serviss (1883, p. 528) stated that there were "on thefarm of J. L. Stockton, 1 mile northwest of [Wyandotte City, nowKansas City, Kans.], remains of an aboriginal workshop or vil-lage . . . located on a small stream called Jersey Creek, and near alarge spring. It covers about 2 acres. . . . The fragments of potteryare the most numerous . . . composed of a mixture of clay, sand,and pounded shells. The variety of the combinations of lines anddots is inexhaustible . . ." He reported also several mounds nearEdwardsville, subsequently excavated by Brown for the University 86 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174of Kansas (see Wedel, 1943, p. 102). Parker (1887, p. 72) reportedbriefly on burial mounds on the Eepublican River bluffs 3^/^ milesnorthwest of Junction City, where fire-blackened bones and stones,along with pottery, beads, and other artifacts, were found.In contrast to the cursory notes and observations cited above wasthe work carried out by Udden in 1881-88, at the Paint Creek villagesite south of Smoky Hill River, a few miles southwest of Lindsborg,McPherson County. An instructor at Bethany College at the time,and later a geologist of considerable standing, Udden devoted hisleisure time during some 7 years to explorations here. Like thosewho have since investigated this site, Udden found an abundance ofpottery and a variety of stone and bone artifacts, all of which hedescribed and interpreted with commendable clarity and fullness.He surmised that the former inhabitants had been semihorticultural,and suggested possible relationships to the Wichita or Pawnee. Dis-covery of a fragment of chain mail in one mound and of two glassbeads on another persuaded him that the site had been visited bythe Spaniards, perhaps even by members of the Coronado expedition.The conclusions tentatively reached by Udden in his clear objectivepresentation (Udden, 1900) parallel in gratifying fashion those sincereached by others (see, e. g., Wedel, 1935) with a far richer back-ground for comparative purposes and interpretations. Basically, hisreport stands as one of the few bright spots in Kansas archeologyto date. Despite Udden's disclaimer to the designation of arche-ologist, one wonders whether archeology was not the loser whenhis later career led him into full-time geologic studies.The next decade, 1890-99, saw additional records of archeologicalobservations in various parts of the State, including some that havehad lasting value. In 1890 Mead reported on antiquities about theconfluence of the Arkansas and Little Arkansas Rivers, on the presentsite of Wichita, in Sedgwick County, where he operated an Indiantrading post among the Wichita in 1864 (p. 68). He noted thatplowing turned up many objects of an earlier day, especially largenumbers of small, triangular, unnotched arrowpoints. Other pot-tery-bearing village sites occurred on Chisholm Creek, 2 miles southof Wichita, and on the west bank of the Arkansas 4 miles south ofthe city. Caving of the banks of the Little Arkansas exposed a pot-tery vessel, undescribed, at a depth of 6 feet. Mead saw no obsidianand no painted or glazed pottery, hence no evidence of southwesternrelationships or contacts (Mead, 1890). Four years later, Newlon(1894, pp. 192-193) noted that there were many mounds along ShoalCreek, tributary of Spring River near Galena, as well as village sitesalong the Ozark foothills nearby. To him, the remains suggestedtwo peoples or epochs. Wbdkl] an introduction to KANSAS ARCHEOLOGY 87In 1896-97 Gould investigated the Timbered Mounds on the KawReservation just south of the Kansas State line. Here he at firstsaw "rude edifices" with "faint outlines of structures;" but Willistonconvinced him that these were actually Indian chert quarries. Theold workings were scattered in two localities on both sides of the Stateline. Gould observed that the fusulinid chert occurring here in nodu-lar form also occurred as artifacts in ancient Indian mounds 20miles to the northwest, near Arkansas City, Kans. (Gould, 1898 a,1899). My own observations on these old workings are reportedelsewhere in this paper (p. 476). As regards the Arkansas Citymounds, there are brief accounts from this same period in severalplaces (Jolmson, 1897; Gould 1898 b). All tell briefly of investiga-tions on the Beavers farm 2 miles east of the city, where moundexplorations yielded a variety of pottery, stone, and bone remains,as well as features that cannot be certainly identified at this time.Of especial interest is one of the several pots found; according toJohnson (1897, p. 96), it was "highly decorated and colored, and avery superior piece of work. The decoration consists of three parallellines, one of blue and two of red, running around the pottery at thetop. The rest of the piece is light yellow and the material composingit is light and fine." Assuming that this was an honest attempt, byone with limited archeological experience, to describe what he saw, Jthink it can safely be said that the vessel was not of local Indianmanufacture; but was it, perhaps, a late puebloan piece from theupper Rio Grande? The site is without much question that I havesince termed the Arkansas City Country Club site, in which I madelimited excavations in 1940 and where puebloan glaze-paint sherdswere found. Unfortunately, I do not know the present whereaboutsof the vessels described by Johnson, and so am unable to check hisobservations. Elsewhere (p. 351), the results of my mvestigations onthe site are presented in detail.To this period, beginning about 1895, belong the investigations ofJ. V. Brower in the lower Kansas River drainage. "Working withfarmers, local collectors, and other enthusiasts, Brower acquired avery large collection of archeological materials that was subsequentlycarried to Minnesota. He concluded that two cultures were repre-sented: the Quiviran remains, characterized largely by crude stonetools and absence of pottery, and occurring mainly south of KansasRiver from Geary County westward ; and the Harahey culture to theeast and north, wherein pottery, stone, bone, and other materials,sometimes evidently associated with semipermanent house remains,were characteristic. For the most part, Brower's inferences werederived from surface materials and from the results of uncontrolleddigging by his local supporters, generally no better trained in sci- 88 BUREAU OF AMERICAN ETHNOLOGY [Boll, 174 entific methods than he was. It now seems possible that some of his "Quiviran" specimens were actually workshop and quarry materials ? blanks, rejects, etc.?for which definite dating or cultural attributionare not yet possible. Others, however, suggest an early, possiblypreceramic, lithic industry. Among the objects Brower describedor figured as Harahey are included those of several distinct cultures ; some belong to complexes with rectangular earth lodges and evident"Upper Kepublican" affiliations while others appear to have Wood-land, Hopewellian, or other connections. I believe that most ofBrower's conclusions are either wholly untenable or else are stillsubject to verification or correction. However, anyone undertakingarcheological work in the area covered by him can ill afford to neglectaltogether his published reports, for they contain many provocativeand often profitable leads. There are two of these reports (Brower,1898, 1899), published as volmnes 1 and 2 of his series of memoirsof explorations in the basin of the Mississippi.Another interesting development of this period was Martin's dis-covery in 1895 of a chipped blade or point in direct association withthe bones of a large fossil bison. The locality was about half a milenorth of Smoky Hill River, on Twelve Mile Creek, some 12 miles eastof Russell Springs, in Logan County. The bones were in a bed of "fine silty material, blue-gray in color. Overlying this were twentyfeet of the so-called plains marl . . . The bone bed, when cleared off,was about ten feet square, and contained the skeletons of five or sixadult animals, and two or three younger ones, together with a foetalskeleton within the pelvis of one of the adult skeletons. The animalshad evidently all perished together in winter" (Williston, 1905 a, p.336). In another accoimt, Martin gives the vertical depth of theartifact, which lay beneath and in contact with the right scapula ofthe largest animal, as 25 feet (McClung, 1908, p. 250). Williston'sinitial identification of the bones as Bison antiquus Leidy was laterchanged by Lucas to Bison occidentalis (Williston, 1905 a, p. 335),which Skinner and Kaisen (1947, p. 171) regard as "a later animalon the plains of North America than antiquus^ but [which] was thesuccessful surviving species that gave rise to the recent plains bison."So far as I know, the circumstances of this find have never been ques-tioned, and the association is apparently accepted by paleontologistsqualified to judge the merits of the case (Sellards, 1940, p. 387) ; butas in many other comparable occurrences, the possibility of postglacialsurvival of the species involved has been raised (Romer, 1933, p. 79).According to Schultz and Eiseley (1935, p. 312), "In the case of Wil-liston's find this assumption of a sub-Recent dating is certainly opento question, the site having been reputably diagnosed as Pleistoceneand containing, besides Bison occidentalis., Elephas columbi and Wbdbl] an introduction to KANSAS ARCHEOLOGY 89Platygonus compressus.'''' With reference to the latter two species,Williston's report on Martin's find reads as follows : "The materialcovering the [bison] skeleton is the characteristic plains marl, inwhich and not far distant from the site of the arrowhead I have ob-tained bones of Elephas primigenius and Platygonus compressus^ bothcharacteristic pleistocene mammals of the so-called Equus Beds, orSheridan Stage" (Williston, 1905 a, p. 337) . Here, as elsewhere inAmerica at that time, I suppose there would have been no questionabout the Pleistocene dating of the site had there not been evidenceof human associations.The point itself has since been lost, though there are photographsof it. From the illustrations by Williston (1905 a, p. 336) , it appearsto have had a relatively short broad outline, with convex edges andslightly concave base; there are no shoulders nor is there definitefluting of the faces. There is nothing outstanding in the chippingand workmanship. It does not conform closely to any of the recog-nized early Plains types, such as Folsom fluted, Clovis, San Jon, Eden,Scottsbluff, Plainview, etc., although Howard (1935, p. 144) thoughtit resembled "a rather crude type of Folsom-like point . . ."; noris there any specific resemblance to points reported from Nebraska inassociation with Bison occidentalis (Barbour and Schultz, 1932, fig.165). Curiously enough, the find is not mentioned in some of themore recent surveys by archeologists (see, e. g., Roberts, 1940; Worm-ington, 1957) of the Early Man problem in the New World ; but Sel-lards (1940, p. 387) devotes a page to a good summary, and in anotherplace (1952, p. 47) terms it the "first discovery in America of anartifact known to have been used by early man in hunting bison ofa species now extinct."Before the close of the century, Williston and Martin made anothernoteworthy contribution to Kansas archeology. In 1898, they ex-cavated a seven-room pueblo ruin in the Beaver Creek valley, about10 miles due north of Scott City, and several good papers reportedtheir findings here (Williston, 1899; Williston and Martin, 1900;Martin, 1909; see also Hodge, 1900). The excavations yielded con-siderable quantities of maize, as well as metates, pottery, incised claypipes of late puebloan style, bird-bone flageolets, an iron axhead, anda variety of objects of chipped and ground stone and of bone. Theremains of a fireplace and chimney were found where they had beenbuilt into the wall of one room ; in another was a typical pueblo grind-ing trough ; and nearby irrigation ditches were reported to follow thelines of earlier ditches found by the first white settlers in the valley.Williston's first conjecture was that the ruin marked the site of anearly Spanish establishment; but Hodge correctly noted that thearchitecture and other features were typically puebloan, and sug- 90 BUREAU OF AJVIERICAN ETHNOLOGY [Boll. 174gested that the ruin was that of the fortified spot known as Cuartelejo,to which late 17th and early 18th century Pueblo Indians fled fromthe upper Rio Grande (see p. 23). In 1925 a granite shaft waserected at the spot by the Kansas Society of the Daughters of theAmerican Revolution, with the following inscription: "This marksthe site of the Picurie Indian pueblo 1604 [sic] which became an out-post of Spanish civilization and a rendezvous for French tradersprior to 1720." The results of my own investigations here for theNational Museum in 1939, which were concerned chiefly with refusedeposits near the old foundation and exploration of various lesserfeatures nearby, are detailed elsewhere in the present paper (p. 424).Shortly after the turn of the century came another find which at-tracted attention far beyond the borders of the State since it ap-parently bore on the then lively question of glacial man in America.This was the accidental discovery in February 1902 of skeletal partsof two humans beneath 20 feet of undisturbed soil on the Concannonfarm near Lansing. The site was a small bench overlooking theMissouri River bottoms, just inside the bluff line at the mouth of adeep ravine. Soon after the discovery, the site was examined by anumber of distinguished geologists and anthropologists; and in asearch for further relevant geologic data, Fowke under auspices ofthe Bureau of American Ethnology, made additional excavationsnearby. Before the end of the year, no less than six geologists hadpublished their views on the find (Sellards, 1940, p. 386). Appar-ently, no one questioned, then or later, the occun-ence of the bones asreported by the finders; according to Fowke (Hodge, 1907-10, pt. 1,p. 759), they "denoted an intentional burial."Williston, who first announced the discovery (Williston, 1902 b,1905 b), denied that the bones were of glacial age, as claimed in somecontemporary newspaper accounts. In his view, the materials over-lying the bones were certainly river-deposited, not aeolian, and thusdated from a time when the Missouri River flowed at a much higherlevel than at present. He held that the bones were indeed old, repre-senting people who were contemporaries of the mammoth, mastodon,extinct bison, camel, and other ancient forms. Upham, Winchell, andHaworth were also among those who believed the bones were veryold, the first two indeed arguing vigorously for their contemporaneitywith the lowan stage of glaciation (see Chamberlin, 1902, p. 745, andWinchell, 1903).Opposed to belief in such great antiquity were Chamberlin, Salis-bury, Calvin, Holmes, and Fowke. In a detailed discussion of thegeology of the site, Chamberlin (1902) contended that the fill con-sisted of "lodgment deposits derived from the upper slopes" of the Wbdbl] an introduction TO KANSAS ARCHEOLOGY 91 nearby bluffs, and of "silts blown up from the Missouri bottoms. . . .The antiquity of the burial is measured by the time occupied by theMissouri River in lowering its bottoms, two miles more or less inwidth, somewhere from 15 to 25 feet, a very respectable antiquity,but much short of the close of the glacial invasion." Calvin and Salis-bury endorsed this position (Chamberlin, 1902, pp. 777-779), andTodd (in Winchell, 1903, p. 293) also seemed inclined to accept itin preference to Winchell's stand. Holmes (1902, p. 751) summarizedthis view as follows : The preferred interpretation of the phenomena, then, is that the relic-bearingdeposits of the Concannon bench were not laid down in glacial times by thesilt-charged waters of the Missouri, but that they are a remnant of delta-likeaccumulations formed in comparatively recent times within and about themouth of the tributary valley by local sub-aerial agencies, all save the moreprotected portions having been removed by late encroachments of the ever-changing river.The bones represent an adult male and a child of 6 or 7 years. Theadult skull was repaired and first examined by Dorsey (in Holmes,1902, pp. 744-745), who noted that "in its general shape the skullbears a striking resemblance to the crania of the Plains Indians, forexample, the Blackfoot." " All the bones, it appears, subsequentlyimderwent a careful scrutiny by Hrdlicka. He concluded (Hrdlicka,1907, p. 52) that "the Lansing skeleton is practically identical withthe typical male skeleton of a large majority of the present Indiansof the Middle and Eastern States." In support of this statement,he sought to show in a tabular summary that there were no significantdifferences between the Lansing skull and a selected series of recentcrania from the same general region?including Ponca, Kansa, andtwo Pawnee?in the National collections. However, comparison ofthe given measurements and indices with similar items in the catalogof Siouan, Caddoan, and other crania published 20 years later(Hrdlicka, 1907, p. 52; cf. 1927, pp. 66, 80), raises some puzzlingquestions. The two Pawnee skulls in the 1907 comparative table, forexample, are longer and higher than any of the three listed in the1927 catalog; and the cephalic index given for the first two?74.3 ? compares with indices of 80, 81.5, and 82.9 in the catalog. The Poncaskull ("796, National Museum") of the comparative table is clearlyNo. 225285 of the catalog, in which it has the second lowest cephalicindex of 11 male Ponca averaging 79.3. For the Lansing skull,Hrdlicka gives a cephalic index of 73.5. I find no "Kaw" (Kansa)crania in the catalog. It is difficult to resist the suspicion thatHrdlicka's comparative table is slanted toward a desired end, thatis, to prove a point?the lateness of the Lansing skull; if not, theapparent errors and discrepancies remain to be explained away. In "HrdliCka's catalog of human crania (1927, p. 92) includes no male Blackfeet For aseries of 11 male Piegan, close relatives of the Blackfeet, the averages in the catalog, 92 BUREAU OF AMERICAN ETHNOLOGY [Boix. 174any case, I am unconvinced that Hrdlicka proved his point with theexamples he used."In actuality, the Lansing skull resembles less the recent groupsselected by Hrdlicka than it does certain prehistoric crania from burialmounds along this stretch of the Missouri. For example, there appearto be close similarities to the Kansas City vault mound (Hopewellian)series described by Stewart (1943, pp. 261-265, and table 13), whonoted further that the Lansing skull is higher headed than the averageSiouan or Caddoan skull. Precise classification or typing of the speci-men is, of course, a problem for the physical anthropologist, who todayhas much more comparative material at his disposal than Hrdlickahad. Meanwhile, it seems to me that the Lansing skull is in muchmore congenial company among such older, generally long-headed,types as are found with Hopewellian, early Woodland, or perhapsArchaic horizons than among the Siouan, Caddoan, or other recenttribes of the eastern Plains.It is interesting to note that the lively controversy briefly touchedon here revolved mainly about two points: (1) the glacial or post-glacial age of the deposits, and (2) the similarity or dissimilarity ofthe bones to those of modern or late Indians of the region. On bothscores the preponderance of professional opinion was, and presumablystill is, against great antiquity for the Lansing remains. Needless tosay, the depth of overburden is of itself not necessarily indicative ofgreat age, and unfortunately there is still no reliable method of datingaccurately the length of time involved in the deposition of such ablanket. Still, I find it difficult to believe that an undisturbed depositthat some competent geologists, like Williston, sincerely regarded as Insofar as they are directly comparable to the table HrdliCka (1907, p. 52) uses to showLansing Man's physical relationships, bear out Dorsey's observation as quoted. Thecephalic index of the Piegan skulls ranges from 73.1 to 78.8, and averages 75.8 as against73.5 for the Lansing skull." On January 21, 1954, with M. T. Newman, associate curator of physical anthropology,U. S. National Museum, I reexamined the four skulls used by HrdliCka in the above com-parison. Although we checked none of the measurements, partial answers to certain ofthe puzzling questions now suggest themselves. All four specimens have been renumberedsince the report on the Lansing remains. The two Pawnee skulls are numbered 225291(old number 550) and 225292 (old number 531). Inside the first, now sexed as a female,we found a note clearly in HrdliCka's handwriting, which said "Not Pawnee (mixed withwhites?or a white?)." The second also contained a handwritten note, initialed "A. H.,"and reading as follows: "This is not a Pawnee (mixed with white? or Algonquian?)."The "Kaw" skull. No. 152 of the comparative table, has the permanent number 225290 ; on the skull, in Ink, the identification is doubly questioned, thus "Kaw??". Thus, threeof Hrdlifka's examples turn out to be of doubtful ancestry or tribal affiliation ; the fourth,as already noted, is near the end of a group range and not an average. 1 suppose thereis no way of determining when HrdiiCka revised his Identifications ; perhaps the natureof the revisions had something to do with the omission of these particular specimens fromthe final catalog. In any event, the dubious nature of the evidence In this case seemsclear; the reader may judge for himself the validity of any conclusions based thereon byHrdlKika. Wedbl] an introduction TO KANSAS ARCHEOLOGY 93 early postglacial could be as young as the anthropologists strove tomake it appear. Hrdlicka's selection of skulls from tribes who almostcertainly entered the region in the past five to eight centuries, andHolmes' remark that "the time required by the river to lower its bedfive or ten feet might possibly be expressed in hundreds rather than inthousands of years" (Holmes, 1902, p. 751), doubtless reflect the wide-spread opposition among professional anthropologists of the periodto the idea of ancient man in America. Chamberlin, who had no suchax to grind, was apparently willing to concede a considerably greaterage?"a very respectable antiquity"?to the deposits ; and this wouldseem to fit better the modern views as to similarity of the Lansingskull to prehistoric physical types of the lower Missouri Valley two orthree thousand years, or more, ago. If Lansing Man was not glacialor early postglacial, neither was he recent Siouan or Caddoan ; and thepossibility of a respectable antiquity seems very good. He may wellrepresent one of the earlier ceramic or late preceramic peoples of theregion. Should the time ever come when a comprehensive programof integrated archeological and geological investigation can be carriedout on the buried creek-valley and terrace sites along the borders of theMissouri River trench, with particular reference to prehistoric man,I believe the Lansing locality will have to be further considered alongwith such sites as Sterns Creek, in Nebraska, and with others stillunreported and undescribed.In the decade following Lansing Man's discovery, there seem to havebeen no systematic excavations or otherwise outstanding developmentsin Kansas archeology. Brief notices of local digging or of casualobservations continued to appear, however, especially in the pagesof the Archeological Bulletin. This little journal, quarterly publica-tion of the International Society of Archeologists, ran to nine volumesbetween the years 1909 and 1918. For a time it was published inCouncil Grove, and some of its most active supporters and contributorswere residents of Kansas. And, as before, other notes continued toappear in the transactions of the Kansas Academy of Science, withoccasional short papers published by the State historical society.In 1904, Richey reported that the historical society had receivedIndian relics from the Kansas, Republican, Smoky Hill, Verdigris,and Marais des Cygnes Rivers, and commented on recent findings atthe Paint Creek site near Lindsborg and on Gypsum, Holland, andTurkey Creeks nearby to the east. Of greater interest are his observa-tions on ? a very interesting village site and fort on the Verdigris river . . . three milesnorth of Neodesha, near to and east of the river. The lodge sites occupy a con-siderable area, and the village seems to have been an important one . . . Shells,stone mauls, flint arrow-points, hammers, rubbing stones, scrapers, pitted stones. 94 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174 flint chips, and other objects were found on the site . . . The form of this fortis almost that of a horseshoe, with the opening toward the east ... A pieceof the butt plate of a gun and an old iron ax beveled only on one side wereunearthed near the fort ; also bullets and trinkets probably traded to the Indiansby white traders, were found.These articles, of both Indian and white origin, were said to have beenplaced on display at the society's museum (Ricliey, 1904, pp. 135-136).In the same publication, GrifRng described briefly findings by himselfat village sites and in mounds on Wildcat Creek, and on the Republicanand Kansas Rivers, further noting that "I have either opened orassisted in opening more than 100 of these burial mounds" (Griffing,1904, p. 134). In 1905, Mead wrote about remains on WhitewaterRiver near Augusta, including "a former soil surface two feet thick"beneath 8 feet of clay or loess; "on the surface of the black soil andunder the 8 feet of clay, I found the remains of a camp, containingbroken pottery, charcoal, ashes, burnt bone, and stones. . ." (Mead,1905). There are also terse notices of antiquities along DiamondCreek, in Morris County, by Morehouse (1910) ; of ancient house pitson the Missouri River bluffs and of pottery sites along Wolf River,in Doniphan County, by Plank (1910) ; and of a refuse-littered oc-cupation area with "at least 20 tipi sites" on a hill near Sparks, byDinsmore (1912; seep. 132).Among all these amateurs, the name which appears most often isthat of George J. Remsburg. He contributed frequently, if usuallybriefly, to the pages of the Archeological Bulletin concerning antiqui-ties in northeastern Kansas, chiefly in Atchison, Leavenworth, Potta-watomie, and Riley Counties (Remsburg, 1893, 1894, 1909, 1910, 1911,1912 a, 1912 b, 1915 a, 1915 b). These included village sites, historicand other burials, and mounds, some of which seem to have beenchambered. He also reported (Remsburg, 1912 b, p. 122) rock carv-ings of which he was informed in Labette and Woodson Counties.Probably his most important contribution was the identification of re-mains at Doniphan with the Kansa village visited by Bourgmond in1724, and the assembling of all data he could get from early residentsregarding the features of the old Indian townsite (Remsburg, 1919;see also p. 101). He claimed also (1911, p. 67) to have located thesite of the second and later old Kansa town farther downstream, justnorth of Fort Leavenworth; but I have not been able to verify this,and the only remains I had opportunity to examine in this localitywere of much earlier, presumably Hopewellian, affiliation (Wedel,1943, p. 102). Probably the numerous short statements publishedby Remsburg contain leads that would still repay follow-up work.In 1913, the presence of glacial man in the Kansas plains wasproposed again, this time by a Minnesota geologist and one-timepresident of the Geological Society of America. Winchell, one of Wbdbl] an introduction TO KANSAS ARCHEOLOGY 95the chief protagonists of glacial antiquity for the Lansing Man, hadby this time carefully reexamined Brower's extensive chipped-stonecollections from the Kansas River valley; and, stimulated by cor-respondence between Brower and Thomas Wilson, curator of pre-historic archeology in the National Museum, he set out to demonstratethat the "coarsely chipped large artifacts which Mr. Brower at-tributed to the Quivirans (Wichita Caddo)" were in reality "dis-tinctly paleolithic and manifest all the characteristic features of thepaleolithic artifacts of Europe." Nearly all were patinated in oneway or another, he claimed, and many had been rechipped by laterpeoples. He inferred that there had been two stages, plus severalsubstages, of ancient culture near the Kansan chert beds : the oldest,a paleolithic complex, which antedated the Kansas glaciation, anda later neolithic stage, or stages, occupying post-Kansan time. Lan-sing Man, he insisted, was from the lowan glaciation, and "probablytook part in the making of some of the Early Neolithic implementsfound on the Kansas upland interior." All of this seems some-what farfetched today; and the chief merit of Winchell's book(Winchell, 1913) is perhaps in its halftone illustrations of the roughlyfashioned cores, quarry blanks, blades, and other chert artifacts thatonce occurred in such extraordinary abundance along the belt wherethe Kansas River traverses the Flint Hills upland. As a sign of thetimes, Winchell's preface censured "the powerful influences thatare localized in Washington" for denying glacial man and paleo-lithic implements in America. In a. reply that Winchell probablynever saw, one of those "powerful influences," who shall be namelesshere, wrote on the fly leaf of his book, "This work is the best illus-tration of utter error and silly foolishness ever issued from the press."The next two decades, roughly from 1912 to 1932, were essentiallya continuation of the preceding ones, with sporadic observations, littlesystematic fieldwork, and no comprehensive publications. Early inthis period, from 1912 to 1914, F. H. Stems was engaged in an in-tensive archeological survey of the Missouri River bluffs zone ineastern Nebraska for the Peabody Museum of Archeology and Eth-nology, and this work was carried southward into the two Kansascoimties of Doniphan and Atchison.^^ Sterns established the ex-istence of the Rectangular Earth Lodge culture, since designated theNebraska Culture, in Northeastern Kansas. He also noted a largemound near Iowa Point, which had been extensively looted by relic " This Btudy remains In manuscript, and only one of the two short published papers basedon It makes brief mention of Kansas findings (Sterns, 1914). My remarks here are basedon notes made by Dr. W. D. Strong, who has discussed Sterns' materials at some lengthand has related them to later findings In Nebraska archeology (see Strong, 1935, pp. 49,250-266) ; and on a brief examination of the Sterns' collections from northeastern Kansas,DOW In Peabody Museum, Harvard.484172?59 8 96 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174hunters; and 3 miles south of White Cloud, on Cedar Creek, he re-ported that rains had washed out pottery bearing wickerwork andcord impressions, and with punclied rim bosses. He made a few ob-servations on a late site on Wolf Creek, presumably that describedelsewhere in the present paper as the Fanning site (see p. 131), notingthat the pottery liere was shell tempered, without cord markings, andresembled that from a large campsite, now known as the Leary site, onthe lower Nemaha Kiver (see Hill and Wedel, 1936). For the WolfCreek site he suggested a Kansa origin.At about the same time, Gerard Fowke briefly surveyed this por-tion of the Missouri Valley, from Doniphan County, Kans., to thevicinity of Omaha, but apparently without doing any excavating(Fowke, 1922, pp. 151-154) . The Taylor mound, near ^Yliite Cloud,Kans., which he was told was the largest mound in the State andfrom which local collectors had taken 50 or more skeletons, is evi-dently the same mound Sterns reported. From my own observa-tions here in 1937, I believe that a good deal of information is stillto be gotten from this structure, since the mound appears to be atleast 5 or 6 feet deep and 30 yards or more in diameter, and most ofit was said to be still untouched by local digging. Near this mound,Fowke reported a very large house pit, still visible in 1937; othersoccur nearby in some numbers. Near Iowa Point, Fowke reporteda mound, or mounds, with slab-cist and stone-vault( ?) burials. Healso mentions the Leary site on the Nemaha, erroneously ascribing itto the Pawnee, who were supposedly dealt a crushing defeat here in1837 by the Iowa and Oto (cf. Hill and Wedel, 1936, p. 66) ; and hecould see no difference between the remains here and those found inthe house pits nearby, "unless it may exist in the markings in tliepottery . . ." Fowke seems to have shown less insight and exer-cised a good deal less restraint than Sterns in his evaluations of thearcheological potentialities of the region and of the significance ofthe materials he saw. He concluded this section of his report on thefollowing note: "Any estimate of age must be only a guess at thebest, but it is a safe guess that no earthwork, mound, lodge site, orhuman bone along this part of the Missouri River has been here aslong as 10 centuries."To what extent Sterns and Fowke were aided and perhaps in-fluenced by local persons in their northeast Kansas observations is notclear ; but it may be presumed that both were in contact with MarkE. Zimmerman of 'WTiite Cloud, self-styled State archeologist ofKansas, and his assistant, Ed Park. These two men, with no pro-fessional training and possessing more enthusiasm than sense of scien-tific problem, did considerable collecting and some digging in north-eastern Kansas and in adjacent districts. Zimmerman's reports con- Wbdbl] an introduction TO KANSAS ARCHEOLOGY 97tain some useful data on village sites and other local antiquities, butare beclouded with long discourses on various esoteric matters andstrange theories ; they contain, in short, a great deal more chaff thangrain (see, e. g., Zimmerman, 1918). Yet, as we shall see presently(p. 620), these observers had arrived at a local scheme of culture se-quence which is substantially in line with present archeological views.Elsewhere in the State, iUlison and Eandolph about this time pub-lished a short paper on antiquities in the vicinity of Pittsburg. Theseincluded chiefly chipped stone, with only passing mention of pottery.They also proposed a local sequence going back to perhaps 12000B. C, and involving certain climatic variations (Allison andRandolph, 1927). Farther west, Paul Jones, newspaper editor ofLyons, issued the first of several books calling attention to abundantarcheological remains in Rice and nearby counties, and maintainingwith very good reason that these marked the location of the Quiviransettlements seen by Coronado (Jones, 1929, 1937 ; cf . Wedel, 1942, andBolton, 1949, pp. 284-304). These remains were also the subjectof a paper by Thobum (1927), and were briefly mentioned by Moore-head (1931, pp. 85-88) in his survey of archeological materials in theArkansas River valley. Moorehead's survey was a rather super-ficial one; it recorded a number of village and burial sites reportedby local collectors in Butler County and elsewhere in and near theArkansas drainage, but there are disappointingly few and vague de-tails regarding the nature of the remains. He also reported at somelength on observations by Thobum at "an ancient irrigation canal"in Clark County, near which were seen "two low earthen mounds, atpoints on its course, which were certainly the ruins of such [puebloan]structures" (Moorehead, 1931, pp. 90-91) . There has been, to date, nosubsequent confirmation of these puebloan ruins or of an "ancient"irrigation system in the locality.Mention must also be made of the work of certain nonprofessionals,which promises to contribute materially to our understanding of cer-tain aspects of Kansas prehistory. One of these was Floyd Schultz,formerly of Clay Center, whose excavations and surveys in the lowerRepublican River valley resulted in a large cataloged collection which,with accompanying records, is now in possession of the University ofKansas Museum of Natural History. Publication of this materialshould clarify our knowledge of "Upper Republican" and earlier com-plexes in the Kansas-Nebraska region. Similarly, the carefully donecommercially oriented excavations of G. L. Whiteford at a village andburial ground near Salina open interesting vistas on relationshipsbetween the prehistoric rectangular earth-lodge peoples of east-cen-tral Kansas and the more advanced cultures of the lower Arkansasvalley region (see also p. 512) . 98 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174 jFor present purposes, our survey of archeological work in Kansasprior to the investigations comprising the body of the present paperis substantially complete. We need mention only the further fact thatin 1934 an expedition from the Nebraska State Historical Societyopened three earth-lodge ruins at a village site on the Solomon Kiver 3miles south of Minneapolis, and made brief tests at other sites nearGlen Elder and at Udden's Paint Creek site near Lindsborg (Wedel,1935). The writer's investigations for the United States NationalMuseum were initiated in 1937 and continued until 1940; thefindings are detailed in the pages which follow. The results of morerecent investigations and studies by professional anthropologists atthe University of Kansas and by Missouri Kiver Basin Surveys per-sonnel will be considered in the relevant sections of this report. It isgratifying to record in closing, however, that, with the exception of afew years between 1942 and 1946, the State university has included aprofessional anthropologist on its staff since 1937. Thus, there isavailable today within the State competent professional advice andassistance to those local persons and others interested in the archeologyof Kansas. AECHEOLOGICAL INVESTIGATIONSIN NORTHEASTERN KANSASTHE DONIPHAN SITE (14DP2)Our investigations in northeastern Kansas began with a brief ex-amination of a site on the Missouri River bluffs at the village of Don-iphan. The locality, an attractive and historic one, centers in a littlevalley surrounded on the east, north, and west by prominent hills.Doniphan itself, with less than 200 inhabitants, retains little trace ofits own picturesque past. First settled as a trading post in 1853, itsoon became, on removal of the Indians, an important shipping pointfor grain, produce, and other commodities. With establishment of aUnited States land office in 1857 came a short-lived boom that sawsingle lots selling as high as $2,000 ; the townsite was surveyed, streetsgraded, and much residential and commercial construction undertaken.A year or two later, when the land office was transferred elsewhere,began a steady decline, temporarily arrested by the development of aflourishing grape-growing and wine industry in the 1870's. At itspeak, the town boasted two railroads, grain elevators, flour and saw-mills, two packing houses, a newspaper, general stores, saloons,agencies, offices, several hotels, and three churches; and in 1857, asagain in 1871, its population was estimated at upward of 1,000. In1859, Abraham Lincoln delivered a campaign speech in the town. Asmany as four and five landings daily were made by steamboats, and the Wbdbl] an introduction TO KANSAS ARCHEOLOGY 99 river warehouses did a rushing business in furniture, tools, farmingimplements, and other items.Today the railroads are gone, and the river channel is a mile or moredistant from the old landing. Most of the business houses, stores,residences, and other establishments have been burned down or re-moved, and the erstwhile commercial and industrial interests have longsince shifted to Atchison, a few miles downriver. Gone, too, are thesplendid Belleview and other vineyards that once covered the adjacenthills, though fallen-in wine cellars still remind one of the industrythat once produced thousands of gallons of beverage yearly. A storeor two, a school, a church, and modest dwellings widely scattered alongthe old streets and over the valley slopes are Doniphan today.The valley in which Doniphan lies, nowhere much exceeding half amile in width, is drained by a small unnamed creek rising a mile or twonorth of the town and flowing due south to empty into the MissouriRiver bottoms near the old steamboat landing. Virtually dry duringour 1937 dig, the creek was a running stream in the 1860's and 1870's,and is said to have occasional remnants of once-perennial springs atseveral points along its short course. It is now deeply entrenched, sothat the gently sloping valley bottoms consist of pleasant terraces,beyond which the ground rises on the east and west to ridges 120 feetor more in height. These ridges terminate in steep south-facing bluffs,against the base of which the Missouri flowed before 1900. The riverbottoms are partly under cultivation, but more of their area is over-grown with cottonwoods, willow, cattail, and the like. A fluctuatingbody of water known as Lake Doniphan marks the old course of theriver. The terrain east and north of Doniphan is rough and hilly ; tothe west, the valleys of Independence Creek and its tributary, RockCreek, lie just over the first ridge, beyond which the hills and ravinescontinue. Limited stands of hardwood forest fringe the streams andcover the steeper slopes of the bluffs, but most of the land is in culti-vation (pi. l,a).Although the former existence of a large and important Indianvillage on this spot has long been known, its precise location and extentseem never to have been set forth adequately by those who recognizedit in the early days. Our investigations disclosed relatively littlearcheological material on the west side of the creek, and despite re-peated inquiries, we were not able to learn of the discovery of ash beds,graves, cache pits, lodge sites, or other prewhite evidences of occupancyon the valley bottoms. On the rising ground east of the creek, how-ever, above the point to which Doniphan seems to have spread and justout of sight of the river valley, we found a group of 15 cache pitsscattered among the trees of a hillside orchard. On the winding ridgea hundred yards or more to the east two pithouse ruins were found ; 100 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174and beyond these, on a spur running toward the river, were severalsmall burial plots. All this suggests that the main area of Indianoccupancy lay somewhat to the east of the Doniphan townsite, or elsethat the tangible evidences of that occupancy farther downhill havebeen completely obliterated by tlie present community. The latteralternative seems rather more probable, since the lower slopes andterraces of the creek valley would hardly have been passed over byIndians in search of convenient habitation sites. I suspect that theprincipal village actually lay in the valley, and that the caches wefound represent but one periphery of the habitation area. The gravesat a distance of a few hundred yards to the east are also understand-able as parts of a cemetery belonging to a village in the valley ; andthe two house sites between, as we shall show, undoubtedly pertainto a distinct and earlier occupation of the spot and were not con-temporaneous with the caches and graves. How far up the creekvalley these remains occur, or formerly did, I am unable to say;and I have no basis whatever for estimating the extent of the area offormer Indian occupation.Our interest in the remains at this spot derived largely from thefact that the site is believed, with very good reason, to mark theprincipal village of the Kansa Indians when they were visited byBourgmond in 1724 (Kemsburg, 1919). The location cannot be veri-fied from Bourgmond's narrative alone, or from Renaudiere's state-ment that 30 leagues above the Quans [Kansas] river ? ... a small river [Independence Creek] flowing from the north is found;here is the great village of the Quans [Kansa], consisting of 150 lodges ad-joining the Missouri. There are fine prairies to the south and many mountainsto the west . . . [Margry, 1886, pt. 6, p. 393.]Eighty years later, however, Lewis and Clark (Thwaites, 1904r-5,vol. 1, pp. 66-68) camped near the mouth of Independence Creek,and so named it, on July 4, 1804, observing that it entered the Mis-souri "... a mile below the 2nd old Kansas Village." They com-mented further that ". . . this Town appears to have cov^. [covered]a large Space."; and note specifically that ". . . M. de Bourgmond,a French officer who Com'', [commanded] a fort near the Town ofthe Missouris in about the year 1724 and in July of the Same yearhe visited this Village at that time the nation was noumerous andwell disposed towards the French." ^* In 1819, Major Stephen Long's " By the Missouri River, Doniplian In 1937 was approximately 443 miles from theMississippi, and 63 miles above Kansas River. On Chart No. 89, Missouri River mouthto St. Joseph, Mo., l!>30-.'32, mile point 454 on the 1800 thalweg lies ca. 350 yards southby slightly west of the old Doniphan boat landing. As against this figure of 454 miles,measured along an unstralghtened channel that presumably approximated that taken byearly boat travelers up the river, we have Lewis and Clark's several estimates that thesecond old Kansa village was 405 and 431 miles from the Mississippi and 67 miles aboveKansas River. WKDKL] AN mTRODUCTION TO KANSAS ARCHEOLOGY 101 expedition ascending the Missouri (James, 1823, vol. 1, p. Ii3) alsopassed "... a small stream called Independence Creek. A littleabove and on the south [right] side of the river, is the site of an oldKonza town, called formerly the village of the Twenty Four . . ."Other travelers up the Missouri, both before and after Long, like-wise noted in their journals or on their maps, the existence of this "old," "ancien," "upper," or "2nd" Kansa village. The tribal identityof the principal historic Indian remains at Doniphan thus seemssettled beyond question; and Bourgmond's visit further establishesthe site as probably the earliest specifically identifiable Indian villagein the State.As to the archeological evidence for a former protohistoric Indianvillage of some size on the Doniphan townsite proper, it appears thatour rather meager and inconclusive findings may not reflect the truefacts. Historical works concerning the locality generally give littleinformation beyond passing reference to hilltop burials seen or openedon the bluffs. An early history of Doniphan County (Gray, 1905)records a visit to the hills near Doniphan by a group of medicalstudents in search of skeletal materials for study purposes?a practicewhich, if common, may help to explain the seeming paucity of burialstoday. A later note by Remsburg (1909, p. 2) states that he exploreda mound in this vicinity and found remains of a former scaffoldburial; also, that there "are numerous stone grave mounds on thehills about Doniphan." Still later, in the only published statementdealing specifically with the village site itself, Remsburg (1919, pp.1-11) cites a number of old residents to the effect that when the town-site was being developed there were "numerous hut-rings or lodgecircles of an ancient Indian village . . . exactly similar to those ofthe later day villages of the Kansas Indians at Manhattan, Valencia,Council Grove, and other places." The lodge circles, he says further, . . . with firepits in the center, were plainly visible in many places . . . [andwere] especially noticeable where the public school building now stands. Theearth in many places was intermingled with charcoal, ashes, and other debrisof the Indian village. Mr. [Luther] Dickinson says that as near as he canremember the rings or circles where the wigwams stood and which were quitenumerous, were about twenty feet in diameter and in the center of each wasa cavity filled with ashes and charcoal.Other informants recalled that "large masses of charcoal, pottery,and other burnt substances" were exposed by caving of the creekbanks ; that in and about Doniphan masses of limestone slabs occurredon and under the ground surface; and that tipi sites, ash, pottery,stone artifacts, etc., were often exposed during roadbuilding andin preparation of the A. & N. Railroad grade through the town.Mounds and stone-covered burials were said to have been plentiful ' on the adjacent hills ; and on the authority of McCoy and Morehouse, 102 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174Remsburg says the Kansa frequently buried their dead beneathstone slabs. Curiously enough, there appears to be nowhere a detaileddescription or adequate illustration of the innumerable artifacts saidto have come from the townsite; and it is utterly impossible to de-termine whether one or several cultures were represented, or to evaluatethe remains in light of present knowledge of Missouri Valleyarcheology.The remains described below, as already indicated, were found on theeast slope of the creek valley near its mouth and on the overlookingridge. All were within 300 yards of the river face of the bluffs, and notmore than half a mile from the mam road through Doniphan. Theylay on property owned by George Meidinger, to whom we are underobligations for permission not only to dig but also to camp in thetemporarily abandoned farmstead area on the site. The excavationswere carried on from July 1 to 23, 1937.CACHE PITSThe cache pits we opened were scattered down a sloping hillside(fig. 5) below the Meidinger farm buildings, running from the 900-footcontour westward to the 870-foot contour, at a point about 250 yardsfrom the main road through Doniphan. All were in an area measuringabout 65 by 130 yards. One pit had been sectioned by the windingdrive to the farmstead ; the others lay north of this road, and all wereconcealed from the main river valley by a slight shoulder of the hill.The location was well chosen for purposes of drainage ; but it has anorthwest exposure and must have been windswept and bleak duringthe winter months.With exception of the pit sectioned by the driveway, there was noclear indication of remains on this spot prior to actual tests. Closeinspection of a group of molehills, however, disclosed bits of charcoaland one or two small sherds, and a little digging revealed darkenedearth below mixed with more charcoal, burned clay daub, flint chips,and bone fragments. This sort of mixture, consisting obviously ofrefuse, characterized the fill in all of the pits, and set them off clearlyfrom the surrounding undisturbed red-brown clay subsoil.That most of the pits opened on the slope here were actually oldcache pits secondarily used as trash dumps, is indicated by their gen-eral regularity in form and the conformity in size, shape, and otherdetails to the similar pits found in other Plains Village Indian sites.Characteristically, they were circular in plan, with the floor slightlyexceeding the mouth in diameter, and of medium size and depth. Thediameter varied from 3 to nearly 7 feet, the depth from 2 to 5 feet.There was no evidence of burning of the walls, or of grass, skin, stone,or other floor coverings. The fill, generally less compact than the en- Wbdbl] AN INTRODUCTION TO KANSAS ARCHEOLOGY 103 104 BUREAU OF AMERICAN ETHNOLOGY [BULL. 174 closing subsoil, contained quantities of animal bone, lenses of ash,mussel shells, worked stone and bone, and occasional charred cobs andkernels of corn. Bits of brass and iron, and occasional recognizableobjects thereof, as well as glass beads, occurred in most of the pits,though not in large numbers. From several were taken lumps ofgrass- and twig-impressed clay daub, in all probability from the ruinsof earth-covered dwellings that once stood nearby. Sherds were dis-appointingly scarce and small.Location of the pits opened is shown on the site map (fig. 5) ; theirdimensions and contents are summarized in table 1.Table 1. ? Summary of dimensions and contents of pits opened at the Doniphansite mDP2) PitNo. 2344A66 DiameterTop Bottom Inches407410390X7040604312SX5036425078544550 Inches42 80128X5044625455575048 Depth Inches2445505543436036345331(?) Remarks Ashy fill, with lumps of clay daub; animal bone, copper scrap,sherds, etc.Clay daub, iron, sherds, bones, etc.Little admixture; some bone, metal objects.No data.Sherds, catlinite, iron, clay daub, etc.Copper; iron; stone club head, etc.Lead cross; copper, iron; glass beads; catlinite; pipes; worked andrefuse animal bone; sherds; etc.Sherds, flints, iron, copper, etc.; shape unusual, possibly two over-lapping or intersecting cache pits.Cylindrical neck 10 inches deep; sherds, iron, copper, catlinite, etc.Cylindrical neck 15 inches deep; flints, sherds, etc.Bone refuse, stone, metal, bone tube, catlinite, etc.Bone refuse, worked bone, sherds; no metal or glass.Bone refuse, charred corn, etc.Catlinite pipe, bone needle, copper, iron, chert points, etc.Sherds, flint and antler artifacts, metal, etc.Pits 3 and 4 possibly were not cache pits. Aside from their greatersize (table 1), both were oblong in plan, with walls vertical or inslop-ing at the base, and they lacked the regular shape of the usual caches.The fill, moreover, contained a much smaller proportion of refuse andartifacts. That they were from the same general time period as thecaches is evidenced by the presence of bits of metal along with aborig-inal materials. It was surmised that these basinlike affairs mighthave been borrow-pits, from which earth was taken to plaster housewalls or to cover earth lodges, and that subsequently the holes wereincidentally filled up with earth containing a little trash. Since thetype of house used by the people who built the caches is not known,the above surmise as to borrow-pits must be regarded as a guess.So far as the presence of orchard trees here permitted, we carefullytest-pitted most of the area about and among the caches in a searchfor house sites. The likeliest spot included pits 1 to 4 and the inter-vening area, where lumps of fire-hardened clay daub, many of themgrass- or twig-impressed, were scattered throughout the topsoil.Despite careful stripping of this area to subsoil, however, we found Wbdel] an introduction TO KANSAS ARCHEOLOGY 105no trace of postmolds, fireplaces, floor lines, or other house features.Uniformly negative, also, were our tests elsewhere in the orchard andsouthward across the farm driveway. Bits of clay daub were notedin several of our tests, but in no instance could they be traced to arecognizable habitation imit. The negative results of our searchcannot be considered final and conclusive, and it is still possible thatremains of house units coeval with the caches could be found by moreexhaustive testing. It is perhaps equally likely, however, that thedwellings from this period were mainly on lower and less exposedground to the west or northwest, or that, instead of substantial earth-covered pithouses, they were surface affairs with bark or mat cover-ings which would leave few or no traces after the lapse of a centuryand a half or more. The clay daub we found may have originatedin house ruins of an earlier period, located on the higher points andridge summits, whence the detritus was carried downhill by erosion,early-day farming, or other agencies.HOUSE SITESEast of the cache pit area, on the sloping ridge summit, we openedtwo pithouse sites. House 1 lay atop a rounded knoll (pi. 1, &) about300 yards east northeast of the Meidinger farm buildings (fig. 5),at the 950-foot contour. House 2 was on a narrow saddle about 200yards nearer the farmstead, just above the 900-foot contour. North-ward, the ground slopes away to a short, dry branch of the Doniphanvalley ; to the south is a much sharper drop to a deep, heavily woodedravine opening onto the Missouri River bottoms. Here again ourinvestigations were hampered by presence of fruit trees, so that house1, the larger of the two, could not be completely cleared. Despitethis fact and certain anomalies of form, there can be no doubt, how-ever, regarding the nature and cultural affiliations of these two struc-tures, and of their separateness temporally and culturally from thecaches and from the burials yet to be described.House 1, slightly over 36 feet in diameter, was circular in plan(fig. 6). The floor was generally poorly defined because of extensiveactivity by burrowing animals ; at the periphery it was 24 to 28 inchesbelow the present ground surface, tliis depth increasing to about 31inches near the center in the vicinity of the fireplace. The fireplacewas a simple unlined circular basin 30 inches across, filled with sev-eral inches of clean white wood ash, below which the ground wasfire-reddened to a depth of nearly 6 inches. Four large postmolds,marking the location of the primary roof supports, lay at an averagedistance of about 10 feet from the hearth and formed a central quad-rangle measuring roughly 12 by 18 feet. These postmolds were 9 to12 inches across by 13 to 18 inches deep, and contained charcoal frag- 106 BUREAU OF AMERICAN ETHNOLOGY [BOLL. 174 HOUSE I 4 8 12 HOUSE 2 Figure 6.?Floor plans of houses 1 and 2, Doniphan site, 14DP2. Solid circles indicatepestholes; rayed ellipses, fireplaces; small nucleated squares, orchard trees. Houses notshown in true relationship to each other. Wedbl] an rNTRODUCmON TO KANSAS ARCHEOLOGY 107ments and other trash. Smaller postmolds, 5 to 8 inches in diameterby 6 to 12 inches deep, and spaced at intervals of 15 to 48 inches, layat a radius of 17 to 19 feet from the fireplace. Undoubtedly some ofthe longer intervals in this series, as shown in our plan, once con-tained one or more posts, of which we were unable to find satisfactoryevidence. The wall of the house pit rose sharply to the ground sur-face a foot or two outside this row of secondary roof-wall supports.Some difficulty was experienced in locating the entranceway, butit seems certain this could not have been toward the east or west.Toward the south, however, mixed earth continued as a strip 3 to 4feet wide for some 5 or 6 feet beyond the outer circle of postmolds;and two postmolds flanked this strip about 2 feet beyond the circle.This I suspect was the entrance, though if so, it was poorly centeredwith reference to the house plan generally. The mixed fill roserapidly, but evenly, to the ground surface, so that a very short pas-sage, or one mostly at or near present ground level, is suggested.Immediately north of the northwest center postmold was a shallowcache pit 54 inches in diameter, from which came a fragment of claypipestem, hematite, a few sherds, and other refuse. There were noother caches in the excavated portion, and the only unusual featureof the house otherwise was the presence of a single postmold justinside the outer circle in each of the southeast, southwest, and north-west quadrants. Their significance, if they were used otherwise thanfor bracing posts, is not clear.As to the fill within the house pit generally, a typical section southof the fireplace showed 7 inches of loose, light-gray topsoil, obviouslydisturbed by the plow; 8 inches of dark-gray humus, without cul-tural admixture; 12 inches of light-gray, poorly compacted soil; and4 inches of soil heavily mixed with charcoal, burned daub, refuse, anda few artifacts, all of this admixture increasing in abundance down-ward toward the old floor line. It is probable that most of the darkerunmixed fill is of aeolian origin, and was deposited, perhaps withaccretions by modern farming operations, in the depression left bythe collapsed dwelling. I see no reason to suppose that the abandonedhouse pit was used to an appreciable extent as a trash dump by otherinhabitants of the village or by later peoples ; and it is probably safeto conclude that the materials found on and just above the floor and inthe single cache pit were left by the original occupants of the struc-ture. All of these materials, to be considered fully in another sectionof this report, were of aboriginal origin, and include nothing thatcould possibly be attributed to European trade contacts.During the excavations in and about the house pit, three burialswere uncovered. Burial 1, lying just outside the south wall of thehouse and about 7 feet west of the probable doorway, was that of a 108 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174 cliild 6 to 8 years old. The skeletx>n was extended and supine, withhead to south, the right arm at the side, and the left bent over thepelvis. A few inches from the right hand were the remains of a logor wood slab, not certainly worked. The clavicles, humeri, and lowerribs were heavily copper stained. Over the knees were several layersof wrappings. The outermost was of grasses or rushes that do notappear to have been woven, as no certain weft elements could be de-tected. Under this were several thicknesses of heavy dark-brownskin or leather, among the folds of which lay copper or brass cones,hawk bells, and small circular beads. Beyond much doubt, thesearticles had once been sewn onto the leather, though none remainedclearly attached. The grave pit was about 50 inches long, and thebones lay, fully articulated, from 17 to 19 inches below groundsurface (pi. 2, IFEETFigure IS.?Plan of house 1, Fanning site. Solid circles indicate pestholes; rayed ellipsefireplace; numbered circles, cache pits; broken line, approximate edge of house. 136 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174 all, or none may have been intrusive. Such potsherds and othermaterials as came from the house fill did not differ from those foundin the caches ; if the latter were earlier or later, no great time lapse ormarked cultural differentiation is indicated.The structure that once stood here, to judge by the postmold patternand bits of burnt grass-impressed clay, was in all probability an earth-covered semisubterranean affair of characteristic eastern Plains type.Despite persistent search, no additional house sites were found. Ourtests, for reasons already stated, were far from conclusive, and itwould not be surprising if further examination disclosed the remainsof other similar structures. At the same time, it is clear that withcultivation has come accelerated erosion, and at this late date shallow-lying house pits may well have been mutilated beyond any hope ofrecognition. There is, too, the strong likelihood that the usual habita-tions here were neither semisubterranean nor earth covered; if barkor mat-covered lodges placed on the surface were the rule, any tracesthat survived the active occupancy of the site may have disappearedwhen the ground was first broken.CACHE PITSCache pits evidently occur on the site in considerable numbers (fig.14). No less than 46 were opened during our 3-week stay (table 2).In general, they conformed as to size, shape, and contents to the cachesfound on most village sites of the later semisedentary horticulturalIndians of the eastern Plains. All were circular in plan, or nearly so,with constricted orifices ; maximum diameter came at or just above thebottom, which was usually flat. A few had vertical walls, with nosignificant variation in diameter from top to bottom. Size variedconsiderably ; diameter ranged from 28 to 94 inches at the mouth, andfrom 28 to 91 inches at the bottom ; depth, from 10 to 90 inches. In thedeeper pits there was sometimes a cylindrical neck from 1 to 3 feetdeep below which the walls flared outward and down. Marks pre-sumably left by the aboriginal digging tools were noted on the claywalls in several instances. In no case were the walls or floor fireblackened or hardened, and in only two or three instances was thereany evidence of a charred grass or other vegetal floor covering.As to contents, the pits generally contained an abundance of pot-sherds and broken animal bone, with lesser quantities of worked bone,chipped and ground stone, rejectage of artifact manufacture, musselshells, trade metal, and charred vegetal foodstuffs. This materialoccurred in an ashy mixed fill, usually quite soft, with bits of charcoaland quantities of dark-gray soil intermixed. Most of our artifacts,to be described presently, came from these pits. Wedel] an introduction TO KANSAS ARCHEOLOGYTable 2. ? Summarized data on cache pits. Fanning site 137 138 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174So far as our work is concerned, there seemed to be no significantdifferences in pits or in pit contents from the various parts of the sitetested, except perhaps in one detail. We opened 40 caches north of theroad, of which at least 17 randomly scattered examples contained bitsof copi^er, iron, or glass. By contrast, of the 6 pits opened south ofthe road, none yielded trade goods. It might be surmised that a slighttime difference is indicated, the older part of the site lying south ofthe road. I think such a conclusion would be premature at this stage,since we did relatively little work in the latter area, and the samplingcan hardly be considered adequate. Three areas presumed to be middens because of comparativelyheavy refuse concentration were also tested (fig. 14). One lay 40yards west of the house site ; another, about 90 yards to the southweston the west hillside, had been partly cut away by the road ; and a thirdwas 35 yards south of the road on the east slope of the ridge. About275 square feet of the north midden, No. 1, were excavated, revealingabundant refuse to 8 or 10 inches and less to circa 15 to 18 inches.Here and there pockets of ash and charcoal were noted, as also occa-sional small clusters of sherds evidently from a single vessel. Cachepit 30 lay near the center of the excavated area, and pit 31 was at thesouthwest corner but nowhere was there a recognizable vertical breakin the deposit. Below, there was only clean undisturbed subsoil, withno suggestion of house featui'es or any other premidden construction.The materials recovered were uniform throughout, and did not differsignificantly from those in the cache pits elsewhere on the site.The roadside midden, No. 3, some 12 to 15 yards across, was similarto the foregoing in all respects, except that the deposit reached asomewhat greater depth. Approximately 350 square feet of this areawas dug over. There Avas nothing in the potsherds and other artifactsrecovered, or in the observed manner of their occurrence, to suggestthat the remains belonged to another horizon than that indicated inthe house site and the cache pits opened. Both middens, throughprolonged cultivation, had been so reduced from their original heightand so widely spread that whatever subtle pottery or other changesmay have once occurred from bottom to top could no longer berecognized.The third midden, No. 2, lying south of the road, was found tohave been erroneously identified. Excavation disclosed a group offour closely spaced cache pits, one of them quite small, but all con-taining imusual quantities of broken pottery and stone. Betweenthe pits there was no refuse below plowsole; and on the uphill side,there was virtually no trash beyond the pits. We concluded, there-fore, that the supposed midden was a relatively recent development ? Wbdbl] an introduction TO KANSAS ARCHEOLOGY 139 a superficial mantle of material dragged from the upper parts of thecaches by farm machinery and further spread down the slope bywashing rains.The location of the burial gromid, if distinct from the village site,is not known. A few soft bits of human bone?chiefly ribs and verte-brae?were recognized at one point in the north midden at a depth of18 inches, but no grave outline could be traced. Beyond this we en-countered no suggestion of burials. According to local collectors,stone-covered burials once occurred at the south end of the ridge, butno details concerning these could be obtained. A search of the spotrevealed a few large stones nearby which had apparently been thrownor dragged down the slope during farming operations, but no proofwas forthcoming that these had actually been in association withburials. Moreover, even if the former presence of such graves herewere verifiable, it would still be necessary to establish, through asso-ciation with culturally indicative artifacts, their direct connection withthe village site remains. Mere proximity of culturally unidentifiedgraves to the village site is obviously no proof whatever of a directtieup.^ VEGETAL REMAINSAs was to be expected from its general nature and probable age, theFanning site yielded relatively little material of vegetal origin. Thefew specimens found owe their preservation primarily to the fact thatthey were charred and, as discarded materials, lay in refuse deposits incache pits well below plow line or other shallow subsurface dis-turbances subsequent to their deposition. Both domesticates and non-domesticates are represented in our sample.The only domesticate certainly present was corn, which occurredchiefly as scattered kernels in a number of the caches and here and therein the midden deposits. Of the beans reported in the field notes fromseveral cache pits there is no trace in the vegetal material at hand ; theyhave either been lost or else were erroneously identified in the field.From pit 11 came perhaps half a pint of shelled corn, most of it insmall lumps fused by the burning. Among these lumps was one shortfragment, about 40 mm. long, of a lO-rowed ear ; cob diameter is about20 mm., and the total diameter of cob and kernels is 37 mm. None ofthe kernels show any denting. No stalk or husk fragments were notedin the field, despite the fact that the plant was doubtless extensivelygrown by the natives.Of the wild fruits, roots, berries, and other nondomesticates doubt-less readily available to the natives here, only three species are repre-sented in our excavated material. These include black walnut (Ju-glans nigra) , found as shell fragments in pits 17, 25, and 29 ; hazelnuts{Corulus americana) , represented by scattered shell fragments in pits 140 BUREAU OF AMERICAN ETHNOLOGY [Bull. 17410 and 17, and by numerous shell fragments and half a dozen kernelsfrom pit 25 ; and papaw (Asimina triloba) , whose characteristic seedscame from pits 10, 17, and 25.Since our 1937 investigations at Fanning, additional charred vege-tal materials from the site have come to my notice. In June 1955,Fenn Ward of Highland, Kans., called at my office and informedme, among other things, that he had taken from an eroded cache atFanning a quantity of shelled corn in which were a number of "bis-cuits," both com and biscuits having been preserved by charring.On his return to Kansas, Mr. Ward sent me a series of these curiousobjects. They vary a good deal in size and shape. Some are in theform of small irregular pellets 15 to 20 mm. in diameter; others,slightly larger, superficially resemble a dried plum or prune; stillothers have a flattened circular shape, measuring 40 to 45 mm. in diam-eter by 15 to 20 mm. in thickness. The outer surfaces are more orless smoothed and dull black in color, though here and there a glossyarea may be seen. The interior varies from fairly compact and finegrained to a coarsely vesicular structure vaguely reminiscent of acharred black walnut. Under a glass, some of the freshly fracturedinternal surfaces have a somewhat "nubbly" appearance. Each of theseven whole, or nearly whole, specimens I have examined is piercedby a small cylindrical hole 2 to 3 mm. in diameter ; and from two ofthese holes I took fragments of twisted two-ply cordage.Samples of the corn and "biscuits" were submitted to Dr. Volney H.Jones, University of Michigan Museum of Anthropology. Underdate of November 15, 1956, Dr. Jones reported as follows on thismaterial (Eeport No. 372, Lots 4652-4654, of the EthnobotanicalLaboratory, Museum of Anthropology, University of Michigan) : The three lots to be described here . . . were recovered from the FanningSite, Doniphan County, Kansas. All three lots are carbonized but otherwisefairly well preserved.1) This lot consists of 15 more or less intact corn kernels and a number offragments of kernels. The kernels are of rather large size, mostly broadlycrescent shape, but some few tend toward wedge shape. The converging anglesof the sides suggest that they are from ears having from 8 to 10 rows of kernels.The dimensions are : Width: range 9 to 10.5 mm., mean 9.8 mm.Height: range 7 to 8.5 mm., mean 7.8 mm.Thickness: range 4 to 5 mm., mean 4.5 mm.These kernels fit very comfortably into the conception of the race of cornwhich has been designated the Eastern Complex. Among the characteristicswhich can be accommodated readily here are the low row number, large kernels,and broad crescent form. It can also be noted that the kernels on charringhave lost their germ areas, a trait which seems to be characteristic of theEastern Complex. The only negative feature is that some few of the kernelstend to a rectangular form which is not typical of this race. These are in theminority and may be butt or tip kernels which are atypical. Wbdel] an introduction TO KANSAS ARCHEOLOGY 141The characteristics and regional and cultural associations of the EasternComplex have been discussed by Brown and Anderson (1947), Jones (1949),and Jones and Fonner (1954). In aboriginal times it had wide distributioneast of the Rocky Mountains, persisting up to historic times in the EasternWoodlands and Prairie areas. Archaeologically it has very close associationwith the Mississippi Pattern.2) The second lot consists of a number of tubers wliich had been perforatedand strung. These vary considerably in size, from about 1.75 cm. to about 4 cm.in greatest dimension. Some of the smaller specimens are complete and globularto egg-shaped. The larger ones have been sliced and are disc shaped with theperforation in the center.These apparently represent some food product which was gathered and pre-pared for drying and preservation. In the Prairie region a number of plantsfurnished underground parts which were used for food. One which suggesteditself immediately was the "prairie turnip" or tipsin {Psoralea esculenta), thepreparation and use of which is described by Gilmore (1919, pp. 92-93 andplate 16). It should be noted that in his illustration the product is strung bybraiding the tops of the plants together rather than perforating and threadinga cord through them. Another product is the water chinquapin {Nelumiolutea), the use of which among the Osage is described by La Flesche (1928, p. 55,and plate 6). His report includes an illustration of the sliced roots strung on acord. However, both of these products can be eliminated from consideration, onthe basis of form and structure.In every respect, the tubers from the Fanning site compare well with thoseof the groundnut, Apios americana (synonyms: Apios tuberose, Qlycine apios)and an identification to this species seems warranted. Specimens from ourcollections were charred for comparison, and checked very closely in size, form,and structure. Particularly significant is that one of the entire tubers fromthe Fanning site shows points of attachment of rhizomes or root-stocks at bothends. The groundnut grows in this manner with the tubers strung along theroot-stocks like a string of beads. This accounts for the common names of "rosary roots" and the French "les racines des chapelets." This characteristicdoes not apply to the other underground food products of the Prairie regioninsofar as we are aware.Groundnuts were used for food almost wherever available throughout theEastern Woodlands and Prairie. Gilmore describes their use by the Indiansof the upper Missouri River country, and illustrates the plant and its tubers(1919, pp. 94-95, and plate 17). He also has further discussion of it in Wedel(1936, p. 60) where he reports identification of it from the Hill Site, a historicalperiod Pawnee site near Red Cloud, Nebraska. The only other archaeologicalfind of this plant known to us is from certain of the so-called Ozark Bluff-dwell-ings of Arkansas, this identification also being made by Dr. Gilmore. This latteridentification is reported in a paper by Beardsley (1940, p. 512) in which shesummarizes the data on the uses of the groundnut by Indians and presents ananalysis of its food value and manner of preparation and use. It should be re-marked that in none of the accounts of the preparation of the groundnut is itreported that it was strung as was the Fanning Site specimen. Perhaps this canbe attributed to incompleteness of the accounts or to oversight, as such stringingwould be a convenience in transporting and in storing and removing from cachepits.3) The cord on which the groundnuts were strung is about 1.5 mm. in diameterand composed of two strands twisted together in a clockwise (S-twist) direction.The twist is loose and uneven, indicating that no great amount of care was exer-cised in its manufacture. Cord of this nature is generally made by rolling the 142 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174fiber strands on the thigh under the palm of the hand. The material from whichthe cord is manufactured is coarse inner bark from some tree or shrub, but inits present condition it does not seem feasible to attempt identification to anyparticular plant. It definitely does not have the structure of rawhide or sinew.FAUNAL EEMAINSMammal bones from the Fanning site represent 13 or more species.The following, listed in order of their relative abundance, have beenidentified in the material returned to Washington : White-tailed deer {Odocoileus virginianus) 120Bison (Bison bison) 29Dog (Canis familiaris) 18Dog or wolf (Canis) 10Raccoon (Procyon lotor) 8Beaver (Castor canadensis) 10Black bear (Euarctos americanus) 7Gopher (Oeomys hursarius) 6Elk (Cervus canadensis) 5Puma (Felis concolor) 1Lynx (Lynx rufus) 1Woodchuck (Marmota monax) 1Opossum (Didelphus marsupialis virginianus) 1White-tailed jackrabbit (Lepus totcnsendii) 1Miscellaneous unidentified fragments 29With exception of the bison and jackrabbit, none of the above speciescan be regarded as typically Plains forms. The list, in fact, repre-sents a fair sampling of the animals characteristic of the hardwoodforests and tall grass prairies of northeastern Kansas ; and probablyany of them could have been easily taken along the timbered streamsand bluffs and on the prairie uplands within a short distance of thevillage site.Bird bones were much less common than those of mammals, andrepresent only two species. The humerus, a carpometacarpal, severalossified leg tendons, and other bones of the wild turkey [Meleagrisgallopavo) were taken from the cache pits; and there is also onefemur of the red-shouldered hawk {Buteo lineatus). Turkeys musthave been very plentiful in the heavy woods near the Fanning site,and it seems rather strange that no more of their remains were un-covered.The coracoid of a snapping turtle {Ohelydra serpentina) was takenfrom pit 42. It is remarkable chiefly for its great size. Comparisonwith the same bone from snapping turtle skeletons of known size inthe national collections suggests that the specimen from Fanning(USNM 381876) came from an animal whose carapace length closelyapproached 500 mm. (20 inches).The field notes make no mention of finding of bones of fish, andthere are none in the Fanning site collections returned to the NationalMuseum. WEDBL] AN INTRODUCTION TO KANSAS ARCHEOLOGY 143MOLLUSCAN BEMAINSThe shells of various fresh-water mollusks were fairly plentifulon the site, broken and weathered remnants being turned up at almostany point we tested. Whole shells, unworked and generally suitablefor identification, came mainly from the various cache pits. Manyof the specimens, when found, still retained considerable portions ofthe periostracum, or epidermis. The following species were present,listed in the order of their indicated abundance among the wholeshells returned for identification to the National Museum : Amblema costata Raf 21Lampsilis ventricosa occidens Lea 16Quadrula qnadrula Raf 9Lampsilis fallaciosa Smith qLampsilis siliquoidea Barnes 6Quadrula pustulosa prasina Conrad 2Pleurobema coccineum Conrad 2Proptera purpurata Lam 1Tritigonia verrucosa Raf 1These mollusks were presumably obtained from nearby Wolf Creekand its tributaries, which I suppose were clean sandy streams prior tobreaking up of the natural vegetative cover and consequent heavysilting of the watercourses. There is no way of determining whether,or to what extent, the mussels were used for food. Their burned orcrushed shells were certainly extensively utilized in pottery making,the local wares being almost exclusively shell tempered, as we shallpoint out presently. Possibly the need for the shells in connectionwith the native ceramic industry was the primary or sole reason forgathering the shellfish. POTTERYDespite the fact that the Fanning site has been frequently visitedand surface-hunted by collectors for upward of half a century andhas been intensively farmed for an even longer period, ceramic re-mains still occur in relative abundance both on and under the surface.During our own brief investigations, sherds were the most plentifulmaterials encountered in the cache pits, middens, and on the groundsurface. There can be little question, therefore, that the native in-habitants of the village site were well supplied with pottery of theirown manufacture. There is evidence, too, that they were in contactwith neighboring peoples who made pottery of quite different characterthan that typical of the Fanning site. We found no whole vessels, andonly two or three pieces could be restored from the sherds we un-covered. I suspect that large-scale excavation of the middens wetested and careful working over of the materials so recovered wouldbrmg to light many additional restorable vessels. 484172?59 11 144 BUREAU OF AMERICAN ETHNOLOGY [BOLL, 174 a -^^^-^- f g ---? - Figure 16.?Vessel shapes indicated by sherds from Fanning site. WBDEL] AN INTRODUCTION TO KANSAS ARCHEOLOGY 145Our pottery sample from Famiing comprises 2,026 sherds, including1,340 body pieces, 632 rim fragments, and 54 handle fragments. Sev-eral hundred vessels are probably represented. With regard to theirgeneral characteristics?paste, tempering material, surface treatment,vessel shapes, rim form, etc.?the great majority of these sherdsapparently represent a single somewhat variable ceramic tradition.There are, to be sure, some differences in the amount of aplastic added,in the thickness and quality of finish of various pieces, and in otherdetails. I doubt, however, that these variations are significant enoughto warrant the setting up of distinct types. The most obvious distinc-tion appears to be that between decorated and undecorated sherds ; andon this basis I have set up two categories. These have, I think, somevalidity over and above their descriptive convenience; and I believe,too, that they will help bring out some of the differences observablebetween the Fanning pottery and that of other related ceramic com-plexes in the region. It is doubtless true that decoration was usuallylimited to certain portions of the vessels, notably the upper body andthe lip, so that some of the plainware sherds may have come fromvessels otherwise represented by decorated fragments. On the otherhand, decorated sherds are decidedly in the minority ; and most of themappear to be from vessels that were somewhat smaller, a trifle thinnerwalled, and perhaps more carefully finished than were most of thosefrom which came the body sherds and rim sections that I have classedas plainware. Fanning Plain(Sample: Body sherds 1^50 ; rimsherds 594; handle fragments 54; total 1,798)Paste :Tempering : Shell, usually in thin flat particles 0.25-2.0 mm. in diameter,seldom over 4 or 5 mm. ; sparingly to abundantly used, shell flakes usuallyoriented in plane parallel to sherd walls ; exposed edges often show thinflat cavities where shell particles have been leached out ("hole-temper-ing").Texture: Medium to fine; usually even, well-compacted, sometimes moreor less laminated ; fractures usually clean, with little tendency to crum-ble ; very rarely, sherds show tendency to split along line parallel to theirsurfaces.Hardness: 2.5 to nearly 5, with most of the sherds apparently around 3.5 to 4.Color: Core variable from tan to dark slate gray ; surfaces variable from tanto orange-brown to dark gray or nearly black; surfaces of large sherdsoften show irregular firing clouds.Method of manufacture: Undetermined; no evidence of coiling in the ma-terial at hand.Surface finish : exterior surfaces usually well smoothed, sometimes almostattaining a polish; commonly with horizontal striae left by the smoothingimplement ; interior surfaces uneven and less carefully smoothed.Form :Vessel shapes: Medium to large full-bodied or globular jars, with roundedbase, constricted neck, recurved rim, and often?perhaps typically?with 146 BUREAU OF AMERICAN ETHNOLOGY [BOLL. 174Strap handles, are apparently the common form (fig. 16; pi. 8, 6). Someevidence of deep bowls (fig. 16) . Rim: Outcurving or straight outflaring, rarely vertical ; from 2-4 cm.high ; unthickened and undecorated (fig. 17) . Lip: Usually rounded and somewhat thinned, but occasionally nearly flat.Base: No direct evidence, but presumably rounded.Neck: Constricted.Vessel size: Jars apparently range in diameter from 10-12 cm. upward toapproximately 30-35 cm. (estimated from sherds) ; the relative abundanceof heavy rim, handle, and body sherds suggests that there were a goodmany large plainware utility vessels.Thickness: Body sherds average between 4 and 7 mm.; occasional pieces,especially from necks of large jars, may reach thickness of 12 mm.Handles : Strap handles are highly characteristic, loop handles and lugs muchless so ; probably two and sometimes four per vessel. The ends are blendedinto the vessel fabric, the upper end usually 5-10 mm. below the lip and thelower end on the upperbody below the neck. There is no evidence of rivetedor tenoned attachments, and in no case does the handle rise directly from thelip. Two forms of strap handles may be recognized: (a) Tapered straps (57specimens), narrowest at the lower end and widening as they rise to the rim.In size, the smallest widen from 15 to 25 mm., the largest from 3-5.5 cm. ; they are from 2 to 5 times as wide as they are thick, and have a flattened e^m^ f ^'^ gFigure 17.?Rim profiles, Fanning site. Wedel] AN INTRODUCTION TO KAJSTSAS ARCHEOLOGY 147 cross section. (&) Straight-sided straps (34 specimens), in which width isfairly uniform except for the slight widening at points of attachment. Theserange in width from 15-40 mm. (fig. 18) . Loop handles, with circular or flattened elliptical cross section, occur twice.There are two lug handles?heavy protuberances, wider than they are long,and extending out from the body (?) or rim, imperforate, but with 2 or 3deep vertical notches or indentations at their outer end. Thin horizontal tabswhich are flrittened extensions outward from the vessel lip also occur very Figure 18.?Straight-sided (a-c, /) and tapered (d, e, g) strap handles. Fanning site. 148 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174 sparingly, always imperforate and undecorated ; they were probably confinedto open bowls.Decoration : Confined entirely to the lips and handles. Approximately 70 per-cent of the strap handles have broad shallow flutings apparently made bydrawing the fingers vertically down the surface ; 20 percent have narrowerscored, incised, or trailed lines made with a pointed instrument. Lip deco-ration consists of punctate or incised elements (20 percent), closely spacedalong the lip surface and seldom very carefully executed; or of shallow im-pressions apparently made with the finger tip, occasionally with the nail markalso present (37 percent) ; or the lip was left undecorated (38 percent).Fanning Trailed(Sample: Body sherds 185; rimsherds 29; total 214)Paste: See Fanning Plain.StJBFACE FINISH : See Fanning Plain.FoBM : See Fanning Plain. Most of the sherds bearing trailed or inciseddecoration suggest smaller vessels than do the plain sherds ; one or two largevessels are also indicated, but even these seem to have been smaller and oflighter construction than a good many of the larger plainware vessels.Handles: Present, including tapered strap (6), straight-sided strap (1), andloop handles (5) ; loop handles occur mainly on small vessels. ^ ////// Figure 19.?Trailed potsherds from Fanning site. Wbdbl] AN INTRODUCTION TO KANSAS ARCHEOLOGY 149Decoeation: Body decoration consists of broad (3-5 mm.) to narrow (0.5-1.5mm.) trailed lines, usually running vertically to obliquely from the neck down-ward over the rounding shoulder (figs. 19, 20). In some cases, the lines are so g \4^ ...^ ^Figure 20.?Trailed and punctate potsherds, Fanning site. 150 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174 narrow and deep as to suggest incising, and tliey may start with a deep sharpindentation ; others are blunt ended, and some are faintly applied with littlepressure. In some cases certainly, and perhaps characteristically, the linesoccur in blocks of varying sizes and shapes, with the lines in adjacent blocksslanting in different directions (pis. 9, 10) . The lines are nearly always closelyspaced at intervals of 3-7 mm., but occasionally are as much as 20-30 mm.apart. In 26 sherds, short punctates, rounded to elongate, occur in a row be-tween contiguous trailed lines (pis. 9, 10). Chevron arrangements also occur(pi. 9, 6). In three instances, including the single restored vessel (pi. 8, b),blocks of punctates occur immediately below handles, and these blocks areflanked by trailed line areas. Curvilinear lines are present in two instances,but there are no circles. Lip decoration consists of finger (?) impressed units(21 percent) or of small punctations or diagonals (41 percent) ; there is con-siderable variation in the execution, but the best and most carefully done lipornamentation from the site occurs on these rims. About 25 percent of thelips are undecorated. Handle ornamentation consists of vertical trailed lines(7 specimens), or of several rows of circular punctates or short indentations(4) ; only one specimen has the fluting characteristic of plainware handles.Where four handles occur on the same vessel, as on USNM 381751 (pi. 8, 6),scored and punctate decoration alternates, with scored handles opposite eachother and punctate handles opposite each other.Miniature Vessels(2 specimens)Both of these are incomplete. One is a fragment of small jar, full-bodied and with diameter apparently exceeding height, and having aconstricted neck, a low recurved rim, and loop handles. The lip isnotched, apparently with the finger tip ; and body decoration consistsof shallow trailed lines and round punctates. Measurements as recon-structed by projection of various curves were approximately 60 mm. inheight by 77 mm. in maximum diameter. In practically all respects,this little vessel conforms to what has been described herein as FanningTrailed ware (see fig. 20, 5, and pi. 10, h).The second piece is a shallow open bowl, slightly longer than wide,and with dimensions of 52 by 45 by 18 mm. At the ends and sides, therim has been prolonged slightly into horizontal flat tabs, 4 in number.Below each of the end tabs there is a small perforation. That similarbowls of larger size were probably also in use here is suggested by abowl rim fragment with a flat lip tab (pi. 10, A, and fig. 21, d, e).Other SherdsFourteen sherds, including 9 rim fragments and 5 body sherds,differ markedly in all respects from the foregoing materials. All aregrit-tempered, in marked contrast to the locally made shell-temperedpottery ; and the body pieces all bear simple stamping, done either witha scored paddle or with a thong-wrapped implement. The rimsinclude 6 specimens with a thickened or braced rim (pi. 11, a-f) . Fourhave parallel incised lines running horizontally along the rim exterior; Wedbl] AN ENTRODUCTION TO KANSAS ARCHEOLOGY 151 Figure 21.?Perforated disks {a-c) from reworked sherds and miniature pottery bowl,Fanning site.one has parallel lines apparently crossed at intervals by paired slanteddiagonals ; and the sixth has herringbone indentations instead of thehorizontal lines. Three, and possibly four, specimens are from vesselswith multiple handles, and it may be presumed that these were fromspecimens with cloistered rims. The lips all have either diagonal in-cisions or punctates, the incisions being either on the top of the lip orelse running downward on the inner upper margin of the rim. Thethree uncoUared rims also show the diagonal incisions on the innerupper rim (fig. 22, ^) . Eight of the nine rimsherds conform in paste, tempering, shape,and decorative treatment to pottery characteristic of Lower Loupsites in east-central Nebraska ; and those with multiple handle scarsare immediately recognizable to anyone with experience with thewares of that area and tradition (cf. pi. 11 with Dunlevy, 1936, pis.2c, 3c and 3e, 6-8; Wedel, 1936, pi. 6; Wedel, 1938 c, pis. 4 and 5).Closely similar pieces occur in collections at the Nebraska StateHistorical Society from a number of sites in the Lower Loup andPlatte valleys, including 25BU2 (Bellwood I), 25BU4 (Barcal),25PT1 (Larsen), 25NC1 (Burkett), 25NC3 (Wright), and 25CX1 152 BUKEAU OF AMERICAN ETHNOLOGY [BULL. 174 4 Figure 22.-GrIt-tempered rim and handle sherds suggesting Lower Loup Focus connec-tions, from Fanning site. Wedbl] an introduction TO KANSAS ARCHEOLOGY 153(Gray). The Fanning pieces are compact, hard, and well made ? better, at any rate, than those from Lower Loup sites showing con-siderable White trade material, and about as well made as those fromthe earlier sites of the complex. We shall return to this point again.Two other rimsherds, both shell-tempered, suggest a blending ofLower Loup practices with the local pottery tradition. Both showa decided thickening of the rim, and have deep diagonal incisions onthe inner surface below the lip ; on the outer rim below the lip, one hasherringbone punctations, the other a single row of diagonal punctates.Except for the shell tempering, these pieces might well be consideredLower Loup trade items (pi. 11, ^, h).PEEFORATED POTTEBY DISKSOther than pottery vessels, the only ceramic products indicated inour work at Fanning consisted of perforated disks. There are 40of these, all made from reworked potsherds ; all but one were fashionedfrom plainware sherds. They are present in all stages?some arefairly well shaped, others are perforated but not otherwise worked,and a few show the first attempt at drilling. They seem to have beenmade by chipping sherds into a roughly circular form, perforatingthem centrally, and then grinding down the edges to produce asmoother outline. They range in diameter from 25-55 mm. ; the per-foration is always biconic, but varies from a well-smoothed hole to avery irregular one. In several instances, these disks occurred in somenumbers in single cache pits. Thus, 5 were taken from pit 37, 4 frompit 3, and 2 each from pits 2, 20, 25, and 28. From pit 36 camie 7sherds, each with the central perforation, but otherwise without modi-fication ; none was circular, nor was there any evidence of an attemptto shape them. In one or two of these, the perforation lies near theedge of the sherd, and thus there is a possibility that these are rem-nants of broken jars repaired by crack lacing, evidence of which alsooccurred in several other plainware body sherds; but the fact thatthere were 7 such sherds together in one cache, and that there wereother instances of clusters of perforated disks, suggests that the pit36 materials were intended for the manufacture of disks also. Thepurpose for which these objects were intended is unknown (fig. 21,a-c; pi. 12, o-q) . WOBK IN CHIPPED STONEThe great majority of the chipped-stone artifacts from Fanningare made of a blue-gray chert, much of which has a characteristicbanded appearance. Many of the objects, particularly among thenumerous end scrapers, include on their nonworking surfaces more orless hard white calcareous material which was apparently the outsideof nodular cherty concretions. I suspect that this material was ob- 154 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174tained from some of the chert-bearing formations in the Flint Hillslimestone beds of the lower Blue Kiver drainage. There are alsoa few pieces of reddish-brown foraminiferous chert that may beFlorence flint ; but the present specimens lack the very characteristicbanding found in the Florence flint artifacts collected at sites inthe Arkansas Kiver drainage of central and southern Kansas. Noobsidian was found.It appears likely that the blue-gray chert mentioned above was car-ried to tJie site in partially worked condition. Numerous unre-touched spalls and flakes up to 10 or 12 cm. in length were found;and in pit 42 there was a compact mass of 63 such flakes, presumablystored away for future manufacture of implements. No nodulesfrom which such flakes might have been struck came to light anywherein our workings.Projectile points.?Chipped points are represented by 41 whole and1 fragmentary specimens. Of the former, 36 are of simple triangularunnotched shape, widest at the base, which is usually straight or veryslightly concave, and with straight to slightly convex lateral edges;all save one or two are bifacially flaked. Workmanship and qualityof flaking vary considerably. The size range is as follows: length13-37 mm., mean 23.3 mm. ; width 12-16 mm., mean 14.3 mm. ; thick-ness 2-3.8 mm., mean 3.3 mm.; weight 0.5-1.0 gr., mean 0.85 gr.(pi. 12, Z, m).Three other points are roughly triangular but have curved edges,with their maximum width somewhat above the base. They arelarger and of much cruder manufacture than the points in the fore-going group. Two were made from thin flakes and are retouchedmainly along the lateral edges; the third is bifacially flaked, but re-tains the marked curvature and asymmetry of the original flake. Theyrange in length from 34-38 mm., in width from 15-22 mm., and theirmean weight is 2.75 gr.The two remaining points are of altogether different type, and wereprobably not a part of the site complex. Both are thick stemmedand coarsely flaked, strongly reminiscent of the large corner-notchedand other heavy stemmed Woodland point types found at Woodlandand Hopewellian sites in this general region. The larger is corner-notched, measures 50 by 32 by 7.5 mm., and weighs 11.5 gr.; it wasfound in pit 10. The smaller, a surface find, is notched above thebase and has an expanding stem ; it measures 40 by 27 by 7.5 mm. andweighs 9 gr. I believe that both points are from a much older periodof occupancy tlian that represented by the small triangular pointswhich decidedly predominate on the site.Bipointed ohjects.?There are four of these from the surface andtwo from the excavations in midden 1. They are fusiform, widest Wbdbl] an ENTRODUCTION TO KANSAS ARCHEOLOGY 155and thickest at the middle, and are bifacially chipped to a rather bluntpoint at each end (pi. 12, k). They range in length from 17^6mm., in width from 7-10.5 mm.; the width to length ratio variesfrom 1 : 2.5 to 1 : 4. Their bifacial chipping distinguishes them froma group of somewhat similar objects found at the Leary site and de-scribed (Wedel, 1936, p. 49) as gravers. To what use these objectswere put I cannot say ; they may have been fish gorges (see fig. 24, /) . Drill points.?There are three of these, all from excavations. Theyare long, narrow, moderately thick objects, bifacially chipped to apoint at one end. Length varies from 29-47 ram., width from 9-15mm. In two the working tip is long and evenly tapered, the other endbeing either broken or else much less carefully flaked. The third has ashort stubby point and a rounded dressed butt. The tip and adjacentlateral edges of the longest specimen appear to be somewhat bluntedfrom use.Knives.?Bifacially chipped objects with cutting edges to whichthe term knife may be reasonably applied are surprisingly uncommonin our collection from Fanning. There are, in fact, only four suchspecimens, all incomplete. Two are oblong with the long edges moreor less parallel, and each has one end missing ; they measure 60 by 41by 11 mm. and 48 by 25 by 9 mm. The edges are irregular and uneven,and have only a moderately good retouch. One lateral edge of thelarger appears to be somewhat blunted from wear. Another piece isevidently the tip of a pointed blade, showing two oppositely beveledlateral edges ; whether this is from a four-edged diamond-shaped knifeof the so-called Harahey type or from some implement with but twooppositely beveled edges and a rounded notched butt such as occurcommonly in the Little Eiver Focus sites in Rice County (p. 270) I amunable to say. The fourth specimen from Fanning is a roughly tri-angular thick flake with one curved, uneven, and coarsely chippedcutting edge ; its size is 62 by 40 by 13 mm.EticL scrapers.?These ubiquitous little implements were the mostnumerous of all the artifacts found at Fanning, with exception of thepotsherds. They are a rather variable lot, but all share the unre-touched plane under surface, a more or less steeply chipped broadworking end, and a convex or ridged upper surface. Unlike the endscrapers characteristic of the Little River Focus sites in Rice Comity,however, the Fanning scrapers are seldom very well made and veryfew show much or careful retouching of the convex surface. They looklike strictly utilitarian tools made to a well-established basic pattern,but without much effort devoted to fashioning fine implements (pi.12, h-j).The 129 complete end scrapers found may be divided into fivegroups: 156 BUREAU OF AMERICAN ETHNOLOGY [B0LL. 174Group I (38 specimens).?These are fashioned from spalls of vary-ing size and shape, with the convex surface generally retaining muchof the limestone matrix that originally enclosed the chert. All are ofbanded gray or blue-gray stone. The shape is largely accidental, theconvex surface consisting usually of two or three longitudinal flakescars with little or no further modification of the lateral edges. Thebroader end is generally chipped to a rounded working edge. Nearlyall specimens are asymmetrical, the highest part of the ridge beingto one side or the other of the longitudinal midline of the specimen.Dimensions are: length 32-66 mm., mean 45 mm.; width 15-37 mm.,mean 24.6 mm. ; thickness 8-12 mm., mean 10.9 mm (fig. 23, a).Group II (16 specimens).?These are very similar to the foregoing,except that they have fine retouching on one or both lateral edges andthus tend to have an ovate form. They are also somewhat shorter,narrower, and thinner. Dimensions: length 27-43 mm., mean 33.8mm.; width 16-32 mm., mean 22.2 mm.; thickness 4-19 mm., mean8.1 mm. (fig. 23, 6)Group III (6 specimens).?These are strongly ridged or keeled,with a single longitudinal flake scar on each side of the ridge. Theyare triangular in cross section?in effect, prismatic flakes with retouch-ing only at the broad end. Dimensions : length 33-66 mm., mean 44.3mm. ; width 18-30 mm., mean 21.5 mm. ; thickness 8-9 mm., mean 8.3mm. (fig. 23, c). Figure 23.?Top and side views of chipped end scrapers, Fanning site. Wkdbl] an introduction TO KANSAS ARCHEOLOGY 157Group IV (61 specimens).?Scrapers of this, the largest, group arerelatively broad and low-backed, with the convex surface consistingof more than two flake scars and the lateral edges as well as the broadend retouched. Dimensions : length 23-60 mm., mean 36 mm. ; width15-35 mm., mean 24 mm.; thickness 4-14 mm., mean 7.7 mm.(fig. 23, 6?). _Group V (8 specimens).?These have a fairly prominent convexsurface, usually rounded by the allover flaking by which the final formwas achieved. In cross section, they lack the angular profile thatcharacterizes most of the preceding scrapers ; and they show generallymore modification of the original flake than do any of the others.Dimensions: length 26-65 mm., mean 39.2 mm.; width 19-33 mm.,mean 26 mm.; thickness 6-16 mm., mean 11.1 mm. (fig. 23, e).Side scrapers.?These are unifacially chipped spalls of irregularand largely accidental form, many of which retain a thick whitelayer on one surface. Nearly all are made from Flint Hills (?) chert,generally the same material that was used in groups I and II of theend scrapers. They differ from the end scrapers in that the final re-touch is all on one or both lateral edges, never at the end. The qualityof retouch varies greatly ; sometimes it is carefully done and providesan even curved cutting edge, whereas on other pieces it is irregular,coarse, and haphazard. All these specimens are planoconvex in crosssection.Two groups may be distinguished, viz, single edged and two edged.The former, represented by 28 specimens, includes spalls with littleor no modification other than the edge retouching. The size rangeis as follows: length 35-68 mm., mean 53 mm.; width 16-52 mm.,mean 28 mm. The second group consists of larger and somewhat morecarefully finished implements, some with a strongly ridged back, andall with retouching on two lateral edges. The 15 specimens have thefollowing size range: length 48-97 mm., mean 61.6 mm.; width 20-47 mm., mean 30 mm. (fig. 24, a-e) .Flake knives.?Distinguishable from the foregoing by their smallerand more slender character are seven prismatic flakes with very finelyretouched edges. The edges are usually retouched or chipped from theplane surface as well as from the slightly ridged back. Size range : length 45-68 mm., mean 52.4 mm. ; width 14-21 mm., mean 17.1 mm. ; thickness 5-7 mm., mean 5.7 mm.WOBK IN GEOTJND AND PECKED STONEMaterials used in manufacture of implements considered in thissection include sandstone, catlinite, pumice, hematite, diorite, quartz-ite, and granite. The first of these was presumably obtained fromthe Dakota sandstone outcrops beyond the Blue River, 100 miles or 158 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174 Figure 24.?Side scrapers and bipointed object, Fanning site.more to the west. Pumice occurs as irregular slightly worked lumpsthat may have been gathered along the shore of the IVIissouri River, asit was by the historic Indians of the valley. The nearest naturaloccurrence of catlinite was the famous quarry in southwestern Minne-sota, whence the stone may have reached the Fanning locality byintertribal trade. Hematite was obtainable in Missouri and southernIowa, but the exact locality from which originated the pieces found atFanning is uncertain. Diorite, Sioux quartzite, and granite wereobtainable as boulders in the glacial drift of northeastern Kansas, andprobably the materials of which our artifacts were made were takenat no great distance from the Fanning locality. Wedel] AN INTRODUCTION TO KANSAS ARCHEOLOGY 159Abraders.?There are several kinds of abrading tools in our collec-tions. Most common are six worked fragments of oblong shape, eachwith one or two flat surfaces bearing a shallow longitudinal groovesuch as would be produced in the smoothing of arrowshafts. Thelargest piece is 65 by 37 by 25 mm., and is broken at both ends. Thispiece, and two much smaller fragments, suggest some care in shaping,and I suspect they are from paired shaft-smoothers such as were foundin both broken and unfinished state at the Leary site (Hill and Wedel,1936, p. 45 and pi. 7, a-c) and which are also known from manyother sites in the Central Plains. Other fragments at Fanning areperhaps from broken pieces modified by re-use until their originalform is obscured.Four oblong flattish pieces have one to several short deep narrowgrooves on one or more surfaces or else bear short broad grindingfacets. The narrow grooves suggest awl sharpening; the broaderfacets could have been produced in rounding off the ends of sticks, boneimplements, or other materials, but not in finishing arrowshafts orsimilar long objects.Three pieces of sandstone, including a shaped irregular block withdressed edges and measuring 87 by 55 by 18 mm., have smoothed sur-faces that suggest their use as whetstones or rubbing tools. Two otherblocks of very fine-grained red sandstone appear to have been roughlyshaped ; their surfaces bear smoothed areas, and one has also numerousnarrow grooves not over 1.5 mm. wide and of equal depth (fig. 25).These two specimens measure 11 by 10 by 3 cm. and 12.5 by 6.8 by 3cm. Still another smoothed block, of light-gray limestone and meas-uring 95 by 80 by 16 mm., has a single straight groove on one face.Five lumps of pumice vary in size and shape, and were apparentlynot dressed to form. All show smoothed or worn facets that mighthave been produced from hide dressing or by rubbing of other softmaterials. They range in size from 40 by 25 by 15 mm. up to 125 by105 by 65 mm. Figure 25.?Sandstone sharpening block, Fanning site. 484172?59- -12 160 BUREAU OF AJVIERICAN ETHNOLOGY [Bull. 174Celt.?From the surface of the site came the only celt found. It iselliptical in cross section, widest at the blade edge, and tapers evenlytoward the butt, which is broken. The surface is well smoothed, es-pecially near the blade, but there are numerous pittings and irregulari-ties on much of the surface. It measures 60 by 42 by 19 mm., and isof diorite.Pitted stones.?There are two of these, each with a single pit, andboth are of limestone. One appears to be the half of a nearly circu-lar, thick block, with edges battered to shape ; one face is unfinished,the other is smoothed and bears a small pit, bisected by the break. Thepit is 20 mm. in diameter by 10 mm. deep ; the block measures 98 by 64by 45 mm. The second specimen is roughly quadrilateral, with theends and lateral edges pecked to shape. One side is unfinished; theother is smoothed and has a shallow pit, 24 by 4 mm. The block meas-ures 122 by 85 by 56 mm.Rubbing stones.?Two stream-worn pebbles bear the marks of bat-tering on their peripheries, and each has one smoothed face. The lat-ter suggests that the specimens were used as rubbing or smoothingstones. Whether the pecked peripheries indicate some effort at shap-ing the objects, or are the result of use as pecking or hammerstones, Icannot say. The stones measure 88 by 78 by 35 mm. and 67 by 56 by37 mm.Sinew stone ( ? ) .?This is an irregularly shaped fragment from aflattened waterwom quartzite cobble. All surfaces exhibit somesmoothing. Where the fractured edge meets the natural surface, thereare four shallow grooves or notches, each 4r-5 mm. wide, where somesoft thin material or narrow object appears to have been drawn re-peatedly across the edge. On another edge there appears a similarsingle notch, plus faint suggestions of several others. The piece meas-ures 83 by 43 by 35 mm. (fig. 26) . Milling stones.?Flat slabs of tough stone used for grinding werefound in three pits; and elsewhere in the excavations and on thesurface were recovered handstones, or muUers, probably used withi Figure 26.?Notched quartzite "sinew" stone, Fanning site. Wedel] an introduction TO KANSAS ARCHEOLOGY 161 such slabs. The largest slab, found in pit 43, is of granite ; it measures63.5 by 32.5 by 10 cm., and most of the perimeter shows evidence ofdressing by pecking or battering. The working surface has a shallow,smoothed, elongate depression, evidently produced by a more or lessrotary motion with the handstone. The two smaller slabs are ofpink Sioux quartzite, and measure 23 by 12.7 by 5 cm. and 27 by 15by 3.8 cm. ; they were taken from pits 8 and 22, respectively. Bothshow evidence of shaping around the edges, and one surface of eachis worn smooth, but not in the same maimer as the large slab, I aminclined to think these also served in grinding, with the muller usedin a back and forth motion over the entire surface of the slab. Al-ternatively, they might have been "anvil stones" used with a hammerfor the crushing of cherries, berries, and other food substances; butI doubt that such usage would produce the sort of worn surfaceobserved.Three mullers, one of which is broken, were made of quartzite.Two appear to have been worn smooth from rubbing, the other isunsmoothed. The two complete specimens measure 16 by 10 by 5.2cm. and 13 by 8 by 3 cm.; the third seems to have been originallya thicker specimen of substantially the same size and form. The com-plete specimens are from pit 34 and house test 1 ; the fragment wasa surface find.Dinsmore (1912, p. 54) lists "5 mano stones" among the surfacecollections he made on the Fanning site.Pounding stones.?This term includes two specimens which givesome indication of having been hafted and used for pounding orchopping. One is an irregular pillow-shaped block of limestone, 12by 8 by 5 cm., with slightly constricted middle. At one place in thisconstricted zone is a short 15 mm. groove, polished from wear thatmight have come from the rubbing of a haft. There is no wear ofthis sort elsewhere on the piece, which is otherwise unfinished andunsmoothed, but there is some evidence of battering on the blunt endsurfaces.The second specimen is a quadrilateral slab of diorite, 12 by 8 by2.5 cm., which has very shallow notches midway of each long edge.The faces are unfinished; one end is sharpened somewhat, the otheris battered to a blunt surface, as from pounding. It suggests a crudechopping tool, probably hafted and used perhaps for the crushingof bones.No trace of grooved mauls was found during our investigations,but there are reports of surface finds that suggest (Dinsmore, 1912,p. 53) that hammers of more definitive sort than the objects notedabove may have belonged to the site complex. Dinsmore, for example,observes (Dinsmore, 1912, p. 53) that "I found one maul, weight 8i/4 162 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174lbs. of red granite; one small beautiful round mallet, weight 8 oz."Since these were surface finds, it cannot be positively asserted thatthey were used by the Indians who dug the numerous cache pits andmade the shell-tempered pottery. I have not seen Dinsmore's col-lection and can give no further details regarding the objects; butthe larger suggests a grooved maul, the smaller possibly a war-clabhead.Catlinite objects.?There are 7 pieces of worked catlinite, but nofinished or complete artifacts. Three pieces are pipe fragments.Two are from the juncture of bowl and stem, but give no informa-tion as to the size and form of the original pipes. A third piece isapparently the end of a pipe stem ; it has a small comb on the uppersurface, which was perforated but has been mutilated by cutting.The exposed stem bore has a diameter of 10 to 11 mm., and showshorizontal grinding striae. None of these three pieces exceeds 25mm. in maximum dimension.A quadrilateral block measuring 63 by 63 by 14 mm. has most ofone face ground smooth, as are the edges; the reverse is unworked.On the worked face, deep narrow grooves mark out what was evi-dently intended to be an elbow pipe blank. I am inclined to thinkthat the block must have been accidently split after its initial dressingand marking but before the roughed out pipe could be detached, andso was discarded.The remaining three pieces are presumably rejectage. Two showevidence of deep sawing at the edges, where they were finally detachedby snapping them off; and there is also some smoothing of the sur-faces. The third piece has three adjoining ground edges, and onebroken and unsmoothed surface. I am unable to determine whetherany of the cutting shown was done with metal tools, but am inclinedto doubt this.Dinsmore (1912, p. 54) reports "one small pipe" from the surface,but gives no further details concerning it.Hematite.?Seventeen pieces of worked hematite came from var-ious locations in the site. All are hard massive material giving a darkred-rown streak; they vary from 18 to 306 gm. in weight. Allshow rubbing facets, and on some the present surface appears to bealmost entirely produced by prolonged rubbing or grinding. Nonehas been intentionally shaped, although one has cutting grooves ap-parently designed to remove one comer. Undoubtedly all were usedas sources of pigment. WORK IN BONE AND ANTLEBHoes.?Scapula hoes are represented by 10 specimens; 7 of theseretain most of the distal extremity, or head, 1 is split through thehead, and 2 are large slivers off the anterior and posterior borders Wedel] an rNTRODUCTION TO KANSAS ARCHEOLOGY 163 of the bone. They have been modified by partial to complete removalof the scapular spine and of the ridge at the posterior border, and bysharpening of the vertebral border. The head is either unmodified orelse shows slight notching and working down of the edges of the glenoidcavity. On most, there is a high polish from use and the workingend of the blade has been worn back to a deep broad V-shaped notchwith smoothed edges. No effort seems to have been made to smoothdown the scars left in removal of the spine and ridge, beyond elimi-nating the jagged edges. In one case, the spine appears to have beenhacked away by repeated blows from a steel tool, perhaps an ax ormore probably a heavy knife. In another, numerous fine lightscratches and one or two deep narrow cuts run along each side of thespine, suggesting knife cuts intended to weaken the bone for removalof the spine by blows from the side.Specimens sufficiently complete to indicate the length range from29 to 39 cm. Six were fashioned from the left scapula, four fromthe right. All are bison.Knife or scraper.?This is a segment of bison scapula with oneedge thinned for cutting or scraping. The opposite edge is muchheavier and consists of the worked down scar left in removal of thescapular spine ; and this extends 30 mm. beyond the end of the blade.The surfaces are much eroded, but traces of wear are still visible.Including the projecting back, the specimen measures 145 mm. inlength and it has a width of 73 mm. It may have been used in cut-ting or shredding soft materials, such as squash, or in scraping.Shaft straighteners.?Four segments of worked and well-smoothedbison rib were evidently shaft straighteners. The longest measures55 mm. and is cut off square at one end, very likely with a metal saw.The other end is broken through a hole 8 to 9 mm. in diameter, slightlyelongate from wear in a direction parallel to the long axis of the bone.Seven deep narrow notches have been cut into one edge. Anotherpiece, 45 mm. long, is broken at both ends through holes that also showtheir heaviest wear in a longitudinal axis. It has 11 fine notches alongone edge, and the concave face of the bone bears numerous fine striae.A third piece is 80 mm. long, broken and slightly smoothed at oneend ; at the other, the break is through a heavily worn hole (pi. 12, e) . The fourth specimen is apparently unfinished ; it is an 18-cm. segmentcut off square at both ends. About midway of its length a conical pithas been drilled into the concave face, which is otherwise heavilystriated as if from scraping.Knife handle.?A rib segment 12.3 cm. long is broken at one end andcut square at the other. Transversely across each face, 7 to 8 cm. fromthe broken end, is an incised line ; between these lines and the brokenend, one edge of the rib has been slotted and the cancellous tissueremoved to a depth of 9 mm. The slot is approximately 1.5 mm. 164 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174 wide, and so was presumably intended to accommodate a metal blade.The opposite edge has 15 shallow narrow notches ; and the piece is gen-erally well polished from use.Worked rib fragment.?This piece, 47 mm. long, may have beenpart of a shaft straightener or knife handle. One end is very raggedlybroken off, the other is cut rounding. Surfaces are well polished.Near one end, on the internal face, are three lightly incised transverselines. From each end of this group, 3 similar lines run obliquelyacross the surface to converge or cross 25 nmi. from the broken end.Beyond this point, the surface of the bone is broken away, and thelines cannot be traced further.Needle fragments.?Here I include 4 specimens, 3 of them brokenand 1 unfinished. Despite their incomplete nature, they seem to meto suggest the presence of implements perhaps not unlike the largemat-weaving needles described by Skinner (1921, p. 245 and fig. 17)for the Menomini. All were made from strips of split mammal rib,probably bison, dressed down so that only traces of the cancelloustissue are discernible on the convex surface of the needle, which wasalways curved.Our longest specimen (USNM 381883) measures 193 mm. and is 9mm. in maximum width ; thickness nowhere exceeds 2 mm. It showsa long even taper to a well-finished heavily polished tip. The butt isbroken off, so that neither the original length nor the presence orabsence of an eye can be determined. A second piece (USNM 381814) , broken at both ends, is 143 by 85 by 3.5 mm. ; it shows a very slighttaper toward one end and at the other has been broken off through aneye 2 mm. in diameter. Still another specimen measures 73 by 9 by 2.5mm. and is also broken at both ends. One break is through a 3-mm.hole, whose remaining portion is heavily worn ; a second hole has beencrudely drilled 10 mm. from the break, but shows no wear (pi. 12, d).Surfaces are well polished from use.T\niat I take to be a needle blank is a 210-mm. sliver split from alarge rib. One edge is formed by the natural rib border ; the other isuneven, and follows a deep cut made after the rib was split. Thewidth is irregular, but nowhere exceeds 11 mm., and the cancellous tis-sue has not been worked down nor is the tip finished. The object seemstoo light and thin for an awl or punch ; but its proportions and sizewould yield a finished tool not much different from the needles in-ferred from the fragments noted above.Awls.?Awls are very poorly represented in our collections. Thesingle complete example is split from mammal leg bone, probably deermetapodial. The head is partly worked down, and is from the proxi-mal extremity. The specimen has a long even taper, and is highlypolished (pi. 12, c). Its length is 13.5 cm. A second specimen is the Wbdbl] an introduction TO KANSAS ARCHEOLOGY 165rounded butt of what appears to be a "rib-edge" awl; it is wellsmoothed, slightly flattened in cross section, and shows very fainttraces of cancellous tissue on the convex surface. Length is 67 mm.The distal end of a deer metatarsal, 50 mm. long, has been deeplygrooved transversely across the shaft and snapped off raggedly. Theanterior and posterior surfaces above the condyles bear deep longitu-dinal scorings, and there are ends of at least four other such scorings.These marks suggest that this is the discarded end of a bone groovedfor splitting preparatory to the manufacture of awls.Tubes.?There are two of these, one broken and one apparently un-finished. The first (USNM 381708) is a 65-mm. length of bird bone,split, and broken at one end. The other end has been cut neatlyand smoothed off. The entire piece is highly polished. The secondspecimen is of thick-walled mammal leg bone, and measures 37 by 14mm. The ends have been irregularly cut off, but are unsmoothed.Worked penis hone.?The well-polished os phallus of a raccoon hashad the anterior extremity cut off but is otherwise unmodified and hasno perforation. Its length is 98 mm.Projectile points of deer antler.?Five of these interesting objectscame to light in our excavations. Each is the neatly dressed tip of adeer antler, carefully fashioned by whittling and grinding into asymmetrical conical form, with polished point and socketed base.Basal diameters range from 10 to 13 mm., with sockets up to 17 mm.deep. Length varies from 48 to 63 mm. ; and the longest specimen has,additionally, a thin basal tang 7 mm. wide by 19 mm. long (pi. 12, g).Four of these points were taken from pit 1, the fifth from pit 42.From pit 37 came a 59-mm. section of antler lacking the extremetip. A narrow groove encircles the piece 30 mm. below the broken tipend ; below this groove, the antler has been split or whittled off at oneside so that the cancellous tissue is exposed. Apparently, this repre-sents an attempt to obtain a projectile point blank with lateral tang.There is no polish and the piece was never finished.There is also a short thick tapered section 33 mm. long, cut fromnear the end of a deer tine, with battered tip. It may have been thebeginning of a projectile point shorter and heavier than any of thefinished specimens. Two antler fragments, 9 and 11 cm. long, eachwith clean-cut distal extremity, probably are scrap from manufactureof points.Hammer of antler.?This was fashioned from the basal portion of alarge deer antler, with the brow tine attached. The tine forms thehandle; the remaining portion of the main shaft, with the "burr"battered or chopped away, is the striking head. The specimen, 20.5cm. in length, shows little use polish, but is of a very convenient sizeand shape for use as a hammer or small mallet. It turned up in therefuse bone from the site, and its exact provenience is imcertain. 166 BUREAU OF AAIERICAN ETHNOLOGY [Bull. 174METAL AND GLASS OBJECTSIron, brass, and glass objects of European origin were found at anumber of points in our excavations. They were not abundant, andthey included none of the larger items, such as hoes, gun parts, traps,and axes, that would be expected if the natives had been in direct andregular contact with an established trading system. That the Indiansat the Fanning site had relationships with white men seems a reason-able conclusion ; but I believe it is equally reasonable to regard theserelationships as more tenuous and infrequent than would be expectedto be the case after the establisliment of trading posts above the KansasRiver on the Missouri in the mid-18th century.Iron is represented by 13 pieces, some very small and all very heavilyoxidized; and this, plus their generally fragmentary nature, makesidentification of the original articles impossible in most cases. Fromhouse test 2 came a short knife blade (USNM 381837) measuring ap-proximately 55 by 19 mm., with an 11 by 17 mm. tang for insertion inthe handle. Another smaller fragment (USNM 381945) from pit 42is perhaps another knife blade. From pit 11 was taken an incompletethin disk 55 mm. in diameter with a 5-mm. central perforation. Otherpieces are unidentifiable; one or two from shallow findspots are per-haps late intrusions, but the majority were inclusive in the aboriginaldeposits (pi. 13, &, c).Brass objects were in general much better preserved. Seven smallscraps are probably rejectage. A thin concave disk 40 mm. in diameterhas a central opening 17 mm. across, bordered by 3 concentric ridges(pi. 13, i) . Two coils of light wire, one with 3 turns, the other with 4,and with overall diameters of 18 mm., were found in pits 18 and 42(pi. 13, h). Thirteen crudely twisted cones range in length from 21to 65 mm. ; each has a small opening at the apex for passage of thethong or cord by which they were attached to garments or other items(pi. 13, d-f) . Two twisted and crushed tubes, the larger 43 mm. long,were presumably also for stringing; they were apparently made byturning up the edges of rectangular sheets and overlapping them.Hawkbells include a mangled specimen originally about 16 mm. indiameter (pi. 13, g) with 2 encircling lines, 2 holes in the lower half,and strap loop fastened by soldering ; a slightly larger example with2 pairs of encircling lines and with the strap loop fastened by insertinginto the body and then bending the ends over; and half of a thirdspecimen without recognizable markings or other distinctive features.There is also a strip of brass 35 mm. wide, which tapers toward abroken tip and has a broad tang at the other end, suggesting a knifeblade ; total length is 156 mm., and the convex edge is smoothed. Frompit 17 came 5 small circlets made by bending brass strips 2 to 3 mm.wide by 10 to 11 mm. long into a circle ; one still contains a bit of darkorganic substance that may be leather (pi. 13, j). Classed as annular Wedbl] an mTRODUCTION TO KANSAS ARCHEOLOGY 167beads when found, these are identical to objects found closely spaced onthong fragments from a burial at Doniphan (p. 126) and to othersdescribed as "fringe clips" from Arikara burials near Mobridge, S.Dak. (Wedel, 1955, p. 159).Two blue glass beads, both badly weathered, are listed in the catalogof Fanning materials; but unfortunately these have been misplacedand no further description can be given. One is from pit 22, the otherfrom pit 23. My field notes state that "a pea-sized blue glass bead"was found in the west corner center pesthole in house 1. A "tubularshell bead . . . trade ware" is listed from pit 31, and 2 from the sur-face ; but these also have not been found.STTMMAEY AND CORRELATIONSOur investigations at the Fanning site confirm earlier observationsregarding the presence here of important archeological remains, and,by extending our knowledge of their nature, they permit more precisecorrelations with other sites than have been possible heretofore. Anumber of points are still illuminated poorly or not at all, and someof these will probably remain so for some time; but progress hascertainly been made, and the general position of the materials in thepicture of eastern Plains prehistory is measurably clearer.The Fanning site, as indicated by low refuse mounds and abundanttrash-filled cache pits, marks the location of a semisedentary Indiancommunity of moderate size. Neither the number and density offormer house units nor their usual nature were disclosed by our work.The single structure indicated by a definite posthole pattern and bitsof burned grass-impressed clay was presumably an earth-covereddwelling with four center posts, like those of the protohistoric Pawnee.That other similar units once stood here, their traces now whollyobliterated by long cultivation, is possible ; but it is equally likely, Isuspect, that the more usual habitation was a surface structure withbark or mat covering, all evidence of which vanished with earlyplowing of the site. There is no indication whatever that the com-munity, which probably did not much exceed two or three hundredpeople, was in any way fortified.Subsistence was based in part on cultivation of maize, of whichthere is direct evidence. Use of beans and squash is inferential. Thebison scapula hoe is the only horticultural tool found, and the gardenswere probably in the valley bottoms at a little distance from the vil-lage. Wild fruit, nuts, berries, and tubers were undoubtedly used inseason ; and there is direct evidence that the groundnut {Apios ameri-cana) was harvested and strung for storage. The presence ofnumerous cache pits suggests surpluses to be cared for and drawn onduring the nongrowing season. Hunting was of considerable im-portance, with deer and other woodland forms taken in greater 168 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174abundance than the bison. Birds were used to lesser degree; offishing, there is no evidence, unless certain bipointed chipped flintsare correctly identified as fish gorges.The inferred subsistence economy and the probable nature of thecommunity pattern indicate that the natives here were participatingin a way of life widespread throughout the eastern Plains in very lateprehistoric and early protohistoric times. In the absence of fortifi-cations and in their probably prevalent use of a house type other thanthe earth lodge, the Indians at Fanning differed from their Pawneecontemporaries on the Lower Loup. In these respects, as also in thenature of their material culture inventory, there are much closer tieswith the Leary site on the Nemaha River, some 15 miles to the north-west. These similarities, especially prominent in the pottery com-plex, extend to other categories as well; and they serve to place theFanning site, as Leary has been (Hill and Wedel, 1936) , in the OneotaAspect. The Oneota Aspect is considered in more detail elsewhere inthis paper; but some remarks regarding the relationships betweenthe Fanning and Leary sites are relevant here.The pottery complex, well represented at both sites, may be notedfirst. At Fanning, as at Leary, 95 percent or more of the pottery isshell-tempered. The single restored jar from Fanning suggests thesame basic vessel shape indicated by the longer series at Leary butthere is no evidence of elliptical or horizontally elongate vessels fromthe former. Fanning site rims include more high straight forms thanthose from Leary. Trailed, incised, and punctate body decorationoccurs at both sites ; and this includes the characteristic use on shoulderand upperbody of blocks of parallel lines alternating either with otherblocks in which the lines slant in a different direction, or with areas ofpunctates. Especially characteristic at Fanning is the use of a singlevertical row of shallow punctates flanked on each side by zones ofparallel trailed lines. At Fanning, the decoration is usually less skill-fully done and occurs on only about 15 percent of body sherds as com-pared with nearly 33 percent at Leary. The circle and the cross, bothof which occur at Leary, are absent at Fanning. Lip decoration, bydiagonal incisions, punctates, and finger impressions, is found on 60 to70 percent of Fanning rims and on more than 90 percent at Leary;but the deeply notched or crenate lip so common at Leary is rare atFanning, as is the use of punctate elements on the inner upper surfaceof the rim. Handles are common in both, but occur more plentifully,in much more variety, and in greater elaboration at Leary. There areno "crested" or "tailed" handles from Fanning. Fanning pottery in-cludes relatively more heavy strap handles, these often wideningmarkedly toward the top, commonly with shallow fluting, and probablyassociated with a larger proportion of high heavy rims from large Wedel] an introduction to KANSAS ARCHEOLOGY 169 utility vessels. At Fanning, about 90 percent of handles were attachedat their upper end below the lip, whereas at Leary some 65 percent wereattached at the lip. There are no vertically bisected vessels, or sawedsherds at Fanning. As this recital of the major differences suggests,the pottery complexes at Fanning and Leary are by no means identical ; and of the two, that at Leary seems technically much the better andartistically the more advanced and unrestrained. At the same time, itshould be emphasized (1) that the two are sufficiently similar to bereadily distinguishable from any other pottery tradition of the trans-Missouri Plains, and (2) that in those details in which they are dis-similar, each shares its apparent peculiarities with other more distantOneota manifestations.With respect to nonpottery traits, close comparison is rather un-satisfactory because our excavations at Fanning produced a generallymore limited series of artifacts than is available from Leary. Commonto both, if sometimes in variable proportions, are the following : de-cided prevalence of small triangular unnotched chipped projectilepoints; abundance of end scrapers; unbeveled knives probably pre-dominant over beveled, but neither abundant; flake knives; plain-shafted straight drills, not common; bipointed chipped objects (fishgorges?) ; grinding stones (mostly small to medium at Leary withrotary grinding motion or pounding indicated, some evidence of to-and-fro grinding on larger slabs at Fanning) ; grooved mauls (crudeat Leary, reported as surface finds from Fanning) ; ground dioritecelts with elliptic cross section; paired sandstone shaft smoothers;sandstone sharpening blocks; pumice abraders; catlinite for pipes(disk pipes at Leary, pipe fragments only at Fanning) ; hematite withscraped and worn facets; hammerstones ; bison scapula hoes; boneawls scarce, apparently usually of split mammal leg bone ( ? ) ; bisonrib arrow straighteners or wrenches; polished bone tubes; needles(eyed forms from Leary, large flat mat-weaving (?) type fromFanning) ; and conical socketed antler tip projectile points, sometimeswith basal tang at Fanning.The above list includes a very substantial majority of the artifactmaterials from the two sites. A few artifact categories from Learyare not reported from Fanning. These include incised tablets and diskpipes of stone ; a chisellike object and a cylindrical section of antler,the latter with dressed ends ; and an ulna pick. The edge-slotted boneknife handle from Fanning, on the other hand, is not duplicated fromthe Leary site; nor are there materials clearly of white origin fromthe latter. I suspect that if we knew more about what was collectedat Fanning in prior years and what may still remain unfound, thesimilarities between the two sites might well be yet closer.There are, of course, other Oneota manifestations in the trans-Mis-souri region of which some note must be taken here. These include 170 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174abundant and important materials in Saline County, Mo. (Chapman,1946), the Stanton site in Stanton County, Nebr., and Occupation Bat the Ashland site, Saunders County, Nebr. (Hill and Cooper, 1938,pp. 267-271)." Detailed comparisons of these Avith Fanning cannotbe attempted here ; but it may be noted that Stanton and the Missourimaterials seem generally rather more like Leary than like Fanning.This applies to both pottery and nonpottery traits. In certain minordetails, as in presence of many high straight vessel rims and in ex-treme scarcity of punctate or other decorative elements on the innerupper rim, Stanton more closely resembles Fanning. In general, Ihave the impression that Fanning exhibits a somewhat simpler and lessartistic pottery complex than Stanton, Leary, and the Missouri Oneotasites, and lacks as well some of the more distinctive nonceramic traitsthere represented, such as disk pipes and inscribed catlinite tablets.It is worth noting, too, that Wliite contact materials are associatedwith the Oneota complex at Stanton and in Missouri, as at Fanning ; but they were not found at Ashland.The general chronological position of the Fanning site is not diffi-cult to establish. As adequate evidence indicates (Mott, 1938), theOneota Aspect is a relatively late manifestation throughout most ofits area of occurrence. It includes sites with, and others without.White contact materials. At Fanning, the discovery of small quanti-ties of iron, brass, and glass beads in some 17 of 46 refuse-filled cachepits, confirms a post-European contact dating. Beyond this, the prob-lem is not so simple. None of our finds can be exactly dated. Theyinclude, as already indicated, no gun parts, traps, axes, or other heavygoods. Knife blades are present, but otherwise only small ornamentsand trinkets seem to be indicated. Such items, I suspect, could as wellhave come from the packs of itinerant traders or trappers as fromvisits to an established trading post. The fu'st such post in the generalregion was Fort Orleans, founded by Bourgmond near present MaltaBend, Mo., in 1723 and abandoned a few years later. At this time,however, the only fixed Indian village of which we have record alongthe Missouri in present northeastern Kansas was that at Doniphan.The meager evidence we found at Doniphan, where White contactswere also indicated, faintly suggests possible contacts or relationshipswith an Upper Mississippi manifestation ; but it disclosed no site com-parable to Fanning or Leary. I believe that the Fanning site was "Griffin (1937 b, p. 297) long: affo suggested the "probability" that certain late ceramicand associated materials in the OzarJc region of northwestern Arkansas and southwesternMissouri (Harrington's 'Top-Layer' culture) would "eventually be classified In the Oneotaaspect of the Upper Mississippi phase." No detailed description of these remains as awhole has yet been published; but two statements (Delllnger and Dickinson, 1942; Belland Baerrels, 1951, p. 71) are relevant. From these, It appears that there are certainlyaffinities with Upper Mississippi, and also that there is Inadequate basis for further assign-ment to the Oneota Aspect as this is known from Iowa, Wisconsin, Minnesota, Nebraska,Kansas, and Missouri. Wedhl] an introduction TO KANSAS ARCHEOLOGY 171 already abandoned by 1724 ; and I suspect it may also have precededthe year 1700. In this case, its White contact materials doubtlesscame from some of the traders or trappers who were venturing insmall parties up the Missouri and its tributaries by the closing decadesof the 17th century.Additional evidence in support of this general placement derivesfrom the association at Fanning of local potterywares with severalgravel-tempered collared and cloistered rim sherds clearly assignableto the Lower Loup Focus of east-central Nebraska. Unfortunately,these exotic sherds do not permit precise dating, since the age of theseveral Nebraska sites from one or more of which they may have comeis not definitely fixed. In other words, they indicate contemporaneitybut not the exact time of that contemporaneity.As to tribal identity of the people to whom the Fanning site shouldbe attributed, nothing conclusive can be offered here. That the Oneotamaterials in general represent the remains of the Chiwere Siouans,and may also have entered strongly into the material culture complexof the Dhegiha, is generally accepted by midwestern workers. Spe-cific tribal correlations for the known trans-Missouri sites in Kansasand Nebraska, however, remain unproved. The lateness of the Fan-ning site, if my pre-1700 guess date approximates reality, and occur-rence there of White contact materials, would seem to narrow consid-erably the range of tribal possibilities. Documentary evidenceindicates that the Kansa had a considerable settlement in 1724 at pres-ent Doniphan ; and the Marquette map of 1673 suggests that the tribewas already in the general region by that time. Perhaps if we hadfuller data regarding the nature of Kansa material culture at Doni-phan, our problem would be somewhat easier. Lacking such a tieup,it still appears likely that the Fanning site was the location of theKansa as of roughly the time of Marquette or perhaps a little later.Identification of the Fanning site as Kansa is not without its diffi-culties. The Kansa were closely related to the Osage; and Osagearcheological materials of the late I7th and early 18th centuries withwhich comparisons should be made are not available. The Osage as-semblage from Vernon County, Mo. (Chapman, 1946) shows a numberof similarities to Fanning, though probably no more than do the Sa-line County materials attributed to the Missouri tribe ; but the VernonCounty remains are evidently of a later period than those at Fanning.What degree of similarity there may be if and when Osage sites con-temporary with Fanning are studied, I do not venture to predict. Itis, therefore, with some reservations that I concur with the suggestionmade some 40 years ago by Sterns ^^ that the Kansa were probablyresponsible for the Fanning site. ^ Unpublished manuscript, "The archeology of eastern Nebraska . . .," vol. 2, p. 176(Peabody Museum, Harvard University). 172 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174HILLTOP BURIAL SITESAlong the ridges west of the lower Wolf Creek valley are a numberof small inconspicuous burial "mounds." Most of these appear asvery slight elevations on the ridge summit; rocks of various sizes,some of them burnt, are usually scattered about on the surface overan area 5 to 8 yards across. Cultivation has probably reduced con-siderably their original height and subdued their contours, but I knowno description of their appearance before the land was broken. Forinformation on the location of these remains I am indebted to FennWard, rural mail carrier at Highland, who had tested one or two of the"mounds" and subsequently guided us to several of them. Mr. Wardwas also instrumental in securing permission for us to excavate.BA8 MATHEKSON SITETwo mounds on the Ras Matherson farm, circa 2 miles west ofSparks and 400 to 500 yards south of United States Highway 36, wereselected for test. Situated on a commanding ridge, with no con-venient site for a village within several hundred yards, the two wereabout 100 yards apart. The northernmost of the two was examinedfirst. Two trenches, each 3 feet wide and 32 feet long, were dug so asto intersect at a right angle at what appeared to be the mound center.At the outer end of each trench clean yellow-brown clayey subsoil wasfound 12 to 14 inches below ground surface. As the point of inter-section was approached, bits of brick-red burnt earth appeared inincreasing numbers, extending a few inches below the plowed zone.Large irregular chunks of the same material lay under and just eastof the center stake. This burnt earth and occasional bits of scorchedbone, were the only evidence that the ground had ever been disturbedby man.About 30 inches east of our center stake, a layer of stone slabs wasuncovered (pi. 14, d). It measured 58 by 30 inches, the long axisrunning north to south, and was designated burial 1. Five feet to thenorthwest was a similar but smaller slab mass ; and small stones werescattered here and there to a distance of perhaps 6 feet southwardfrom the two slab areas. Removal of the stones from the larger groupdisclosed an elliptical pit 51 by 27 inches with a total depth from thesurface of about 20 inches. The pit fill was dark in color with someadmixture of charcoal, contrasting markedly with the clean, lighter-colored unmixed soil of the walls. The fill was worked over meticu-lously, yielding but two or three very small bits of human bone.Under the second rock mass were the somewhat more plentiful butequally fragmentary remains of another burial ; the bones, includingnumerous skull fragments, covered an area about 21 by 24 inchesacross, and all were within 10 inches of the surface. Many of the Wbdbl] an ESTTRODUCTION to KANSAS ARCHEOLOGY 173fragments were fire blackened. There were no artifacts in association,nor could a pit outline be fomid. This was designated burial 2.Beginning at a point about 6 feet south of our center stake, wenext proceeded to excavate all moved dirt to whatever depth and dis-tance were indicated by burnt earth, charcoal, or other admixture.This resulted in the delineation of a basin 24 inches deep at the center,with steeply sloped curving walls, a north-south diameter of nearly 11feet, and an east-west diameter of 9i/^ feet. Except on the east side,where the grave pit of burial 1 partly intersected the basin, there wasno break in the wall outline. The basin fill contained much fire-reddened earth, particularly at and near the center, but there was notrace of ash. Mixed through the fill were bits of human bone, some ofthem fire blackened; the entire basin yielded somewhat less than acigarboxful of this scrappy material. There were no artifacts, andpractically no stones. So far as we could tell, there had been no dis-turbance of the mound or its contents other than some superficialresults of plowing.About 5 feet north of the above basin lay a smaller but deeper one.Measuring 95 inches north to south by 38 inches wide by 36 mchesdeep, this contained only dark moved earth without artifacts or tracesof bone. Its purpose is not known.From our findings, I would surmise that the large shallow basinwas dug for no other purpose than burial, and that into it werethrown the fragmentary remains of bodies that had been crematedsomewhere nearby. Over the nearly filled basin were placed stones,perhaps by way of discouraging disinterment of the remains by wildanimals. If the rather ambiguous statements of some of the olderresidents are credible, there was also some mounding up of earth overthe burials. The smaller stone-covered areas which we designatedburials 1 and 2 may have been laid down at a different time. Burial2 was certainly later than the main basin, but the well-defined pitnoted as burial 1, wherein was found no fire-reddened earth, couldhave been an earlier feature. The general similarity, however, inthe fragmentary skeletal remains, the stone coverings, and completeabsence of artifacts rather suggests that all the burials were probablyleft by the same people. It is possible, too, that the smaller bone-freebasin north of the large one was dug for further burials, but for somereason was never so used. There is, of course, no way of estimatingthe number of individuals represented in the various basins and gravepits, or of determining their one-time cultural affiliations.The second mound had been dug into a year or two prior to ourwork, by Mr. Ward and others. As our tests progressed, it becameevident that nearly all of the burial area had been worked over.Bone fragments, some of them with fresh breaks, were encountered 174 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174in small quantities; and the presence of a few boulders suggested asetup probably paralleling that we found in the other mound. Theonly undisturbed find worthy of note was a very badly broken skull,dissociated from any long bones, stones, or other remains. No basinoutline could be determined; and, as in the other mound, there wereno artifacts. A. W. GUTHEIE SITEAbout 31/^ miles north of the Matherson mounds, and a mile orslightly more south of Iowa Point, we opened another slab-coveredgrave on the Paul Guthrie farm. The grave was situated on a highpoint about 400 yards north of Cotter Branch of Indian Creek (pi.14, a) and within a few hundred yards of the Missouri River bluff.From the burial point, a narrow descending ridge projects southwardtoward the creek, which carried some water at the time of our visit.Potsherds, projectile points, and other artifacts, apparently repre-senting more than one cultural horizon, have been picked up on theGuthrie farm, where a stone-chambered burial mound was also openedlong ago (Wedel, 1943, p. 160) . We found three grit-tempered cord-roughened sherds on the slope below the burial site. Unhappily, wewere unable to make a thorough search of the tract, and so I can makeno statement as to probable nearby village sites.The grave under consideration had been exposed m plowing a yearor two before our arrival, and stones of various sizes were scatteredabout. We cleared away 4 to 5 inches of loosened topsoil, exposing alayer of fine-gi'ained gray sandstone slabs covering an area about 5feet across (pi. 14, h). Included were slabs as much as 27 incheslong by 18 inches wide, whence they ranged downward to mere peb-bles; none exceeded 2 or 3 inches in thickness. Their source is un-certain but they seem to have ben carried in from somewhere down theslope several hundred yards away.Below the slabs was a subcircular pit slightly over 5 feet in diameter ; the sides were steeply sloped, and the maximum depth was about 30inches. On the southeast and northeast sides, about 10 inches abovethe main pit bottom, was a shelf or bench about 18 inches wide. Onthis bench 10 inches below the largest slab, was a skull fragment;other scraps of human bone were scattered about through the fill inthe deeper part of the pit. There was no evidence of burning on thebones or in the soil, nor were there any artifacts. It is possible thatthese bones were the remains of a body previously exposed on a scaf-fold, and already subjected to weathering and disintegration beforeinterment. It is wholly impossible to determine the cultural horizonto which the burial should be assigned, or its age (pi. 14, c).Our rather inconclusive findings on the Matherson and Guthriefarms appear to be in line with those of local enthusiasts. We weretold that artifacts are generally absent, that the skeletal remains are Wkdbl] an introduction TO KANSAS ARCHEOLOGY 175 almost always very fragmentary and scattered, and that the humanbones frequently show scorching. That the remains in the Mathersonmounds are from cremated bodies seems a reasonable inference, inview of the evidence of scorching and of long-continued burning ofsome of the soil in the fill. At the Guthrie place, on the other hand,no burning was indicated, but the fragmentary nature of the bonesmay be taken as evidence of secondary interment here also. Inneither case, unfortunately, is it possible to relate the burials to anyof the several archeological horizons that are evidently present in thissection of the State.Surface collections owned by Mr. Guthrie included glass beads andother objects of historic age; grit-tempered plain and cord-roughenedsherds, including rims with slight collars whose lower border wasscalloped or noded, all probably assignable to a local Nebraska Culturemanifestation; and several grooved axes, ground celts, a stemmedscraper, and a large grit-tempered mica-flecked sherd with rocker-roughening, all reminiscent of Hopewell remains in the vicinity ofKansas City (Wedel, 1943).MISCELLANEOUS SURFACE FINDS IN NORTHEASTERN KANSASAs is clear from the foregoing pages, our limited excavationsin 1937 along the Missouri and lower Kansas Rivers furnished evi-dence of the former existence in this part of the State of a variety ofsemisedentary ceramic complexes. Others undoubtedly await discov-ery and definition. There is, furthermore, clear indication of stillolder preceramic horizons here as well. None of our fieldwork wasconcerned with such materials, but we were told of occasional finds ofprojectile points and other artifacts differing widely from those usu-ally acquired by local collectors. Through the courtesy of Fenn Ward,of Highland, Kans., I am able to illustrate (pi. 15, a^j) and reportbriefly a series of points that give interesting indication of the natureof some of these early remains. All of those figured are surfacefinds in and near Doniphan County, and are without direct culturalassociation.The five specimens shown in the upper row (pi. 15, a-e) are onlygenerally suggestive of possible early materials. Their shapes andcertain other characteristics set them off from the specimens usuallyidentified in and about the Central Plains with "Woodland and laterceramic manifestations. Thinning of the base is indicated in a and e,but only in the latter can it be said to approach actual fluting. On allfive specimens, the lateral edges have been blunted by grinding for adistance of 10 to 13 mm. above the basal corners ; in e, this bluntingextends from 23 to 33 mm. up the edges. I have not observed thisfeature on any of the points excavated by us in Kansas sites of theceramic complexes.484172?Bfl 18 176 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174Early point types are more strongly suggested by the central blockof specimens illustrated, notably in / to j. The largest of these, /, isof mottled gray and tan chert, measures 9.9 by 2.7 by 0.55 cm., and isfluted on both faces. Secondary chipping along the edges is coarse andirregular ; the indented base and the lateral edges for some 35-38 mm.above the corners are ground smooth. It is a good example of theEastern Fluted type, whose geological context and age remain unde-termined. Typical Folsom points are represented by h and i, the for-mer lacking the base. Both are broadly fluted on each face to within5 to 10 mm. of the tip; and the edges show fine even chipping withflake scars measuring 1 to 1.5 mm. each in width. Much less well made,but falling within the range of Folsom points, is j\ the fluting ispoorly done, but the lateral edges adjacent to the base have beenblunted. Still another early form, the Plainview point, is suggestedby ^, in which the lateral edges are blunted for about a third of theirlength above the base, beyond which the edges converge to the tip.There is a suggestion of a medial ridge on one face. This specimenrecalls the somewhat narrower basal fragment of a Plainview-likepoint reported by Solecki (1953 a, fig. 2) from site 14MH33 in TuttleCreek Reservoir, though the base is more nearly straight in the presentpiece.Since these specimens, as noted, are all surface finds without associa-tions, they must be regarded for the present as suggestive rather thanas definitive evidence of preceramic peoples. Their recurrent discoveryat various locations, and the variations in type, nevertheless hint at thepresence of early big-game hunters in this section of Kansas perhapsas much as 8 or 10 millennia ago. What is needed now, of course, is adiscovery of fluted and other early point types in situ, with geologicand paleontologic contexts that will permit their placement in timeand with artifact associations from which can be determined the localcomplex to which they belong. The glaciated northeast section ofKansas in which these pieces were found would seem a likely area foreventually discovering such associations.SITES NEAR MANHATTAN, IN RILEY AND POTTAWATOMIE COUNTIESManhattan lies on the north bank of Kansas River, approximately120 miles above its mouth and just above the point where it receivesthe Big Blue River from the north. The Kansas and Big Blue, bothperennial, flow in attractive flat-floored valleys from 1 to 3 mileswide, bordered by limestone-rimmed blue's up to 200 feet or more inheight. Both valleys, though occasionally swept by devastatingfloods, have fine terraces that are normally flood-free. Beyond thebluffs are, or formerly were, rolling tall-grass prairies intersected bysmall creeks, which, becoming perennial in their lower courses, at inter- Wkdel] an introduction TO KANSAS ARCHEOLOGY 177 vals of a few miles break into the main valleys from both sides.Practically all these creeks, like the main streams, were formerly well-timbered, as some of them still are, with a variety of hardwoods.Some juniper occurs on the steep north-facing slopes of the valleys.Until the 19th century, it may be supposed that the wooded valleysand nearby grassy uplands provided an abundant and varied faunathat could be easily and conveniently drawn upon by the Indiansdwelling on the valley terraces; and the fertile bottom lands plusan average annual precipitation in excess of 31 inches made possible arelatively sure and productive horticulture. An added attractionwas the chert exposed in limestone ledges along the creek valleys, es-pecially south of Kansas Kiver, this doubtless furnishing the grayto blue-gray stone used extensively in manufacture of chipped im-plements. All in all, it is difficult to visualize a locality better suitedto aboriginal occupancies, whether these depended primarily on hunt-ing and gathering, on the cultivation of crops, or on a combinationof these practices.That an abundance and variety of archeological remains onceexisted in this locality has been known for many years, thanks to thereports of Brower (1898, 1899) and others noted in another section ofthe present paper. These remains included village and campsites,frequently with ceramic materials ; burial mounds of varied construc-tion and content; workshop areas littered with great quantities ofchipped chert implements ; and, according to Brower, aboriginal chertquarries for which he presents no direct evidence of ancient workings.Unfortunately, most of these early-day investigations were made byindividuals with little or no scientific training, and the literaturepertaining to them is sadly lacking in the detailed information neces-sary for establishing associations and defining complexes. Muchfurther investigation on a continuing basis will thus have to bemade in order to define clearly the various native manifestations thatare evidently present, to establish their sequence, and to fit them intothe larger picture of regional prehistory.Our primary interest in the Manhattan locality was the examina-tion of the site of the Kansa Indian village of circa 1800-1830, re-ported just below the mouth of Big Blue River by Say and othercontemporary observers, and a preliminary survey of some of theother sites noted by Brower and others. Our stay was a short oneof less than 3 weeks ; but it produced ample confirmation of what washinted in Brower's reports, namely, that the locality was peopledby Indians at several different time periods and that sites are plenti-ful. Mounds were formerly present on the bluffs, and some of thesewere probably of some antiquity, though the records are vague, in-complete, and indeed often nonexistent; and these relatively con- 178 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174spicuous features, of course, have suffered a good deal more fromcurious amateurs than have the more obscure village and campsites.THE GRIFFING SITE (14RY21), IN RILEY COUNTYJust above Manhattan, Kansas Kiver is joined fi*om the north byWildcat Creek. Heading in the uplands of central Riley County,near the little town of Riley, Wildcat Creek runs southeasterly forsome 20 or 25 miles to its union with the Kansas. It has a pleasantlittle valley with a gentle grade, and through this valley nms theline of the Chicago, Rock Island, and Pacific Railroad. We attemptedno survey of the valley as a whole, nor did we examine two othersites reported by Brower (1899, p. vi)?one upstream, the other down-stream, from the Griffing site. Casual inspection suggested, however,that the lower portion of the vaUey especially, from approximatelyKeats downstream, must have had much to offer a native semiseden-tary population. On both banks of the creek lie attractive terracesthat are normally flood-free and widen as one descends the valley;the creek carries a good flow of water; and the hardwoods growingalong its banks and on the steep south bluffs provided in aboriginaltimes an adequate supply of building timbers, fuel, and cover for game.The Grifl^g site is located about a mile west of the Manhattancity limits, and 11^ miles above the point where Wildcat Creek issuesfrom the bluffs to cross the Kansas River floodplain. Owing to un-satisfactory ground conditions, we were not able to define the exactlimits of the site; but the surface evidences noted suggested that themaximum occupancy was on the terrace that the Rock Island rail-road crosses to the north side of the valley (fig. 27). This is sub-stantially the area sketch-mapped by Brower (1898, p. 24). The ter-race here is about half a mile wide, and lies at approximately 1,050to 1,061 feet above mean sea level.^'' Areas shown as under 1,050 feet,usually separated from the terrace proper by a scarp 10 to 12 feethigh, are apparently subject to frequent overflow and constitute afloodplain; they include old cutoffs of Wildcat Creek, now partiallyfilled in and under cultivation, but not usually utilized for residencein aboriginal times. The creek pursues a sinuous course down thevalley in a channel some 20 feet below the terrace surface. Southof the creek steep wooded bluffs, their summits utilized for erectionof burial cairns by the Indians, overlook the site; northward, thevalley wall rises more gradually.According to Brower, 19 slight elevations which he took to belodge sites were visible here in 1897. He reported that the site hadbeen "veiy thoroughly explored" by W. J. Griffing; but there is no " Elevations given here and In figure 27 are approximate. They are based on an as-sumed elevation of 1,060 feet, m. s. 1., for the Chicago, Rock Island, and Pacific Railroadbridge (No. 1464C) crossing Wildcat Creek near the east end of the archeological zonehere included in the Grifi^g site. Figure 27.?])on an ai484172?59 Figureon27.-Map of portion of Griffing site, 14RY21, showing features investigated by U.S National Museum 1937 Ekv^ions shown are based1 an assumed elevation of 1.060 feet, m.s.l., for the Chicago, Rock Island & Pacific Railroad bridge No. 1464 C sho^n at lo^^er right.<84172?59 (Face p. 178) I i Wedbl] an introduction TO KANSAS ARCHEOLOGY 179way of determining what these and other early-day investigationsmay have involved in terms of actual excavation. Brower's mapshows three lodge sites east and north of the railroad bridge, includ-ing one on the south bank. The others were scattered loosely up thevalley for more than half a mile; most of these were on the widernorth terrace, but four are indicated on the south side of the creek.Of these elevations, owing to long-continued cultivation, little ornothing remained at time of our work; but the probable location ofsome could still be approximated by surface concentrations of pot-sherds, flints, and burned house daub bearing grass and twigimpressions.These indications, it seemed to us, were most promising at spotsthat appeared to coincide with two lodge sites shown by Brower oneither side of the railroad north and northwest of the bridge, and alsoat approximately the point where he locates a "prehistoric kiln" atthe bank of an old westward loop of the creek (cf. Brower, 1898, p.24, and my fig. 27 and pi. 16, &).HOUSE TEST 1Our first excavation unit, designated house test 1, was about 70yards south and west of the railroad and 380 yards west by north ofthe bridge, at the edge of the terrace overlooking the creek bottomsto the south (see fig. 27). Here the plowed surface was littered withbits of burned grass-impressed clay daub, charcoal, potsherds, andother occupational debris. We stripped an area about 25 by 40 feetto a depth of 16 inches. Burned clay and charcoal continued belowthe plow-disturbed zone, and in this fill occurred also cord-roughenedsherds, worked hematite, and flints. At 16 inches, the mixed fill gaveway to clean undisturbed soil. Despite careful peeling of the floorof our excavation where this soil change took place, there was not theslightest trace of a fireplace or of a definable living level, and only themost unsatisfactory evidence of any postholes. In the southeast partof the cleared area we found a cache pit ; and from this were taken alarge milling slab, a smaller slab bearing nine pits or "cups" on onesurface, portions of a medium-sized pottery vessel, much burned daub,and flint chips. The conviction persists that we were working in ahouse site ; but the total absence of definable items such as hearth andpostholes precludes any statement as to its size and the arrangementof features in it. HOUSE 1Better success rewarded our efforts on the east side of the railroadright-of-way. Here, on a narrowing strip of terrace between the rail-road and an old channel of the creek (see fig. 27) , and some 200 yardseast of house test 1, we trenched an elevation rising a few inches abovethe terrace surface. Burned clay daub, charcoal, and sherds were 180 BUftEAU OF AMERICAN ETHNOLOGY [Bi:ll. It4fairly plentiful to a depth of about 1-1 inches, where the mixturechanged abruptly to a clean unmixed soil. Beneath the highest pointof the elevation, at a depth of 18 inches, was a well-marked fireplace ? a shallow basin of fire-reddened earth, 24 inches in diameter, contain-ing a little wood ash. Working out on the probable floor level in alldirections from the hearth, we next located 4 large postholes at aradius of 11 to 12 feet. These varied in diameter from 15 to 18 inchesand in depth from 21 to 24 inches. All contained a mixed fill and char-coal ; in one, at approximately the original floor level, vestiges of thepost remained as a 13-inch ring of charcoal with the grain of the woodrunning vertically. Beyond these 4 holes were 43 smaller postholes,averaging 6 to 7^/^ inches in diameter by 6 to 14 inches deep. Thesewere spaced at irregular intervals of a few inches to nearly 5 feet;some of the longer gaps may indicate that we missed one or morepostholes. This series formed an approximate square with roundingcorners (fig. 28), and with a gap on the south marking the doorway.On the south side also were two large postholes, one on either side ofthe doorway and about midway between the door and the southeastand southwest corners of the house. These were 12 inches in diameter ; the one on the east was 18 inches deep, that on the west was 21 inchesdeep (pi. 16, c).Except at the south side, the mixed fill ended a few inches outsidethe square of small postholes, suggesting that the shallow pit overwhich the lodge had been erected had measured approximately 32 by34 feet, with the long axis north to south. On the south, a strip ofrefuse-laden fill 5 feet wide continued for some 12 feet to the end of theentrance passage. The passage was further defined by 2 rows of post-holes, 4 on the east and 5 on the west, these being closely comparablein size and depth to the outer series in the house site.Beneath the floor, and in all likelihood directly associated with thehouse, were five cache pits. The limited material taken from thesedid not differ significantly from that found in the house fill and on thefloor. The three larger pits were located in or near the southeast,northeast, and northwest corners ; two smaller ones lay west and south-west of the hearth. There was also a small pocket cache near the mid-dle of the east wall. All these features are shown on the house floorplan (fig. 28) ; but through some strange oversight I neglected to placethe pit numbers on the field map and so cannot now identify thesefeatures specifically. The sizes of the pits were as follows : Diameter (inches)Pit. No. : Top Bottom Depth (inches)1 36 36 162 34 36 333 24 36 124 12 18 145 36 38 32 Wedbl] an introduction TO KANSAS ARCHEOLOGY 181 "t) ? O o o O ?? ? ? ? ? ? 9> 9 9_?/ /o /sFEET Figure 28.?Plan of house 1, Griffing site (14RY21). Solid circles indicate outer post-holes; heavy circles, primary (center) postholes; rayed circle, fireplace; unnumberedcircles, cache pits; broken line, approximate edge of house floor.OTHEE EXCAVATIONSA few yards east of house 1, on the terrace edge overlookmg an oldbend of the creek, was another concentration of burned clay, charcoal,and potsherds. Cultivation had spread this debris down the bankonto the partially filled creek bed. Tests at the upper margin dis-closed some refuse below plowsole over an area about 35 feet across.This is, without much doubt, the location of Brower's "prehistorickiln." We trenched this thoroughly; but the relative thinness ofdetritus below 6 to 8 inches depth offered little inducement to ex- 182 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174tended excavation. There was no trace of floor lines, hearth, or otherrecognizable house features, and we finally abandoned the spot. Ofhis investigations here, Brower (1898, p. 29) wrote : . . . The narrow walls of a small oven made of clay plastered against a founda-tion of straw and reeds, were obliterated by the weight of an accumulation ofdebris, and many pieces of burned clay were scattered over the surface of theplowed ground. From what appeared to be the floor of the kiln we recoveredseveral large potsherds, one piece being about the third part of a whole earthenvessel, which has been . . . shown in Plate VII. A piece of the clay wall of thekiln is illustrated in Plate VIII.The fragment of "kiln wall" illustrated by Brower is evidentlygrass-impressed clay and suggests to me that he was actually workingin a house site or in house site debris ; the "oven" was perhaps a cachepit. Brower's work here may have contributed to the obscuring orobliteration of other evidence which would have enabled us to identifythe spot as a house ruin ; or it is possible he was digging in a middenarea associated with our house 1 and developed over an exterior cachepit. In any case, the restored pot he illustrated is without much doubtthe same potteryware we found in our various excavations on the site ; and so, whatever the true identity of his "kiln," it undoubtedly pertainsto the same complex suggested by our findings at the Griffing site.We made tests at several other detrital areas within the site asBrower delineated it. One of these was north and west of house test1, just south of the railroad, and again near one of the lodge sitesshown on his map. The subsurface evidences were considered insuffi-cient to justify large-scale excavation. Farther up the valley, severalsmall terrace remnants south of the creek and one larger refuse-bearingtract subject to inundation were also examined, but these were eitherjudged unpromising or were not available for extended excavation.Neither did we devote any time to a cairn on the point of the bluffsoverlooking the east end of the village site, since it had evidently beenrifled long ago. I have little doubt that additional house sites couldbe located along this section of Wildcat Creek, if thorough search weremade for them at various seasons; and while some may have beengutted or badl}'^ damaged by local collectors in, before, and sinceBrower's time, there are probably others that might still repay sys-tematic excavation. Since we were able to define but one house siteand our artifact sample is small and incomplete, additional data thatcould be gotten through further exploration would be well worth while.ABTIFACTSThe artifact yield from our excavations at the Griffing site wassurprisingly small, even granting the limited extent of our work.Pottery remains totaled less than 300 sherds, including both surfaceand subsurface specimens. Stonework, including chipped and ground Wbdel] an introduction TO KANSAS ARCHEOLOGY 183pieces, numbered less than 30 specimens. We found no objects ofbone or shell, and virtually no refuse or scrap in either of these cate-gories. Possibly local soil conditions, drainage, or other factors wouldaccount in part for the general lack of such organic materials; butone wonders whether much of the refuse was removed from the housesand discarded elsewhere?perhaps over the creek banks or in middensat a little distance from the dwellings.PotteryCeramic materials recovered by us from this site consist of 62 rim-sherds and 225 body sherds; no whole or satisfactorily restorablevessels were found. The great bulk of this came from house 1 andhouse test 1, and from the excavated pits within these features.Surface finds include 2 rim and 14 body sherds. There is a good dealof sameness among all these specimens, but sufficient differences seemto exist between the rim fragments to suggest that between 40 and 50vessels are probably represented. All may be assigned to a singleware, here named Riley Cord-roughened, whose description follows(see pi. 17) . Riley Cord-roughenedPaste :Tempering: Rounded siliceous (sand) particles, 0.5-3.0 mm. diam., spar-ingly to abundantly used, and often protruding from the surfaces ; somesherds include also angular particles of indurated clay (or crushed pot-sherds?) varying in color from light gray to nearly black, and havingthe same hardness and testure as the sherd body ; a few sherds possiblyfrom a single pot have shell inclusions, and shell is suggested also bya few "hole-tempered" fragments.Texture: Fine, even, usually fairly well-compacted; fractures rough anduneven, but not crumbly; occasional tendency for sherds to split alongtheir midline on a plane parallel to the surfaces.Hardness: Varies between 3 and 4.5, most sherds apparently just under 4.Color: Sherd cores are nearly all slate gray, rarely almost black, butsometimes light brown or tan ; surfaces usually gray but in many caseswith an orange-brown color ; firing clouds.Method of manufacture: Undetermined, no evidence of coiling in this sample.Surface finish: Exterior surfaces almost invariably cord-roughened; in-dividual impressions are 2-4 mm. apart, and may be vertically appliedor else crisscrossed at various angles ; interior surfaces smoothed, butoften unevenly, and sometimes striated. A noteworthy characteristic isthe disintegration and scaling away of surfaces, especially the exterior,so that the tempering particles protrude from a chalky-textured grayor brown matrix and impart a coarse sandpaper "feel" to the touch(cf. Hill and Cooper, 1937, pp. 232, 259).FoEM : Vessel shapes: No whole or restorable examples found; larger sherds andpartially restored segments suggest small to medium vessels, full-bodied,round in horizontal cross section, with rounded shoulder, constrictedneck, and presumably a rounded base; miniature pots of same generalform indicated ; no recognizable bowl or other shapes. 184 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174Size: Restored segments suggest vessels up to 25-30 cm. in maximum bodydiameter.Thickness: Sherd thicliness ranges from 3.5-15.0 mm.; of 50 random sherdsmeasured, 2 (4 percent) were under 4 mm., 40 (SO percent) were 4-8 mm.,8 (16 percent) were over 8 mm., and the mean was 6.2 mm.Handles: One small plain strap handle from miniature pot present ; probablehandle scars noted on other sherds.Rims: Two general forms are indicated, (a) Most common is a simple un-thickened rim which rises from the neck and flares outward in a gradualcurve to a rounded, or less commonly flattened or narrowed lip. Thedegree of curvature varies considerably; and In the larger, i. e., higher,examples, the rim is nearly straight, but in no case is it vertical or, withone exception, incurving. These rims are 8-40 mm. high, but mostlyunder 25 mm. They number 46 of 62 specimens, or 76 percent. ( 6 ) Thesecond form (16 specimens, or 24 percent) consists of rims which have athickening below the lip so that the profile shows a collar overhangingthe neck. The collar varies from very weak to moderately pronounced,but it is usually much less prominent than in Upper Republican potteryof the Lost Creek and Medicine Creek localities of Nebraska. Two ofthree of these rims have a pronounced flare ; others are nearly vertical ; and in 2 or 3, the curve is inward so that the lip diameter apparentlydid not exceed the neck diameter. Rim height here varies from 16-35mm., but is usually between 25 and 30 mm. Rounded or thinned lips arecharacteristic.Decoration : Occurs very rarely. One body sherd has deeply scored lines sug-gesting concentric upward-pointing chevrons or triangles ; a low recurved rimhas an incised zigzag line on the lip, shallow indentations or scallops on theouter lip edge, and 3 parallel incised lines which presumably encircled thevessel below the neck; a surface rimsherd has remnants of incised linesapparently oppositely slanted in adjacent blocks. The simple rims are usuallyundecorated, but 3 have the outer lip edge scalloped by notching with thefingers. Collared rims have either cord-wrapped paddle impressions on theirouter surface (8), incised parallel horizontal lines (1), or scalloped or notchedlower edge (3).Although our pottery sample from the Griffing site is admittedly asmall one, I feel sure that the potteryware named and described abovewill be found to recur with only minor variations over a considerablearea in the drainage of the Kansas and lower Smoky Hill Rivers.Sherds readily classifiable as Riley Cord-roughened are apparentlya characteristic ware of the Manhattan locality. They were foundat practically all pottery-bearing sites seen by members of our 1937party on the Kansas and lower Big Blue Rivers, as well as on the nu-merous creeks?Lyons, Clark, Humboldt, McDowell, Deep, and Mill,with their branches?joining the Kansas from the south. Local col-lectors and informants maintained that all pottery found in thislocality was the same type ; and such collections as we examined seemedto support this assertion. As we shall see presently, other and distinctpotterywares do occur here, however, but in lesser abundance andon fewer known sites. It should be noted, further, that a good manyof the sites and local collections characterized by Riley Cord-rough- Wkdbl] an introduction TO KANSAS ARCHEOLOGY 185ened sherds also have lumps of grass-impressed house daub, suggestinga direct association such as we established at Griffing site between thisware and earth-covered dwellings, the latter probably typically ofsquare or rectangular form. The wider relationships of Kiley Cord-roughened pottery and its associations will be discussed elsewhere ; butit may be noted here that it has close resemblances to pottery fromMinneapolis on the Solomon Kiver (Wedel, 1935, p. 225) , and probablyto cord-roughened pottery from Salina, from Chase County, and else-where in east-central Kansas.Work in Chipped StoneMost of the chipped stone artifacts were fashioned from the graychert native to the locality. They include projectile points, knives,celts, and scrapers. Relatively few specimens can be said to have beenwell made.Projectile points, four in number, include two triangular unnotchedspecimens from house 1 and two side-notched fragments from housetest 1. The first two have slightly convex bases, convex blade edges,and are moderately well fashioned ; they measure 37 by 15 by 3.5 mm.and 24 by 13 by 5 mm. A third specimen from house 1, 35 by 17 by 5mm. in size, is possibly also of this type ; but the tip has been intention-ally rounded off so that there would have been poor penetrating power.The notched specimens include one small thin flake, retouched onlyalong the edges and base, and provided with broad shallow notchesjust above the base ; and a fragment of a small well-made point brokenoff at the notches. Doubtful projectile points include a fragment oftriangular form 26 mm. across the base, with tip missing ; and anotherthick piece with rounded base and edges, measuring 27 by 17 by 7 mm.Knives include 4 specimens, all from house 1. They are more or lesselliptical to leaf-shaped, with one or more of the longer edges re-touched from both faces. The ends are usually rounded ; one is squared.They range in length from 53 to 69 mm., in width from 26 to 37 mm.,and in thickness from 6 to 17 mm.Two celts were found in house 1. They are thick heavy pieces, withbroad rounded blade and tapering to a narrow rounded butt. Maxi-mum width occurs about one-third the distance up from the blade.Both show wear, with the edge of the shorter specimen blunted fromuse. They measure 140 by 55 by 23 and 127 by 57 by 32 mm. Theyresemble in all particulars one of the forms of heavy chert implementsfound formerly in great abundance on the surface of many sites inthe Manhattan locality ; and their inclusive presence in a habitationsite, plus the indubitable evidence of use, suggests at least one prehis-toric horizon with which artifacts of this sort were evidently directlyassociated. 186 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174Unifacially worked chert objects include end scrapers and sidescrapers. End scrapers, of which there are 12 specimens, fall intotwo groups. Most common are scrapers with a prominent ridge alongthe back, either along the midline or to one side or the other. Only 2have all-over retouching and retain no trace on their convex surfaceof the original surface of the flake used in their manufacture. Threeof these specimens retain parts of the white matrix enclosing the stoneof which they were fashioned. Scrapers of this group have the follow-ing size range : length 44 to 94 mm., mean 59.6 mm. ; width 16 to 39mm., mean 26 mm. ; thickness 8 to 20 mm., mean 12.4 mm.Three other scrapers are thin, low-backed specimens, made from thinspalls that required only edge retouching to make them serviceable.They range in length from 40 to 70 mm., in width from 20 to 33 mm.,and in thickness from 6 to 7 mm.Two side scrapers were made from thick planoconvex spalls, andhave prominent ridges on the back. Retouching of the edges is limitedto the long sides, not to the end, as with the preceding group. Theymeasure 55 by 36 by 12 and 50 by 35 by 20 mm. Both are from house 1.WOBK IN Ground StonbSix specimens only of ground-stone work are present, and thesewere made of sandstone and limestone. Abraders of sandstone wereof two kinds. Two fragments are evidently from elongate shaftsmoothers of the type commonly used throughout the Plains in pairs.The larger fragment, which has one finished end, is 85 by 32 by 30 mm. ; it has a longitudinal groove on each of three sides. The smaller frag-ment, which has a groove on each of its four sides, is from a smootherof comparable size and shape. Some of the grooves seem too narrowand irregular for arrowshafts and probably result from awlsharpening.A different type is represented by two fragments that fit together toform an incomplete block 82 by 42 by 18 mm. The two broader sur-faces are worn concavely, from rubbing or sharpening activities. Oneedge has a narrow V-shaped longitudinal groove. The entire piecehas imquestionably been dressed to shape, but its original size andform are conjectural.A large milling stone was made from an irregular slab of limestone,dressed roughly to shape, and measuring 44.5 by 36.8 by 7.7 cm. Theupper surface has an elliptical concavity, evidently produced by a ro-tary motion rather than by a to-and-fro grinding motion. No mullerswere recovered, though unifaced manos of limestone and other ma-terials have been collected from the surfaces of other sites in thelocality.From the same pit that produced the milling slab came anotherlimestone slab, somewhat smaller, and measuring 25.4 by 22.8 by 6.3 Wedbl] an introduction TO KANSAS ARCHEOLOGY 187 cm. One surface is flat, but not certainly dressed. The other is un-even, and bears 9 shallow pits or "cups," each about 15 to 25 mm. indiameter, scattered irregularly over it.Other Stone ObjectsA flattish waterworn pebble, approximately 55 by 35 mm., has bat-tered facets on the edge which were undoubtedly produced by use asa pecking or hammer stone. It does not appear to have been otherwisemodified.A quartzite pebble 35 mm. in diameter shows no workmanship be-yond the removal of several small flakes. There is no way of deter-mining whether this was accidental or purposeful.COMMENTDespite the very limited extent of our work at the Griffing site,certain inferences seem warranted at this point. That the site wasoccupied by a semisedentary group is suggested by the presence ofearth-lodge remains, these being indicated, as Brower inferred, by theoccurrence of low mounds scattered loosely for some distance along thecreek. The nature of the settlement, open and unfortified, and thepresence of square or rectangular dwellings, further suggest relation-ships to other prehistoric communities of square houses widespread inthe Central Plains. The subsistence economy doubtless includedmaize agriculture, with which was associated the metate-mano com-plex. Hunting and gathering, and perhaps the use of shellfish, canbe inferred. No evidence of fishing was found; but bone fishhookshave been recorded from the immediate locality and I venture the guessthey will eventually prove to be associated with the same culture as isrepresented by the Griflmg site. As to the time of occupancy, it issafe to suggest that it was in pre-White days ; I suspect, though, thatit also predated any Kansa occupancy of the locality. The physicalappearance and relationships of the people remains unknown, as doestheir method of disposal of the dead, but perhaps some form of moundburial can be suspected. Further work at this and related sites nearbyought to extend considerably the range of artifact types and othertraits characteristic of the culture, and will thus make possible a muchbetter cultural and chronological correlation with other prehistoricculture complexes of the region.KANSA VILLAGE SITE (14P024), POTTAWATOMIE COUNTYKansas River and the Big Blue are both near grade and flow throughtheir respective alluvium-floored valleys in meandering fashion.Northeast of Manhattan, a wide sweeping curve of the Big Blue tlirowsthe stream against its eastern bluff line until it is within half a mile ofthe Kansas. Here it turns sharply west for about a mile, then turns 188 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174again to the south and east to its junction with the Kansas. Formerly,its westward direction continued for a mile farther, and its entryinto the Kansas occurred at the east edge of Manhattan, more than amile above its present mouth. At the narrow neck where the twostreams most closely approached each other before their final junction,and on the north bank of the Kansas 2 miles east of Manliattan, iswhat remains of the principal village of the Kansa during, roughly,the first third of the 19th century.The general appearance of the village during its occupancy is welldescribed by Say (James, 1823, vol. 1, pp. 120-130), who estimatedthe number of lodges in 1819 at about 120. Its population seems tohave been in the neighborhood of 1,500 persons. In 1880, the site wassurveyed by the Kansas State Historical Society, and a map, on whichfigure 29 is based, was prepared. At that time (Anon., 1881, p. 287) : . . . the village stretched from the banks of the Kansas river northward for thegreater part of the distance towards the Blue. . . . The situations of the lodgesare yet plainly marked by circular ridges and depressions, ranging in diameterfrom less than 10 feet to more than 50 feet. These, numbering about 160, exclu-sive of those which are quite small, were accurately measured and located.A good many changes have taken place at the spot since it was aban-doned by the Indians, and some have been of drastic nature. The sitelies on a fine terrace rising some 30 feet above the river, and droppingoff a few hundred yards to the east to a lower bottom across which Sayand his party approached the town in 1819. Lateral erosion by theriver has destroyed much of the site ; what remains has been cut up bythe Union Pacific railroad, by an evergreen nursery immediately northof the railroad right-of-waj^, and by years of farming. We were in-formed that the river has cut its north bank back by at least 30 or 40rods within the memory of living persons ; and a curving swale thatmay be the old line of the bend is shown on the 1950 edition of theManhattan topographic quadrangle 400 yards or more south of, butwithin, the present horseshoe bend of the stream at this point. Unfor-tunately, the 1880 map prepared by the Kansas Historical Society in-cludes no landmarks or other points by which correlations can be madewith our partial map of the site, and so the actual extent of streamerosion cannot now be precisely determined.At the time of our investigations, almost no surface traces of theold village were in evidence. We were told that much material, in-cluding iron, catlinite, stone, bone, etc., had been turned up in thenursery grounds, in the cultivated fields just to the west, and in thenarrow triangular field between the railroad and the river bank. Inthe nursery area we were able finally to locate two pits containingrefuse from the native occupation, and these we excavated. Tests inthe field immediately to the west revealed areas with refuse-mixed Wedkl] AN INTRODUCTIOlSr TO KANSAS ARCHEOLOGY 189 ?'*fV,oo.o o oO .. ?3 o op ? o '.-.;O O O O ^?-rPo ?"OqOo ^ ?'' ?e?OOe I'^l^'i /r^^ /V5 /I 5 RIVER Figure 29.?Map of Kansa village site, 14P024, east of Manhattan, between Kansas andBlue Rivers. Large stippled circles, house depressions; small circles, cache pits(?); smallsquares, graves(?). Based on Kansas State Historical Society survey, 1880. 190 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174 soil and traces of what were possibly two house sites. The cut bankof the river showed disappointingly little material; but in the westend of the field south of the railroad, where the stream came closestto the right-of-way, a house site was finally located and excavated.This house lay just east of the point where the river intersects theline between sections 9 and 10. HOUSE 1The house floor we uncovered here was circular in plan and 28 to29 feet in diameter (fig. 30), with a well-defined line separating thehouse fill from the undisturbed surrounding soil. At the center, 18inches below ground surface, was a well-marked fireplace approxi-mately 30 inches in diameter, filled with hard-packed ashes and under-lain by fire-reddened earth to a depth of nearly 9 inches. A secondfireplace, 24 inches in diameter, lay 8 feet to the north, partly over-lapping the edge of an earlier, probably prehouse, basin measuring7i/^ by 4 feet. At a radius ranging between 66 and 78 inches fromthe hearth center were four postholes, each 10 to 12 inches in diameterby 21 to 24 inches deep, and forming an approximate 8 to 9 foot squareabout the hearth. Nine smaller holes, each 4 to 7 inches in diameterby 8 to 15 inches deep, lay farther away, at a radius of 10 to 12 feetfrom the fireplace. The entrance passage, about 5 feet wide, openedtoward the east, being readily traceable for 12 feet beyond the house \\ m\^ 12 4- FEETFigure 30.?Floor plan of Kansa earth-lodge site, house 1, near Manhattan (14P024).^ Solid circles indicate postholes; rayed circles, fireplaces; stippling, prehouse basin or ^_ cache; broken line, approximate edge of house. Wedel] an rNTRODUCTION TO KANSAS ARCHEOLOGY 191 wall. Of the posts that must have supported the walls and roof ofthe passage, the only evidence was two postholes near the house wall.Beyond the end of the passage was a heavily mixed area, which con-tinued some 12 inches below the level of the entrance (pi. 16, a) . There were no cache pits within the house, unless a small pit nearthe northwest wall be so considered. This measured 10 by 16 inches,with a depth of 10 inches ; its fill contained quantities of charred corn-cob fragments, bark, and grass, as did that in the two west mainpostholes. On the house floor, in the basin at the north side, and inthe fill immediately above the floor, were quantities of broken animalbone, as well as gun parts, steel-trap fragments, musselshells, and afew native-made artifacts, but no sherds.Two burned fragments of the nest of a mud dauber wasp were alsofound among the burned house daub on the floor. According toK. V. Krombein, United States Department of Agriculture, theyprobably represent Sceliphron caementarium, the common muddauber of the United States.The general plan of the house floor and the arrangement of postsresembles that found in historic Pawnee sites and strongly suggestsan earth lodge. That the structure which once stood here was prob-ably earth covered is also suggested by the presence of small chunksof fire-hardened grass-impressed clay. The presence of this structureis thus in line with the observations by Say that the Kansa weredwelling in earth lodges, rather than in bark-covered or mat-coveredstructures, when he visited the tribe in 1819.CACHE PITSIn the grounds of the Kansas Evergreen Nursery north of therailroad we examined several holes left in the recent removal ofnursery stock. In and around one were noted bone fragments andpieces of heavily oxidized iron, brass, and other contact materials.Further tests disclosed the presence of two undisturbed pits nearbywhich we excavated. These lay about 65 and 80 yards north of therailroad. Pit 1 measured 42 by 48 inches at the surface, was 48 inchesdeep, and had a floor diameter of 69 inches. It contained a very darkfill heavily mixed with ashes, charcoal, some animal bone, iron, andbrass fragments, but without pottery. The second pit, whose meas-urements I somehow neglected to record, was smaller but had a similarcontent of bones, iron, and brass but no native artifacts or pottery.BUBIALSSome 350 to 400 yards east of house 1, where the village terracedrops away to a lower level, recent cultivation had evidently dis-turbed a grave or burial area. Scattered over the freshly tilled sur-face were many fragments of human bone, all so badly shattered as4?4172?59 14 192 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174to be not worth salvaging. There were also glass beads, copper orbrass buttons, metal-lace fragments, shreds of cloth, and other bitsof former garments. All of this came to light, of course, just at theend of our stay; and in the hour or two we devoted to a search forgraves none was found. A small sample of the cultural materialwas preserved, however, and is described elsewhere in this paper. Itseems possible that further investigations carried out in more leisurelyfashion than were ours would turn up intact or less completely de-molished graves that could be properly worked out and recorded.FAUNAL BEMAIN8Broken and shattered animal bones occurred in some quantity inthe house site and in the pits. Most of these were apparently frombison. Among the less fragmentary materials brought in to theNational Museum were identified the following species, listed inorder of abundance of bones present : White-tailed deer {Odocoileus virginianua) 18Black bear (Euarctos americanus) 2Horse (Equus cahallus) 1Bison {Bison bison) 1Puma {Fells concolor) 1Beaver {Castor canadensis) 1Pocket gopher {Oeomys bursarius) 1Raccoon {Procyon lotor) 1Birds are represented by two bones of the wild turkey {Meleagrisffallopavo). MOIXUSCAN REMAINSThe unworked shells of fresh-water mussels were surprisinglyplentiful everywhere in our excavation, and fragments were scatteredgenerously throughout the plowed soil wherever we made tests. Theabsence of workmanship on the shells, of any shell artifacts, or ofpottery that might have required shell tempering, makes it seemlikely that the mussels were gathered primarily for food. Probablymost or all of the species represented were readily available in theKansas or Big Blue Rivers near the village, or in some of their smallertributaries nearby. Listed in order of their abundance, the followingspecies are represented : Lampsilis ventricosa occidens (Lea) 10Obovaria olivaria (Raf) 3Amblema costata (Raf) 2Pleurobema coccineum solida (Lea) 1Fusconia flava (Raf) 2Quadrula pustulosa prasina (Conrad) 1Proptera laevissima (Lea) 1 Wedbl] an introduction TO KANSAS ARCHEOLOGY l93VEGETAL REMAINSNo identifiable vegetal materials were returned to the NationalMuseum. The only such items noted in the excavations were charredcorncob fragments, bark, and grass from features already noted inhouse 1. ARTIFACTS OF NATIVE MANUFACTUBEObjects of native manufacture were few in number, limited invariety, and of practically no usefulness in tracing relationships withcontemporary or earlier native cultural manifestations. Nowhere onthe site did we find any Indian pottery nor did we learn of any everhaving been found here, either on or under the surface. Workedstone and bone occurred, but they were scarce. Our limited samplingsuggests that the Kansa by the first quarter of the 19th century hadgiven up most of their native material culture and were relyinglargely on the American traders to supply their needs for tools,weapons, and utensils. I have a strong impression that aboriginalmanufactures are proportionately much scarcer on this site than theyare on contemporary Pawnee, Arikara, and other sites of the sameperiod farther to the north. This is, after all, what might be expectedof a tribe that had been for well over half a century in direct contactwith white traders and that, for a considerable time, resided on ornear a major trade artery of the region.Work in bone includes two piercing tools and a possible gamingpiece. Of the first, one is the fibula of a puma {Felis concolor),from which the proximal extremity has been detached and to whicha short, thick, well polished point has been given; the distal endshows no modification. The skewerlike object thus produced (pi.18, a) is 20 cm. long. A shorter heavier implement, made from theshaft of an unidentified mammal bone, is pointed at one end andpierced at the other (pi. 18, &). Suggesting a sewing tool, it is 14.4cm. long and is well polished. The third piece is the toe bone ofa deer, dressed to a conical form and with both articular surfacesremoved. Just above the base of the cone is a 4-mm. perforation ; theentire surface and the tip are well smoothed (pi. 18, c). This maybe one of the pieces from a ring-and-pin game set, but the reason forthe perforation is not at all clear.A deer antler artifact consists of a 70-mm. segment of tine, withthe tip broken off and the other end cleanly cut off with, apparently,a metal saw. The cut end has been smoothed, and the cancelloustissue has been gouged out to produce a cavity 12 mm. in diameterby 10 mm. deep. The piece somewhat suggests a handle, but the socketseems too shallow and broad to have been very serviceable in seatinga cutting, scraping, or piercing element. 194 BTJREAU OF AMERICAN ETHNOLOGY [Bull. 174The only fired-clay specimen we found was a rectangular blockabout the size and shape of a domino and evidently molded with somecare (pi. 18, e) . It measures 48 by 25 by 12 mm. and lacks any mark-ings that would aid in identifying its use.Worked stone was somewhat more plentiful but not much moredefinitive. Two rectangular blocks of sandstone suggest whetstonesfor sharpening metal tools. The smaller, measuring 50 by 31 by 14mm. and broken at both ends, has a rectangular cross section andworn surfaces. The larger is 80 by 52 by 23 mm. ; it is deeply wornand smoothed aromid the middle, and less so at the ends. Also ofgray sandstone is a small pipe bowl, represented by two fragments25 mm. long from the junction of stem and bowl bores. The bowlappears to have been slightly constricted at the middle and was ap-parently fitted with a stem of wood or other soft material ; its inter-ior is fire blackened and has traces of a "cake" at several points.Catlinite is represented by five worked pieces. Four of these, in-cluding a small pipe fragment not further classifiable as to form,are from house 1 ; except the pipe fragment, these look like rejectage.A fifth and larger lump has one cut surface and is otherwise polishedas though from long carrying in a bag. Most of these pieces seem tobe of fairly high quality stone, and doubtless originated in the Min-nesota pipestone quarries.Only two pieces of chipped stone were found. One is small, thin,and very clumsily worked; it has a stem, two broad barbs, and ablunt tip. Possibly a crude projectile point is indicated, but if soit would seem to have been a very ineffective tool. The second pieceis apparently a chopping or scraping implement. It has a curvedunifacially chipped cutting or scraping edge 70 mm. long, abovewhich are two broad shallow notches for hafting; above these isthe thick unmodified back.A small quartzite pebble, apparently dressed to a round flattishshape, measures 32 by 42 mm. ; its purpose is unknown.OBJECTS OF WHITE MANUFACTUEEObjects of White manufacture occurred almost altogether in a frag-mentary state and in poor condition. They indicate, however, thatfirearms, steel traps, cutlery, household utensils, trade textiles, and ar-ticles of personal adornment were present ; and some of these items, itwould seem, must have been fairly plentiful. Although no maker'smarks have been found on the specimens, all appear to have been ma-terials characteristic of the period for which Kansa occupancy of thesite can be documented or from a slightly earlier time.The presence of firearms is attested by 7 gun fragments and 3 gunflints. Three lockplates (pi. 19, /) from house 1 range in length from132 to 150 mm. The smallest of these still carries the frizzen ; and a Wedbl] an INTRODUCnOlSr TO KANSAS ARCHEOLOGY 195gooseneck cock, or hammer, found nearby is very likely a part of thesame mechanism. This probably came from a pistol of pre-1800manufacture. The other two lockplates are smaller than those frommilitary weapons of the period; they may be from light muskets orrifles intended for civilian use or the Indian trade, or less probablyare also from pistols. Four sections of gun barrels vary in lengthfrom 50 to 150 mm., and two of these have been hammered flat. Allappear to be of light stock, suggesting pistol barrels rather than mus-kets. The gunflints, also from house 1, are of very dark gray Englishflint ; all are flat and tabletlike, nearly square, and none exceeds 21 mm.in maximiun dimension by 7 mm. in thickness. They were probablydesigned for pistols, but were usable in light muskets as well.Cutlery is represented by five knife-blade fragments and two incom-plete handle elements of bone. What was probably a butcher knifehas an 18-cm. curved and tapered blade and a 5-cm. tang. Anotherfragment, 13.5 cm. long, may be from a table knife, as are two othersmaller pieces. One of the latter (pi. 19, e) consists of the basal partof the blade and adjoining tang with ferrule; the other is the tang be-low the ferrule, with three transverse rivets and bits of the woodenhandle still adhering. Still another fragmentary table knife from pit3 included part of the blade, ferrule, and tang. Of the two knifehandles represented, one is a tubular piece, stained light greenishinside and out, highly polished on the exterior surface, and with deco-rative beading at the basal end (pi. 18, d). The second is a shortsliver from one side of a handle, beveled along both edges, and showingone rivet hole 8 mm. from the slightly hollowed cut end.Other objects of iron included part of the spring of a large trap (pi.19, d) and the pillar of another. A third piece is either another springfragment or a heavy latch bar. From house 1 came the handle socketsof two axes, made by bending a heavy iron band into a circular or el-liptical shape and hammering the ends together to form the blade,which is missing from both our specimens (pi. 19, a) . The house sitealso yielded two arrowpoints cut from sheet iron. One is a plaintriangle, with the tip broken off; it measures 40 by 20 mm. Theother has a short square stem and sloping shoulders ; including thestem, it measures 44 by 23 mm.Brass is present chiefly in the form of miscellaneous scraps of sheet,most of them with perforations, rivets, rivet holes, or thin cuts andslashes suggesting metal chisel marks. There is also the bail ear froma large kettle, to which the ear was secured with two iron rivets (pi.19, &). The face of a two-piece brass button, 22 mm. in diameter andwith crimped perimeter, was found in house 1 ; traces of the originalwood( ?) backing still adhere to the reverse side.A pointed strip of lead, bent into a hook, is apparently a piece ofscrap metal. 196 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174Earthenware includes five pieces. Two of these are small redwaresherds with a greenish cast and high glaze on the interior surface.According to C. M. Watkins, associate curator of ethnology, UnitedStates National Museum, they are of a type manufactured over a longperiod, probably including the first quarter of the 19th century. Thereare also two stem fragments of white earthenware pipes, as well as afragment showing part of the stem and bowl, with a prominent down-pointing spur on the latter (pi. 18, /) . Beads from house 1 include a short cylindrical glass specimen, 3by 3 mm., and two of tubular shell ( ? ) , 3 mm. in diameter and 8 and17 mm. long.In addition to the Euro-American items described above from thehouse and pits, we obtained similar material from the surface of thedisturbed burial area ; but here no native artifacts were noted. Sevenbrass buttons are all of one-piece style, flat and fairly heavy, and witha stout wire loop soldered to the back of each. They occurred in twosizes. The smaller ones, 3 in number, averaged 19 to 20 mm. indiameter. Removal of the oxide coating from the reverse of oneshowed a depressed band in which raised letters spell out the wordsBEST and COLOUR, each preceded and followed by a 6-pointed star.One of these buttons has a bit of dark woolen fabric still caught inits loop; another was evidently attached to a strip of metal lace.The four larger buttons are 24 mm. in diameter. On the reverse, 2lines encircle the loop ; between these and the outer edge is stampedthe legend RICH GOLD COLOUR, with a conventionalized floralmotif at the bottom (pi. 18, i) . "V\niat appears to be a brass knee buckle lacks the tongue and isotherwise undistinguished; it measures 30 by 37 mm. (pi. 18, k).Beads include a large spherical red specimen with white core, anddiameter of 17 mm. ; a small blue, 3 mm. in diameter ; and 9 of shell,tubular, 3 mm. in diameter by 6 to 9 mm. long.Perishable materials include several scraps of brown and black, ordark blue ( ? ) , woolen goods and three small pieces of metal lace. Thewoolen fabric is a napped plain or cotton weave (over-and-under),evidently machine made, and with most of the surfaces obscured bythe matted nap. In one instance, two pieces of different color havebeen stitched together with cotton thread ; and I suspect that the brownserved as the lining for a dark-blue or black garment.The best preserved piece of lace is a strip 20 mm. wide. It consistsof about 20 cotton warps with which were interwoven fine copperwires used in pairs. The weave is basically an over-two-under-two ; but it is varied by triangular areas in which the copper wire weftscross from 3 to 7 warps before resuming the usual weave. Traces ofthe cotton thread by which this strip was secured to a woolen shirt or Wbdbl] an introduction TO KANSAS ARCHEOLOGY 197jacket are still visible. Two smaller lace fragments are apparentlypart of a similar fabric ; but they lack finished edges and are probablyfrom a wider strip. This lace (pi. 18, j) would seem to have been afiner and probably more costly goods than the copper strip-wire-cottonlace I have elsewhere described from the early 19th century Leaven-worth (39C09) Arikara site near Mobridge, S. Dak. (Wedel, 1955,p. 146). GENERAL COMMENTSFrom our point of view, the excavations here at site 14P024 canonly be characterized as disappointing. As an attempt to apply thedirect-historical approach to the archeology of the Kansa Indians, weran into a dead end. It had been our hope, of course, that the sitewould yield a sample of native pottery and other artifacts of Indianorigin in sufficient quantity and variety to suggest relationships withsome of the earlier sites in northeastern Kansas. A definable complexof this sort obtained at a datable site whose tribal authorship is beyondquestion might have pointed the way to identification of a Kansasequence. As the foregoing pages show, however, we failed to findsuch a complex of aboriginal traits ; and in view of the obvious extentof White influence here, I am pessimistic about such a complex everbeing defined at this site. Our work here, it is true, was extremelylimited; and perhaps sustained observations and more extended dig-ging would still turn up helpful leads to earlier phases of Kansa cul-ture. For the present, we can only remark that, on the basis of arch-eological evidence, the Kansa at this point in history were living, asSay noted, in a large and fairly compact village of earth lodges, thatthey were following a subsistence economy based on maize agricultureand on hunting, and that they were heavily under the influence ofWliite trade. In these respects, they had much in common with othervillage tribes of the eastern Plains ; but their native material culturewas apparently even more completely shattered than was that of thePawnee and Arikara. I see nothing in what we found that wouldrelate the local complex at 14P024 to the earlier Doniphan and Fan-ning sites along the Missouri to the northeast, or to any other knownpre-1800 tribal complex in the region.OTHER SITES AND MATERIALS IN THE MANHATTAN AREAIn addition to the limited excavations above described at the Grif-fing and Kansa Village sites, we made brief tests at several othersnearby ; and members of our party visited a number of other villageand camp sites reported by various individuals in Riley, Pottawatomie,Geary, and Wabaunsee Comities. None of these observations yieldedvery much artifact material ; but what they did produce strongly sug- 198 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174gests that additional research at certain locations ought to be "wellworth the effort.We also examined a large collection in possession of Dr. N. L. Rob-erts, now deceased, of Manhattan. This included a great many graychert artifacts of various forms illustrating the core industry of theKansas River valley, together with small samples of pottery fromvarious village sites and other materials obtained in mound excava-tions. Practically all of this was collected, according to Dr. Roberts,at sites within 10 or 12 miles of Manhattan?most of it from the BlueRiver valley below Juniata Crossing, from Wildcat Creek below Keats,and from Deep Creek near Zeandale. With exception of a few villagesite specimens and small lots from three burial mounds, Roberts'material was not identified as to site provenience. Through the kind-ness of Mrs. Roberts, a representative series of specimens from thiscollection was subsequently presented to the national collection.Though the observations mentioned in the foregoing paragraphscannot be regarded as definitive, they include some that are certainlysuggestive, and therefore merit some further comment here.SITE 14P025 (BBOUS)Lying approximately half a mile west of the Kansa Village site(14P024), this is in all likelihood the location reported by Brower(1898, p. 61, and 1899, p. vi and map) as the Brous site. It is situatedon the same Kansas River terrace as 14P024, about 30 feet above theriver channel and some distance from the present bank. Accordingto Brower (1898, p. 61), this was ? the site of an ancient village, which had been submerged by the overflow of theBig Blue River. A cut on the Union Pacific Railroad through the site exposedthe flint implements and pieces of pottery which lie about 3 feet below the pres-ent surface of the ground, covered with sedimentary deposits by overflowage.Very little cultural material was visible on the cultivated groundsurface. We dug a series of small test pits at random immediatelynorth of the railroad right-of-way, and about 100 yards from a UnitedStates Coast and Geodetic Survey benchmark. No. G-115, elevation1,013.931 feet, m. s. 1. These disclosed a dark-gray topsoil to about a24-inch depth, the color becoming somewhat lighter with increasingdepth. At 24 inches, there was a light gray stratum 6 to 8 inches inthickness, in which bits of charcoal and burned clay were detected. Inone test pit, this layer yielded also flint chips and a few potsherds;elsewhere, other flints and a large stemmed point were found at thislevel. In general, the tests yielded discouragingly little cultural ma-terial. Most of it, however, came from the 24-32-inch level, belowwhich was only clean undisturbed subsoil ; and this, apparently, is themain archeological zone. Its areal extent was not determined, but itmay be as much as 2 or 3 acres, possibly somewhat more. The over- Wbdbl] an introduction TO KANSAS ARCHEOLOGY 199lying soils are probably, as Brower suggested, sedimentary materialdeposited by overflow from the Big Blue, the Kansas, or both.Artifacts from the site include 32 potsherds and 5 stone objects.Many of the sherds are weathered, so that their original surface finishis uncertain; but no cord-roughened pieces can be recognized. Allare tempered with rounded sand particles, usually abundantly presentand of moderately coarse texture ; several have flecks of mica showingon their surfaces. The sherds tend to crumble rather easily. Colorvaries from brown to gray, the paste core usually being considerablydarker than in Kiley Cord-roughened and sometimes nearly black.Hardness seldom exceeds 3 to 3.5. Plain exterior surfaces are charac-teristic of the sherds ; but at least 5 have distinct impressions from anedentate rocker (fig. 31, h). One weathered rimsherd bears on itsconvex outer surface crossed lines that may be dentate stamp impres-sions, below which are traces of a row of punctates. Three otherrimsherds appear to be from bowls. Another sherd has a row ofpunctates at what was evidently the neck of the vessel. All of thesefeatures set the sherd sample apart from the Riley Cord-roughenedpottery found on most local sites, but not noted by us here at 14P025 : and they suggest a Hopewellian tradition or complex.Chipped stone includes one large stemmed projectile point, coarselychipped, from which the extreme tip has been broken (fig. 31, a) . Thestem expands and the base is straight ; the piece measures 63 by 30 by8 mm., and weighs 18 gr. Another piece, roughly triangular in outline,with convexly curved edges and base, measured 75 by 35 by 12 mm. ; itmay be a knife or a projectile point blank. There is a basal fragmentof a second similar piece. Two other coarsely flaked fragments are therounded ends from large blades or blanks.A single piece of ground stone was recovered. This is a small frag-ment of igneous material, with one rounded edge and a smoothly hol- aFigure 31.?Stemmed projectile point and rocker-marked potsherd from Brous site, 14P025,east of Manhattan (USNM 381670, 381671). 200 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174lowed surface. It looks like a sliver from the end of a small grindingor polishing slab approximately 30 mm. thick.The cultural affiliations of this material would seem to be not withthe numerous local village sites on which Riley Cord-roughened pot-tery predominates, but with a Hopewellian or Hopewellian-influencedcomplex probably related to certain sites in the Kansas City locality(Wedel, 1943). Possibly further search on the spot would discloseareas from which a larger and more definitive sample than ours couldbe recovered. In any case, that the site is apparently not unique inthis locality is suggested by our findings on the lower Vermillion Riversome 18 miles to the east. BITE 14P026 (DIKE)Following again in Brower's footsteps, members of our party sur-face-hunted a section of the Vermillion River valley east of Wamego.Here, according to Brower (1898, p. 67), on "a regularly formed ter-race, the Dike Village Site, was discovered, where potsherds and flintimplements were scattered over the surface of the cultivated field fornearly half a mile." Elsewhere (Brower, 1899, p. 103), he figures avessel, almost certainly a restoration, in which the rim decoration ofcrosshatched lines with punctates on the neck below, is reminiscent ofHopewellian pottery. From Brower's text it is not clear whether theDike site was east or west of the Vermillion, but the only site shown inthis locality on his maps is on the east side. Our search yielded noth-ing on the east bank; but on the west side, about 1% miles north ofKansas River, an area with sherds and flints was discovered. To this,located as is Brower's Dike site in section 31, T9S, RUE, is here as-signed the designation 14P026.We did no digging here and only surface-collected materials are athand. They include 13 sherds and about the same number of workedstone specimens. Three sherds, including two rims, are of Riley Cord-roughened type. The other 10 consist of 7 plainware sand-temperedbody sherds ; 2 body sherds with rocker impressions ; and one rimsherdthickly tempered with angular particles of hornblende (from crushedgranite ? ) and bearing on its convex outer surface a series of verticalrocker impressions below which the neck carries a row of crescenticpunctates (fig. 32, a).Chipped flints, like the pottery, suggest two traditions. There aretwo well-made symmetrical projectile points of unnotched triangularform, each with slightly concave base and convex lateral edges; theymeasure 27 by 15 by 3 mm., and weigh 1 gr. each. Much more commonare large stemmed points. Of these, there are six whole, or nearlywhole, specimens and four basal fragments of the same type. Theseall have an expanding stem with straight or convex base ; all are thick,heavy, and coarsely chipped (fig. 32, c-e). The complete or nearly WkdelI an introduction TO KANSAS ARCHEOLOGY 201 Figure 32.?Rocker-marked rimsherd, chipped flint disk, and projectile points from Dike(?)site, 14P026, east of Wamego (USNM 381694, 381695).complete specimens range in length from 43 to 65 mm., in width from20 to 30 mm., and in thickness from 5 to 12 mm. ; their weight variesfrom 5 to 15 gr., but averages approximately 13 gT. The size andweight range of these larger pieces is well within that represented bythe projectile points characteristic of the Kenner site near Kansas City,and the form is generally substantially the same also (Wedel, 1943, p.50 and pi. 12).Two chipped-flint disks are also reminiscent of the Renner site.One is nearly circular in outline (fig. 32, h), somewhat flattened incross section, well made and with the edge retouched all around; itmeasures 46 mm. in diameter by 16 mm. in thickness. The second issmaller, slightly longer than wide, and part of the edge is not re-touched ; it measures 47 by 35 by 16 mm. 202 BUREAU OF AMERICAX ETHNOLOGY [Bull. 174Other flint objects include two knives, both subtriangular in shape,with straight to slightly convex base, convex edges, and one pointedend ; both are coarsely flaked on both faces. They are 74 by 35 by 16mm. and 70 by 45 by 19 mm. in size. A planoconvex end scraper isfashioned from a large curved flake 75 by 35 by 17 mm.; the edgesare irregular and indifferently retouched.A piece of dull red stone suggesting catlinite, but probably a veryfine-grained sandstone, has the marks of pecking over much of thesurface and the edges have been ground to produce a celtlike form.It measures 115 by 63 by 25 mm.Here again our sample is too small to permit positive identificationof the site as to its cultural affiliations ; and a further element of un-certainty is introduced by the fact that all of the material was collectedfrom the surface. To me, the material suggests two complexes ofdifferent age. The Riley Cord-roughened sherds and possibly the twosmall unnotched projectile points affiliate with the rectangular housecomplex characteristic of the locality. The majority of the sherds,the large stemmed points, the chipped disks, and perhaps other itemsare reminiscent of Hopewellian materials. Wliether they representan actual occupation by Hopewellian peoples or, alternatively, reflectHopewellian influences on a native population with other antecedents,must be left for future more extended investigations. It would not beat all surprising if systematic search in this section disclosed thepresence of Hopewellian village sites with close relationships to thosenow known in the Kansas City locality slightly more than 100 miles tothe east down the Kansas River (Wedel, 1943).The likelihood seems strong that a third ceramic complex occurs inthe vicinity of Manhattan, perhaps primarily in sites along the lowervalley of Big Blue River. Brower (1898, p. 65) figures a small-mouthed squat-bodied vessel with two handles, and bearing incisedor trailed and punctate decoration, from the Hill site not far below themouth of Carnahan Creek. This is markedly unlike any of the potterywe found or saw in the area, and the decoration suggests Oneota orOneota-influenced pottery. Further search could perhaps profitablybe undertaken at this site. Brower also mentions (1898, p. 61) a sitenear Rocky Ford where the people "made catlinite pipes and earthenvessels, evidences of which have been catalogued." Farther down-stream, he noted "a considerable group of lodge-circles and two ormore artificial mounds" on the Richards farm; and he stated that"The circles or hut-rings are very distinct though of much greater agethan those at the Kaw village, three miles down the river." At thispoint, on a narrow apron at the base of the east bluffs of the Big Blue,we saw a shallow house ring some 30 feet in diameter and a fewmounds; but we were refused permission to do any testing or to Wedel] an introduction TO KANSAS ARCHEOLOGY 203examine materials said to be from the spot. If in one or more of theseseemingly late villages the existence of an incised potteryware couldbe established, the question of its possible relationship to Oneota ma-terials on the Missouri and to a late prehistoric or protohistoric KansaIndian occupancy would present a challenging and interestingproblem. MOUND MATERIALSOf the numerous mounds and cairns that formerly occurred at manypoints on the Big Blue and Kansas Eiver bluffs, most have apparentlybeen destroyed and few still remain untouched. We opened one sup-posed cairn, with negative results that merit no further commenthere. From such published accounts as exist, and from the oral reportsgiven us, it would appear that some of these structures were perhapserected in historic times and yielded exclusively materials from theperiod of White contact. Others were evidently older and perhapsoriginated at various times and from several cultural levels in the pre-historic period, though it seems likely that historic Indians not infre-quently buried intrusively in such structures. The scant and notaltogether satisfactory data to be presented briefly here were furnishedby the already mentioned Dr. Eoberts, of Manhattan, and concern onlymounds opened by him personally. The artifacts recovered are inthe national collections.Fremont Point Mound No. 1Largest of the mounds opened by Roberts was on a spur of the bluffssouth of Kansas River, known locally as Fremont Point, about 21^miles southeast of Manhattan and some 200 feet above the valleyfloor. The mound is said to have been 10 or 12 feet in diameter andabout 2 feet high. Just beneath its surface was an irregular and dis-continuous layer of limestone slabs and blocks. A circular pit about6 feet in diameter had been dug to a depth of 3 feet below the originalground surface, and this was lined with a rough dry-masonry wall.The earth within the pit and immediately above contained numerousfragmentary human bones, some of which had been scorched and fireblackened. There were no articulated burials and no complete bones.Scattered among the bone fragments and throughout the fill were boneand shell beads, chert arrowpoints, and two shell pendants. No glass,metal, or other White contact material was found. The general im-pression of the excavator was one of complete disorder in the originaldeposition of the human and other materials recovered. No estimatewas made of the probable number of individuals represented ; but Dr.Roberts, a dentist by profession, said he observed more than 300human teeth among the remains.Tubular bone beads to the number of slightly over 800 were recov-ered, these being by far the most common artifact type in the mound. 204 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174Many have been discolored by heat to a black, blue-gray, or white, andbroken or split pieces are common. They were apparently made frombird or small mammal bones by the simple procedure of detaching theends and rubbing down the cut. They vary widely (pi. 20) in size,with diameter ranging from 2 to 10 mm. and length from 6 to 49 mm. ; but 60 percent are less than 3.5 mm. in diameter and 70 percent areunder 16 mm. in length. Decoration occurs on about 365 specimens(48 percent). It consists of 1 to 12 incised lines encircling the tube;these are sometimes very shallow, at other times deep enough to reachthe interior cavity of the bone. In many cases, they are not continuousbut consist of a series of 3 short cuts or nicks whose ends may overlapslightly but usually do not quite meet. Some specimens are cut soregularly and deeply as to suggest an attempt to divide the tube into anumber of smaller segments, but there are no beads as small as thesesegments usually are. In only two instances is the incising in a con-tinuous line to form a spiral.Shell objects are next in order of abundance. There are approxi-mately 180 disk beads (pi. 20). They vary in diameter from 7 to 12mm. and in thickness from 1.5 to 5.0 mm. ; the central perforation isfrom 3.5 to 5.0 mm. in diameter. Many of these are fire darkened andcrumbly, and some of the thinner specimens probably result fromsplitting. All are probably of fresh-water mussel shell. The onlymarine species represented is a single shell of Marginella apicinaMenke, which has had the spire ground away to permit stringing.This species is native to the waters from North Carolina to the Gulf ofMexico and the West Indies. There is also one short thick barrel-shaped shell bead measuring 8 mm. in diameter by 12 mm. long.Two pendants were apparently also made of fresh-water shell.Both are elliptical in outline, approximately 35 mm. long by 17 mm.wide, and are perforated at each end. In the heavier specimen, whichis nearly 7 mm. thick, the perforations run obliquely from one slightlyconcave surface to emerge at the ends (pi. 20) . Projectile points number 35 ; they represent a variety of gray, whit-ish, pink and other cherts, and a rather curious range in forms (pi. 20) . Eleven have a basically triangular outline, with or without side and/orbasal notches. In terms of the shapes defined by Strong (1935, p. 88),these break down further into NBa (2 sp), NBal (3 sp), NBa2 (1 sp),NBa3 (3 sp), and NBa4 (2 sp). Workmanship is variable, but sev-eral are very well made. In size, they have the following range : length16 to 35 mm., mean 24.5 mm.; width 11 to 15 mm., mean 13.4 mm.;thickness 2.5 to 4.0 mm., mean 3.1 mm. The largest weighs just imder2 gr. ; combined weight of the 11 is between 9 and 10 gr.Quite different are 24 stemmed points (pi. 20), all of them small,slender, and well made. The stem results from corner-notching and Wkdbl) an LNTRODUCriON TO KANSAS ARCHEOLOGY 205thus expands toward the base. The base is distinctly convex in eightspecimens; in the others it varies from straight to slightly concave.The lateral edges are usually straight, and in about half the points fineserrations nin from the shoulders to the tip. Size range is as follows : length 18 to 34 mm., mean 25 mm. ; width 8 to 14 mm., mean 9.9 mm. ; thickness 2.5 to 4.0 mm., mean 3.3 mm. The largest stemmed pointweighs less than 1 gr., and the combined weight of all 24 specimensis 15.5 gr. Febmont point Modnd No. 2Approximately a mile southeast of the preceding mound, and alsoon the south bluffs of Kansas Kiver, a second and smaller moundwas opened by Dr. Eoberts. Beyond the observation that it wasvery small and inconspicuous, no details are available as to itsmanner of construction and internal arrangement. It is said tohave consisted almost entirely of earth, and to have contained notrace of human remains. Within the mound was found a smallquantity of red ocher and the worked canine tooth of a bear. Thelatter has been split and modified to such an extent that speciesidentification?as between black bear and grizzly?is no longerpossible. Most of its surface is well polished from use. Near thebasal end are two transverse incised lines. Two conical holes, 3.5 mm.in diameter and 12.0 mm. apart, have been drilled through the root.The specimen is 68.0 mm. long (pi. 20) . Modnd Near Rockt Ford, Pottawatomie CountyThe third mound excavated by Eoberts was situated on the blufftop east of Big Blue River, a mile or slightly more northeast ofRocky Ford. This was apparently on the ridge between the riverand Cedar Creek, not far from the east end of Tuttle Creek dam, anda short distance west of a stone cairn said to have been mentionedby Brower near the mouth of Cedar Creek. The mound was quitesmall?not over 5 or 6 feet in diameter by 1 or 2 feet high ; and it wascomposed almost entirely of earth, with only a few scattered stonesincluded. There was no evidence of a pit beneath the mound, norwas there any trace of human bone. A few small objects of metal,glass, bone, and shell lay at the base of the mound; and it was theopinion of the excavator that if a burial had been made here, it wasprobably similarly laid on the original ground surface and coveredover with earth.Glass beads included 27 pale blue and 3 white specimens, allweathered and faded, and none exceeding 3 mm. in diameter. Metalconsisted of 3 brass wire coils; the two larger, 30 mm. in diameter,consisted of li/^ turns each, whereas the smaller, 12 mm. in diameter,consisted of 5 turns. Shell objects included 12 disk beads, 8 to 9 mm. 206 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174in diameter by 1.5 to 5 mm. thick, with cylindrical perforations3.0 to 4.5 mm. across ; some of these were fire blackened. There wasalso a corner fragment of a pendant, with two finished edges and aperforation. Tubular bone beads numbered eight, equally dividedbetween plain and incised specimens; length varied from 8 to 23 mm.,diameter up to 3.5 mm. Two of the incised specimens have annularcuts; the other two were spirally incised, with successive turns from1 to 1.5 mm. apart. The base of a small stemmed point, with concavebase, and maximum shoulder width of 10 mm., completes the specimeninventory from the mound.1 am not prepared to say just where the three mounds sketchilyrecorded above will fit into the archeological sequence for the Man-hattan locality; but it seems fairly probable that they originatedfrom more than one period of occupation. If we had detailed descrip-tions of the numerous earlier mound explorations hereabouts, or ifthe collections then made and their associations had been preservedsomewhere for later study, patterns of some sort might be apparent.As matters stand now, the uncertainty concerning relationships ofthe mounds opened by Roberts may be regarded as more or lesssymptomatic of the general haziness that exists relative to the histoiyof native occupations in this locality. There are suggestive bits ofevidence, but little of a substantial nature into which one can gethis teeth.The Eocky Ford mound, apparently like some of the small onespreviously opened, suggests historic connections, at least in part;and perhaps it is to be linked with protohistoric Kansa or some stillunidentified tribe resident in, or temporarily inhabiting, the lowerBlue-Kansas River locality. The bone and shell beads and thestemmed point fragment are reminiscent of apparently earlier mate-rials of fairly wide distribution; but since they were found in suchsmall number at Rocky Ford, it is possible they were accidentally orincidentally interred at a relatively late time with the corpse of anindividual not actually associated with them in life. Several otherinterpretations will doubtless occur to the reader, but I am afraidthat none can be profitably evaluated at this late date.With exception of the stone-walled pit said to have underlain it,Fremont Point mound No. 1 appears to have been similar in mostparticulars to two larger mounds for which cross sections are figuredby Brower (1899, pp. Ill, 114). Both of these he calls "Harahey"burial mounds. One was near the Spring Branch village site on Kan-sas River, inferentially close to Manhattan ; the other was at or near theNudson village site on the Big Blue above Stockdale. Each appears tohave had a sort of mantle of stone beneath its surface, and below this afill containing burned and unburned human bones and artifacts. Tn- Wkdel] an introduction to KANSAS ARCHEOLOGY 207eluded in this fill at Spring Branch mound were thirty flint arrow-heads, not further described, a flint knife, fragments of a bone dagger,many shell and bird-bone beads, a spearhead, and an implement ofcatlinite ; and Brower notes further that "exploration of several othermounds yielded similar results." Elsewhere (Brower, 1898, p. 26;pi. 9, fig. 1 ; pi. 25) he illustrates plain and incised tubular bone beads,shell disk beads, and small stemmed and serrate points, apparentlyfrom various sites and all reminiscent of the materials described fromFremont Point mound 1. I get the impression that the bone and shellbeads, and probably also the small stemmed points, are types recurringrather frequently in and about the Kansas Eiver valley in Riley,Geary, Pottawatomie, and "Wabaunsee Counties. Whether or notthey occur in lodge sites associated with Riley Cord-roughened potteryI do not know ; but I have a suspicion they are perhaps mostly from anearlier period and will be found to correlate with one of the Woodlandculture variants still awaiting definition in the Central Plains.Fremont Point mound 1 materials, so far as they go, also show inter-esting similarities to a more extensive and varied series reported fromthe Younkin mound on the Republican near Jimction City, about 15miles west by south of Manhattan (Schultz and Spaulding, 1948;Spaulding, 1949, pp. 106-108). Here groups of fragmentary and oc-casionally burned human bones, an extended skeleton, animal bones,and artifacts were found on a rough stone floor mounded over withearth and rocks. Quantities of plain and incised tubular bone beads,shell disk beads, and small stemmed serrate points match closely thematerials from Fremont Point mound 1. In addition, there was pot-tery, a stone platform pipe, and other material with apparent Hope-wellian affinities, as well as long bone pins with expanded bases, cuthuman and animal bones, perforated and cut deer phalanges, andperhaps other items reminiscent of Archaic specimens. The possibilityof a Hopewellian mortuary complex superimposed on a local culturenot as yet defined but with Archaic affinities may be indicated. Kivett(1953, p. 134) suggests that some of the artifact types, includingstemmed serrate points, shell disk beads, and incised bird-bone beads,are found in Keith Focus Woodland sites, the shell beads being es-pecially abundant in burials of that complex. I have the impressionthat incised bone beads are much less common in Nebraska sites thanthey would seem to be, or to have been, in the burial mounds of the BigBlue-Kansas River locality ; but unfortunately, neither the publishedrecord nor unpublished data at my disposal shed much light on thismatter. I have a strong feeling, too, that when we know more aboutthe archeology of the Republican River valley in Kansas, from which agood deal of promising material is in the collection of the University484172?59 15 208 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174 of Kansas Museum of Natural History, this and other thorny problemsmay be considerably clarified (see, e. g., Eyman, MS) . As to cultural identity of the smaller mound I have designated Fre-mont Point mound No. 2, little can be said. The drilled bear tooth,of course, calls to mind the frequent occurrence of drilled and other-wise worked grizzly canines, or imitations thereof, in Hopewelliansites. Since this locality lies within the demonstrated range of influ-ence of Hopewellian culture, it is quite possible the mound in questionshould be assigned to a local manifestation of that culture, or to acomplex influenced thereby.GENERAL COMMENTS ON THE MANHATTAN LOCALITYThe results of our observations in the Manhattan locality may bebriefly summarized. Our data are admittedly sketchy and incom-plete ; but I believe they give some indication of the varied nature ofthe archeological remains here and may point up the potential signifi-cance of the locality for Central Plains prehistory. Surel}'' presentare ceramic materials from no less than three native culture horizons ; there were probably others. These date at least as far back as theWoodland period. Hopewellian traces suggest by their nature andoccurrence the presence of village or camp sites, some probably al-luvium covered, as well as burial moimds. A few thick, coarselygravel-tempered, deeply cord-impressed sherds that came to our noticewere reminiscent of Woodland ware like that of the Valley or KeithFoci in Nebraska; and the marked abundance of heavy stemmed orcorner-notched points in local collections would seem to imply thatthe Woodland and/or Hopewellian traditions may once have beenstrongly represented in the locality. Much more abundant and un-doubtedly from a later time are the prehistoric village sites of subrec-tangular earth lodges, representing a semihorticultural people whomade pottery of Riley Cord-roughened type. From a later peoplestill, presumably, are the as yet very scantily represented ceramic re-mains with incised or trailed and punctate decoration, whose possiblerelationships to sites reported to have yielded catlinite objects shouldcertainly be investigated. Such work would probably lead into theprotohistoric or historic period, and might point the way to furtherworthwhile operations on Kansa prehistory.With regard to the innumerable large chert implements whose cul-tural affiliations and period of manufacture have long been a puzzle, Ihave little to offer that might be helpful. We examined some of thesites described by Brower, including his "Elliott" site, and can givesome confirmation to his observation that these yield little or no ceram-ic material. That this is proof of a preceramic industry, as Winchellargued, I am by no means convinced ; but neither am I wholly satisfied Wbdel] AN INTRODUCTION TO KL\NSAS ARCHEOLOGY 209 at the moment that the materials can be dismissed as simply workshopdebris to be correlated with the ceramic horizons known or suspectedin the locality. Implements indistinguishable from forms commonlyfound on these largely potteryless chert sites evidently occur in localvillage site complexes, like the two chipped celts we recovered fromhouse 1 at the Griffing site; but in proportion to the extraordinary Figure 33.?Miscellaneous artifacts from general vicinity of Manhattan, a. Stone "mold"(USNM 386S58); J, dentate stamped and zone-decorated sherd (USNM 381702); c,corner-tang knife (USNM 381701), Deep Creek; .\\ \i: '--A ^ > .^- ^v^r:- Figure 43.-Trailed sherds from Tobias site, a, Pit 5 (USNM 388700). (Actual size.) Wbdbl] an introduction TO KANSAS ARCHEOLOGY 245Exotic WaresSo far as we can judge at the moment, the various pottery typesdescribed above can be considered local products, which may havebeen influenced in a few instances by the ceramic traditions of othernearby groups. There are, in addition, two or three sherds certainlyor possibly not of local manufacture. Such foreign pieces, where theymay be identified are of special interest for the hints they give as totrade relations with other regions and the clues they may provide asto chronological correlations. Unhappily, our exotic sherds from theTobias site throw little light on either of these problems.Near the west end of our initial trench through mound 17, at a depthof 11 inches among some large sandstone boulders, in square 60 W 3,was found a small gray sherd (USNM 389024) bearing painted lineardecorations. Mera states (letter of August 12, 1940) that this sherd ? a^ees specifically in paste and paint with late forms of Chupadero Black-on-white, a pottery type confined to the southern half of New Mexico. Although thepaint is normally matte in finish, glazing is not uncommon. Note the transparentglossy stripes on either side and below the brown patch of pigment, where thepaint was applied too thinly and has become largely invisible. This ceramic typeis of little use in dating as it is known to have persisted to some extent fromsometime in the 13th century up until about the middle of the 17th, a surprisinglylong time for a southwestern black-on-white type to survive. [PI- 49, f.]Two other gray sherds (USNM 388901) , both from mound 17, basin2, remain unidentified. The first has a thickly sand-tempered light-gray paste, which looks like local ware, but the well-smoothed surfacebears a faded black paint decoration consisting of wide uneven parallellines, crossed by uneven parallel zigzags. The second sherd, similarlypolished and sand-tempered, has irregular dark spots, but no definiteevidence of a painted design.Other Pottery ObjectsOur findings indicate that aside from the utilitarian vessels and oc-casional miniature pieces characterized in a preceding section, therewas relatively little work in clay. Eight pottery disks were found.Seven of these, ranging in maximum diameter from 20 to 37 mm.,were fashioned by grinding down the edges of potsherds until a cir-cular or subcircular outline was attained. The edges are usually irreg-ular; surfaces are plain, except in two examples which show simplestamping. The eighth specimen was made by pressing out a smallclay pellet ; it is thickest at the center, and expansion cracks radiateout toward the edge. The disks are 5 to 8 mm. thick; none is per-forated (pi. 30, g, K).From the bottom of a pit in square 60, mound 17, came a flattenedoblong lump of moulded clay, irregularly fashioned, with roundedends. It measures 13.5 by 7.7 by 4.5 cm. (fig. 44) ; its purpose is un-known. 246 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174 Figure 44.?Molded lump of clay from mound 17 excavations, Tobias site. Length 135 mm.OBJECTS OF ANTI^RHollowed tines.?There are three of these curious objects, all some-what damaged and restored (pi. 31, e-g.) . They range in length from15 to 23.5 cm., in maximum diameter from 1.9 to 3.2 cm., and the sur-faces have been carefully, but not always completely, smoothed or evenpolished. The tip of the smallest has been broken off ; in the others,the tips have been blunted and smoothed down. All retain the naturaleven curvature of the antler, and have been slightly tapered proxi-mally to a rounded or squared extremity. In addition, each has beenhollowed out for a distance of 6 to 9 cm. from the proximal end, so asto create a scooplike effect. This hollowing was accomplished in thetwo larger specimens (USNM 388682), from pit 3A, and (USNM388836 from mound 17,' basin 1) by hacking or grinding away thecompact surface tissue on the inner curved face of the tine, until thecancellous structure was exposed in a shallow longitudinal hollow.The third specimen (USNM 388926, from mound 17, basin 2) is badlybroken, but seems to have had a much deeper cavity which was on theside of the curve ; the remaining edge of the cavity is well smoothedand even, and most of the cancellous tissue appears to have beencarefully scraped away. This piece has four short transverse linesincised across the outer face about midway of its length. The dressedouter face of each of the two smaller specimens terminates proximallyin a low transverse flange 2 to 7 mm. wide. There is nothing about thespecimens, or in their associations or manner of occurrence, to indicatetheir purpose, and I have been unable to find any record of similar arti-facts from the Plains or elsewhere (fig. 45) . Projectile point.?From mound 17, basin 2, came a well-madestraight conical antler tip point, the only one of its kind found on thesite (pi. 34, i) . It is 52 mm. long, with a circular base 11 mm. in diam-eter, and has a conical socket 12 mm. deep. There is no basal tang.Similar points have a wide geographic and temporal distributionthroughout the Eastern United States. In the trans-Missouri plains, Weobl] AN INTRODUCTION TO KANSAS ARCHEOLOGY 247 Figure 45.?Hollowed antler artifacts from Tobias site. Length of a, 23.5 cm. on outsidecurve.they have been found in the Leary and Fanning sites, both Oneota ; and closely similar specimens with square bases are known from theWright site (protohistoric Pawnee) near Genoa, Nebr. On an earliertime level, specimens from the prehistoric Eenner site (Hopewellian)differ in having a basal tang.Scored object.?The short rodlike specimen shown in plate 35, I(USMN 388683) has been extensively worked down, so that it consistsmostly of cancellous tissue with a thin layer of harder structure oneach side. It has a somewhat elliptical cross section. The ends aredamaged, but one appears to have been bluntly rounded, the otherconically socketed to a depth of 8 mm. or more. About 10 cm. fromthe blunt end was a biconically drilled perforation 4 mm. in diameter.The flattish surface below this hole is split on one side the full lengthof the object, and had 28 short transverse cuts averaging 5 per cm.The piece is well shaped and carefully smoothed on all surfaces ; it is68 mm. long, 10 mm. wide, and 8 mm. thick.Worked tips.?A blunt rounded antler tip fragment 4 cm. long hasfour groups of short closely set transverse cuts on opposing sides.These groups include 8, 8, 6, and 5 cuts, the lower numbers possiblybeing due to extensive rubbing down of the tip. There is no way ofjudging whether this piece was broken from an implement or repre-sents an otherwise unworked tine (pi. 35, k) .Another antler section from near the end of the tine has been splitlengthwise and the cancellous tissue largely removed. The split edgeshave been smoothed; the distal end is blunted, and to a distance ofabout 3 cm. from the tip has nmnerous transverse scratches or knifescars on the exterior surface. Present length of the piece, which isbadly weathered and evidently incomplete, is about 8 cm.Socketed basal section.?The weathered and cracked basal portionof a deer antler, detached just below the "burr," was evidently par- 248 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174tially smoothed by rubbing. About 9 cm. above the proximal end, alateral tine has been detached, leaving a rounded stump. The distalextremity is cut off square, and has a shallow irregular cavity about 8mm. deep. Neither end has been smoothed. The length is 13 cm.Cut tines?There are several of these ; some may be unfinished im-plements, others are probably refuse from artifact manufacture.From pit 3A came a forked tine 11 cm. long, of which the larger branchbears numerous fine lengthwise striae and has had the tip partiallysawed all-round and then snapped off; a lateral tine, also finelystriated, has had the tip broken off. The diameter of the main stem atthe cut is slightly greater than the basal diameter of the finished pro-jectile point already noted ; and the striae, which considerably modi-fied the original antler form to a distance of ca. 12 mm. below the cut,suggest that the tip may have been roughed out by scraping prior todetachment and final polishing. Another tine from which the tip hasbeen neatly severed is similarly striated for a short distance belowthe cut; it, too, may be rejectage from manufacture of a projectilepoint. A third specimen, 6.5 cm. long, has the tip broken off' anda deep cut partly encircling the middle.Eyed and/or notched objects.?This includes 1 complete and 13fragmentary specimens ; all are made of narrow, flat or subcylmdricalstrips of antler or possibly bone, so thoroughly worked down thatpositive identification is not possible. With one exception, all arecurved, and there is commonly a faint suggestion of fine cancelloustissue on the concave surface. The edges are well smoothed, and onnone is it possible to detect any scars or other clues to the manner ofdetachment of the strips.The complete specimen (USNM 388704) is from pit 5, from whichcame five of the fragments. It is evenly curved and well polished,and has a length of 24.5 cm. along the arc or 19.4 cm. along the chord.Near the middle it measures 4.5 by 5 mm. in cross section, whence ittapers toward each end. One end has been neatly squared and per-forated; at the other end, which seems to have been broken or cutthrough a similar hole, is a deep transverse groove across the outerconvex surface of the curve. The concave surface shows faint tracesof cancellous structure; the curve almost exactly duplicates that inseveral unworked antler tines and in the largest hollowed tine de-scribed above (pi. 34, c) .The fragments add little to the foregoing information beyond in-dicating the range of variation in width (4-8 mm.), thickness(2-5 mm.), and cross section (thick elliptical to thin planoconvex).Eight have a small biconical perforation near one end, and in twoinstances there is evidence that the specimen had been previouslybroken off through another similar hole. In a single instance, this Wkdbl] an ESTTRODUCTION TO KANSAS ARCHEOLOGY 249hole is slightly -worn lengthwise as from a cord or string. Threespecimens have a transverse groove across one dressed end ; the edgesof another have one and two notches near the end. All of the frag-ments, whether terminally grooved, notched, or pierced, show at leastone more or less ragged fractured end, usually that opposite the grooveor perforation. In addition to the five fragments from pit 5, thereare two from pit 3, one from pit 1, one from mound 17, pit 8, two frommound 17, basin 2, one from mound 17, basin 2, pit B, and one frommound 17, basin 2, pit C (pi. 34, ?, 6, e-A, j^ k) .The purpose of these objects remains uncertain, though the per-forations and notches suggest points of attachment for a string. Asto the method of manufacture, I suspect they were sawed and splitfrom the outer face of suitable lengths of antler rather than fromribs or other bones. The curvature then would be wholly natural.In none of our excavations, however, did we find any scarfed antlerrejectage from which such strips might have been cut. The inferredmethod of detachment is discussed and illustrated by Kidder (1932,p. 272 and fig. 230).Needle fragment.?This is a thick, curved, shaped, but apparentlyunfinished strip of antler, broken at one end, and retaining the roughcancellous tissue on the concave face. The other end has been squaredoff and dressed, as are the adjacent edges. About 8 mm. from this endis a long narrow eye, 8 by 2 mm., that has been cut or sawed ratherthan bored through. The specimen is 40 mm. long, 12 mm. wide at thebutt, and 4 or 5 mm. thick. It may represent part of an imfinishedneedle for mat weaving. OBJECTS OF BONEIn striking contrast to the sites in eastern Kansas described in thepreceding sections, the Tobias site presents a surprisingly extensive,varied, and interesting series of bone artifacts. (See table 4.) For themost part, moreover, the specimens are in good preservation. Theyare primarily utilitarian; other than beads and tubes for personaladornment, there is nothing that can be regarded as of esthetic oresoteric character. Many of the specimens, however, are very wellmade and carefully finished, and it is evident that the native craftsmenprided themselves on their work?probably to a far greater degreethan did the later-day artisans in competition with the white traderand his wares. The local bone industry is of interest, too, in that itseems to lack certain characteristic late Plains artifact types at thesame time that it presents others new to the area.Most of the implements appear to have been fashioned by cutting,grinding, and rubbing, from the limb bones, ribs, and neural spines ofsuch large mammals as bison, elk, and deer. Other than for beadsand tubes, there seems to have been no utilization of bird bones. So 250 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174 ? Wbdbl] an introduction TO KANSAS ARCHEOLOGY 251far as our observations go, there were no significant concentrationsof types in certain pits or other units that might indicate manufactureby, or ownership within, a single household or family group. That isto say, examples of any of the following types can be expected to turnup in any part of the site where occupational debris has accumulated.Bison scapula digging tools.?These familiar implements, or frag-ments of them, were found in all parts of our excavations, and wereespecially plentiful in the cache pits where they probably outnumberall other artifact categories. Most of the specimens are incomplete;where identification is possible, all the bones seem to be bison. Theseries, numbering upward of 60 pieces, includes not more than 4 or 5distal extremities, suggesting that this part of the bone was usuallyremoved and the resulting stmnp but slightly smoothed down. In allcases, the scapular spine has been hacked off, as also the ridge alongthe posterior or axillary border ; very rarely has there been any effortto smooth over the irregularities left by this trimming process or tobring the ragged scars so left flush with the blade. The proximalone-third or one-half of the scapula, including the entire vertebralborder, was invariably cut off', usually leaving a more or less straightworking edge, less commonly a rounded one. The finished tool, thus,would be subtriangular in outline, with rounded angles and rarelywith the short side convexly curved. The working edge commonly hasa short steep bevel, is sometimes worn back to a concave line, and onmany specimens bears a high polish from extended use. There is nonotching or grooving on most of our specunens, and I cannot say whatthe manner of hafting was. The two largest examples are 31.5 cm.long, with widths of 11 and 16 cm., whence they range downward insize to 16.5 by 8.5 cm.One incomplete digging tool from pit 8 is of exceptional interest byreason of its "non-plains" manner of hafting. In this specimen noattempt was made to remove the head. From the glenoid cavity a deepgroove or socket has been excised about 7 mm. down the dorsal surface ; the ventral margin of the cavity has been battered and chipped,whether in connection with the socketing or subsequently I cannot becertain. The working edge, of which only a portion remains, isbeveled and highly polished; and the scar left in removal of thescapular spine has been rubbed down with some care. The total lengthof the artifact is 24 cm. I am unable to find record of similar socketeddigging tools in other archeological horizons in the Plains area, butas we shall see presently, they do occur farther south in the Arkansasdrainage.Bison ulna pick ( ? ) .?A single specimen is recorded from the site,coming from pit 3A. The rugosities where the ulna was detachedfrom the radius show unmistakable polish from use, and the tapering 252 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174distal end appears to be battered and somewhat blunted from use.There are no other convincing signs of wear on the object. So far asI know, there is no historical reference to these tools, and their identifi-cation as picks rests on inference alone. The broad upper end wouldhave made a very convenient hand grip, and the stout point wouldhave been serviceable in any hard-dirt excavating.Awls.?There are 62 whole and fragmentary implements identifiedas awls; they come from nearly every one of our excavation units.Discounting 8 very fragmentary and unclassifiable specimens, 4 areof split mammal leg bone, 13 are of rib, and 37 are of neural spine;there are apparently none of bird bone.Split leg-bone awls include one sturdy well-made specimen, ofdeer or antelope metapodial, in which the head of the bone has beenunmodified except by the original splitting; it measures 14.5 cm.long, and is from mound IT, basin 2. In three other well-wornspecimens, the joint ends have been partly ground down after split-ting; two are from pit 5, one from pit 8, and length ranges from6.Y to 12.8 cm. (fig. 46, a, h ; pi. 32, j-l) . Rib awls are of two kinds: four specimens are made of the hardinner or outer surface layer of a split rib. One face thus is flat andsmooth, the other shows more or less cancellous tissue ; the edges havebeen rubbed down, the butt rounded off, and the other end sharpenedto a point (fig. 47, a; pi. 32, n). These implements are from 8.3 to13 cm. long, and from 1.3 to 1.9 cm. wide ; two came out of pit 3, twofrom mound 17, basin 2. Nine other specimens are apparentlysplinters from the face, or face and edge, of ribs; they are irregularin form and size, generally showing no modification other than theshaping of one end to a sharp point. They range in length from2.9 to 12.4 cm.Far more abundant and characteristic than any of the precedingare awls cut from the edge of a large rib or neural spine. Typically,these taper more or less evenly from butt to tip. In cross section,they are subtriangular with the angles rounded ; the butts are eitherrounded or subpyramidal. Two sides, representing the convergingsurfaces of the bone, are invariably smooth; the third, except in thesmallest specimens, usually still retains some traces of the cancellousinterior tissue of the bone. Complete specimens range from 6.8 to 15cm. in length. They apparently conform in all respects to the "edgeol rib" awls found at Pecos by Kidder (1932, p. 217), but the contoursof some of the larger and less extensively altered specimens fromTobias suggest strongly that they were cut from the anterior marginof the neural spine of the bison. None of the specimens show anytrace of the slight curvature characteristic of a rib (fig. 47, h, c;pi. 32,e,*). Wkdbl] AN INTRODUCTION TO KANSAS ARCHEOLOGY 253 w M ?Uai Figure 46.?Bone artifacts from Tobias site, a, h, split leg-bone awls; c, 4-sided implement;d, e, projectile points. Length of a, 14.5 cm.Elsewhere in the Central Plains, split-rib awls have been reportedfrom 25CH1, a Dismal Eiver culture site in Chase County, Nebr. (Hilland Metcalf, 1942, p. 196) ; the triangular type is reported from thesame site, and also from the Burkett, Wright, and Gray-Wolfe siteson the Loup (ibid., p. 197; Dunlevy, 1936, p. 197 and pi. 13, l-d).Stemmed implements.?There are eiglit of these. They are approxi-mately square, quadrilateral, or subcircular in cross section, evenlytapered at one end to a thick point and provided at the other end witha constricted, usually well defined stem (pi. 33, e, K-j) . All are wellformed and have been carefully rubbed down. In cross section, one 254 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174 A-l Figure 47.?Bone awls and poIishers(?) from Tobias site, a, flat split-rib awl, length 12.7cm; b, c, awls from edge of neural spine; d-f, awls or polishing tools from edge of neuralspine or rib.short side or edge usually is a trifle thicker than the other, and in atleast two specimens this thicker edge is partially surfaced with can-cellous tissue. From these features, I suspect that the implements, likethe typical triangular-section awls, were cut from the anterior marginof the neural spine of the bison. They vary in overall length from6.3 to 12.6 cm. ; with a maximum diameter of 6.5 to 10 mm. ; the stemis 11 to 23 mm. long, separated by an abrupt shoulder from the blade.All of our specimens are from the mound 17 complex : six from basin 1,and two from basin 2 (fig. 46, e).There is no obvious reason for doubting that these objects representprojectile points, intended for mounting in a hollowed foreshaft orreed arrow. Similar specimens have been reported from the Burkettand Wright sites in Nance County, Nebr. (Dunlevy, 1936, pi. 13, E,and p. 197; Hill and Wedel, 1936, p. 58) and from the Lovitt site,25CH1, in Chase County, Nebr. (Hill and Metcalf, 1942, p. 199 andpi. 9, fig. 2, A and D) . These sites are all on the protohistoric level, Wedel] an introduction TO KANSAS ARCHEOLOGY 255 i. e., early European contact ; I know of no prehistoric occurrence ofthe type in the Central Plains. Kidder (1932, p. 240 and fig. 200, c)reports a closely similar specimen from late Glaze III (ca. 1475-1550)deposits at Pecos.Four-sided implements.?These objects, six in number, are square,square with rounded angles, or quadrilateral in cross section (pi. 33,a-d). One end has been shaped to a thick sharp point, which mayretain the square form and angles of the shaft ; the other is rounded off,and has had the angles rubbed down. One specimen is polished on allsurfaces; the others, though shaped with much care, still bear faintlongitudinal striations on their flat or slightly rounded sides. Twostill exhibit faint traces of cancellous tissue along the midline of oneside, suggesting that they, and probably the others as well, were cutfrom the neural spine of the bison. They are somewhat better madeand more symmetrical tlian the stemmed implements. The blunt androunded end, which in no case shows the abrupt shoulder of the pre-ceding type, nevertheless suggests modification for insertion into asocket or hollow shaft, and these specimens may also represent projec-tile points. They range in length from 9.8 to 13.5 cm., and in maximumdiameter from 7 to 8 mm. As with the stemmed implements, all arefrom the mound 17 complex: four from basin 1, two from basin 2(fig. 46, c).There appears to be no published record of comparable forms fromother sites or horizons in the Plains area. At Pecos, however, closelysimilar specimens were found "at all levels except those of definitelyhistoric period" (Kidder, 1932, p. 225 and fig. 198).Tapered cylindrical implements.?This group comprises four speci-mens ; two are from mound 17, basin 1, and two from mound 17, basin2. All are circular, or very nearly so, in cross section. They taperevenly toward both ends, one of which is pointed, the other rounded ; all are very well finished and remarkably symmetrical (pi. 2>2>,k). Thesmallest specimen, which differs slightly in form and proportions,shows cancellous tissue along one side; the others do not. Lengthvaries from 8.3 to 13.8 cm., maximum diameter from 6 to 9 mm.(fig. 46, d).From the blunted butts of these objects, which differ only in crosssection from the four-sided implements, I suspect that they may alsobe projectile points. They have not been reported, apparently, else-where in the Plains ; a possibly similar artifact from Pecos, describedonly as a "Double-ended implement, round in section" is illustratedby Kidder (1932, fig. 200, f ) . Bipoinfed iinplements.?Included here is a rather variable seriesof nine objects, characterized by relatively small size, generally goodfinish and shaping, and by two pointed ends (pi. 33, o-t). Cross484172?59 18 256 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174 sections vary from subcircular to subtriangular to flattened. Two orthree show cancellous structure on one side, but most have beensmoothed on all sides. The ends are variable : one is commonly morerounded or more abruptly tapered, than the other, and one or bothmay be flattened or otherwise asymmetrical. They vary in lengthfrom 3.8 to 8 cm., and in maximum diameter from 3.5 to 6 mm. Thelarger specimens suggest projectile points of the tapered cylindricaltype, but others are too small or seem too asymmetrical for such use.Gorge hooks might be suggested except that there is no evidencefor the use of fish in the bone refuse from the site. The series prob-ably includes implements of widely diverse uses, none of which areobvious to me. As to provenience, specimens came from three pits,from mound 6, and from mound 17, basin 1.Polishing (?) tools.?Here are included, with some uncertainty,five specimens of two distinct types. Three have a subtriangiilar crosssection, nicely rounded butts, and show traces of cancellous structurealong one flat side (pi. 32, a-c^ ff^h). In these features they are iden-tical with the commonest type of awl from the site (p. 252). Theydiffer, however, from what I have classed as awls in having blunttips, with one or more flattened wear facets or else with fine diagonalstriations (fig. 47, /). As piercing tools they would not seem to havebeen very effective, and the evidences of wear suggest a sidewise rub-bing motion. It is possible that some of the implements counted asawls should be included here, since the two types intergrade (cf.Kidder, 1932, p. 229).Two other polishing (?) tools are made of sections of animal rib.One, apparently the hard outer surface with faint traces of can-cellous tissue on one side, is 11 cm. long and tapers at each end to athick rounding point with worn edges. These points would seem muchtoo heavy and blunt for piercing, and the wear faces indicate thatthe edges of the points were the working part. The second pieceis irregular, with blunt flattened and worn ends ; it is 10 cm. long.Pointed split-rib implements.?There are five of these, all of bisonor elk rib. Two consist of 10 or 11 cm. sections of whole rib, cut orelse broken off at one end, and provided at the other end with a flatheavy point 6 or 7 cm. long. This point was made by transverselyand deeply scoring one surface, splitting away the rib face beyond thescoring, and then tapering the remaining single face of the rib. Theunsplit section is of convenient size and thickness for grasping, andI suppose represents a gi'ip. For what purpose these implements wereintended I do not know, but the sturdy points and grip suggest heavyduty tools (pi. 36,^).From mound 17, basin 2, came a specimen 23 cm. long, split fromthe convex outer face of a rib, and tapering to a long use-polished WBDKL] AN INTRODUCTION TO KANSAS ARCHEOLOGY 257point. The edges have been smoothed, and transverse grinding striaeremain on both surfaces; an miknown length at the butt has beenbroken away. There is a slight indentation on each edge 45 to 70 mm.from the wide end. Maximum width of the piece is 19 mm. (pi.36, e). From pit 5 came a heaA^, much worn point of split rib, 10cm. long and 18 mm. in maximmn width. The cancellous surface andedges have been worked down. The point is much heavier than theusual bone awls, and to me suggests the working end of another longflat implement like the preceding.The remaining specimen is made from the concave inner surface ofthe split rib. The edges have been ground down, but not smoothedor polished. One end rounds off into a thin flat point; the other isbroken but may once have tapered in somewhat similar fashion. Onthe cancellous surface there has been just enough smoothing to removeany sharp edges. The specimen measures 23.5 by 5 by 2 cm. ; thereare no wear facets to suggest the manner of use.Arrowshaft straighteners.?These implements, of which we foundone complete and two fragmentary specimens, were made from sec-tions of bison or elk rib (pi. 36, a). The complete specimen, frompit 3, is 27.5 cm. long; both ends of the original rib were unevenlybroken off, and the fractured edges smoothed. Eleven cm. from thenarrow (proximal) end is a hole 8.5 mm. in diameter. On the innersurface of the bone the hole is slightly elongate toward the distal endof the rib ; on the outer surface it is similarly worn toward the proxi-mal end. Such wear would result if the user, holding the wrench inhis left hand with the concave surface to the right inserted a shaftwith his right hand and then exerted the necessary pressure by bring-ing both hands together. There are numerous fine striations and twodeeper cuts on the concave surface of the implement ; on the convexsurface are 3 or 4 transverse cuts and a number of faint, more or lessevenly spaced, notches suggesting that the specimen had been markedoff originally for a "rasp" and was then converted before completionto another use. Both surfaces are worn from much handling.The two fragments are from moimd 6, and possibly represent asingle specimen, but they cannot be satisfactorily fitted together.Both are from a split rib section : one is from the outer convex surface,the other from the inner concave surface. One fragment is 12 cm.long ; at the center is an 8-mm. hole, very slightly elongated by wearalong the longitudinal axis of the bone. The second piece is 6.5 cm.long; at one end it has been broken off through an 8.5-mm. hole, theremaining half of which is similarly worn.The bison-rib shaft straightener is a common and widespread typein the Plains, having been reported by various observers in numeroussites from Kansas to North Dakota and Montana (see, e. g., Wedel, 258 BUREAU OF AMERICAN ETHNOLOGY [Boll. 1741935, pp. 202, 232; Strong, 1935, pp. 249, 261; Wedel, 1936, p. 82;Hill and Wedel, 1936, p. 55; Hill and Metcalf, 1942, p. 199; Cooper,1936, p. 54; Dunlevy, 1936, p. 197; Champe, 1936, p. 268; Will andSpinden, 1906, p. 170 ; Strong, 1940, pp. 370, 375 ; Mulloy, 1942, p. 72) . It has recently been suggested (George and George, 1945, p. 65) thatperforated rib implements with perfectly round holes were arrowshaf t gages, those with worn holes thong dressers. If this is a valid distinc-tion, most Central Plains specimens, since they show wear facets,would have to be regarded as thong dressers. On the other hand,shaft straighteners of antler have been reported ethnographically forthe Omaha (La Flesche, 1924), and it is a reasonable inference thaton the Plains proper an analogous form of bone, like those under con-sideration, served the same purpose. Prolonged use as a wrenchwould inevitably produce wear facets along the longitudinal axis ofthe holes. Moreover, since breakage is often transversely across one ofthe holes, I think they were subjected to a much greater strain thanwould be caused by the to-and-fro action of a thong. It is possible,and even probable, that the implements served more than one use.Parenthetically, it may be noted that of 87 Arapaho, Kiowa, Sioux,Comanche, and Shoshone arrowshafts which I measured in the Divi-sion of Ethnology, United States National Museum, 67 specimens(77 percent) were under 8 mm. in maximmn diameter, and thus couldbe conveniently accommodated by any of the arrowshaft straightenersin our archeological collection from Kansas.Scored implements.?There are 14 fragments of bison or elk rib,bearing deep parallel transverse cuts more or less evenly spaced alongthe outer surface of the bone; the cuts usually end short of, or justat, the lateral edges of the rib. In three specimens the gTOoves appearto be slightly deeper at the sides of the rib than along the midline,where a somewhat worn zone suggests the use of a rubbing stick. Onthe majority, however, no such action is indicated. Unfortunately,none of the specimens appears to be complete, and so the total lengthof the scored surface originally can in no case be stated. The frag-ments range in length from 3.5 to 34 cm., and bear from 5 to 42 scor-ings. The intervals between scorings vary considerably. The long-est specimen, 34 cm., has 18 grooves distributed over a 17-cm. sectionof the surface; it possibly has all of the original grooves (Wedel andHill, 1942, pi. 10, c) . The next longest example, 31.5 cm., has 42 cutscovering 22 cm., and the grooves apparently continued for an un-known distance beyond the fractured distal end of the bone. Theclosest scoring occurs on a fragment from pit 3, which has 30 cuts in8 cm. Cuts per centimeter of scored surface thus range from oneto nearly four, with about half the specimens averaging just undertwo to just over three. We have 6 examples from 4 pits, two from Wkdbl] an introduction TO KANSAS ARCHEOLOGY 259mound 6, and one and five, respectively, from basins 1 and 2 in momid17 {i>m,a,b,d).In addition to the above, there is one neural process of a bisonwhich has five unevenly spaced scorings on one lateral surface nearthe distal end. It is 28 cm. long, and was found in mound 17, basin2 (Wedel and Hill, 1942, pi. 11a) . Transversely scored rib implements of the type described abovehave been found (Wedel and Hill, 1942, p. 96) at several protohistoricand historic Pawnee sites in Nebraska ; at Dismal River culture sitesin Scott County, Kans., and Frontier County, Nebr. ; at most of theprotohistoric sites so far tested in central and south-central Kansas ; but not, so far as I know, in the Woodland, Upper Republican,Nebraska Culture, or other prehistoric horizons of the Central Plains.Their wider distribution has not been fully worked out; and itappears that the published record by no means accurately or fullyindicates their range. Northward, they have been reported archeo-logically at least as far as the Leavenworth (Arikara) site nearMobridge, S. Dak. (Strong, 1940, p. 370). To the south and south-west, they occur in late prehistoric sites of the Texas Panhandle cul-ture, where adequate descriptions are unfortunately lacking (Holden,1933, p. 48 ; Studer, 1934, pp. 90-91 ; Johnston, 1939, p. 197) . Kidder(1932, p. 252 and fig. 212, d-f) notes their occurrence in Glaze III-Vlevels at Pecos, and Hodge (1920, p. 139 and pi. 44e) records a some-what similar but not identical piece from Hawikuh. Eastward, theyoccur sporadically in a few Oneota sites, though not apparently atLeary or Fanning; and at Madisonville, Ohio (Hooton and Wil-loughby, 1920, p. 62) . They would appear, on present evidence, to berather more plentiful perhaps in the Central Plains than in surround-ing areas ; and in and near the Plains, at least, they seem furthermoreto belong to a relatively late period.At Pecos, Madisonville, in the Texas Panhandle, and elsewherethese implements have been identified as musical rasps, and it ispossible that some of the Central Plains specimens can be similarlyclassed. The wear on some of the specimens from the Tobias site isgreatest along the midline, where the edges of the grooves have beenconsiderably modified by the action, inferentially, of a stick or bonedrawn across the scored surface. For what purpose a stick wouldbe so applied to the grooved surface, unless for rhythmic or musicaleffects, I do not know. So far as I am aware, there is no historicalor ethnographic description of a musical instrument in the CentralPlains that unquestionably applies to the transversely scored ribsrather than to, say, an edge-notched stick. Roberts (1936, p. 24,fig. 5) ascribes a notched stick or bone rasp, used without resonator,to the Omaha and neighboring Plains tribes ; and Dorsey (1904, p. 17) 260 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174 says that the Wichita kept time for certain chants by "the drawingof a stick over a notched club, one end of which rested on a buffalorawhide resonator." It thus appears that some type of musical raspmay have been known in the central Kansas-Nebraska area; and itis possible that the instrument may occasionally, or at one period,or among some groups, have been made of transversely scored animalribs.Elsewhere (Wedel and Hill, 1942) it has been suggested that someof the Plains specimens have few notches, are quite small, and lackany evidence of unusual wear on the scored surface ; and that, on theother hand, they produce in plastic clay a ridged effect indistinguish-able from that on much of the simple-stamped pottery from the samesites. I am still of the opinion that some of the scored ribs may havebeen used in surface texturing of pottery ; and the few-scored neuralprocesses and occasional paddle-shaped scapula objects seem to meto be far better suited to pottery stamping than to producing of musi-cal or rythmic effects. The manner in which they occur in sites nomore strongly supports their identification as musical instrumentsthan it does the suggestion that they were pottery-making tools.Cancellous hone implements.?There are three of these. One, frompit 3, is wedge-shaped, with one edge nearly straight and the peripheryotherwise forming a partial ellipse. At one edge is a small area of thehard outer layer ; otherwise, the piece is entirely of cancellous tissue(pi. 36, g) . It measures 68 by 50 mm., and has a maximum thicknessof about 10 mm. ; the edges are thinned down to about 1 or 2 mm.,which is about as thin as the spongy texture of the bone permits. Thebone is probably from the head of a bison femur or from the neuralspine, in any case certainly not from the nose bone of the bison, as hasbeen erroneously supposed. The specimen closely resembles the paintapplicators used by some historic Plains tribes for decorating tipicovers and otlier articles of skin (Fletcher and La Flesche, 1911, p.354 and fig. 78), though there are no traces of coloring matter in theinterstices. Similar objects have a fairly wide distribution in thePlains, and most of those so far reported archeologically appear to befrom protohistoric and historic horizons (Will and Spinden, 1906, p.171 ; Wedel, 1936, p. 82 ; Dunlevy, 1936, p. 199 ; Hill and Metcalf , 1942,p. 199; and this bulletin, p. 581). At Pecos, Kidder (1932, p. 238)records the same type from late levels, noting further the absence ofpaint traces and pointing out the suitability of the objects for hidescraping.The second object of cancellous bone from the Tobias site is a smallirregular subspheroidal piece measuring 33 by 36 mm. There can beno doubt that it has been shaped by man, but I am unable to suggestits possible purpose. It was found in mound 17, basin 2, pit D. Weuel] AN INTRODUCTION TO KANSAS ARCHEOLOGY 261The naturally rounded head of a bison humerus or femur, with aslightly concave smooth-worn cancellous under surface, appears tohave served as a hide-rubbing tool, as it is known to have done inlater historic days on the Plains (fig. 48). Figure 48.?Top (left) and bottom views of bison epiphyseal hide-rubbing tool, Tobiassite. Length, 10.2 cm.Gamhling chips ( ? ) .?Two oblong objects, each with one surface ofsmoothed dense bone, and the other with traces of cancellous tissue,have rounded ends and dressed edges, but otherwise were not finishedwith especial care. The larger specimen, measuring 17 by 55 mm.,has numerous fine striae on the smooth surface, but without any sug-gestion of a pattern. The other has traces of red pigment on thecancellous surface, and faint transverse striae more or less groupednear the middle and again towards each end of the smooth face. Inform and size these objects resemble the incised Mandan gambling-chips figured by Brower (1904, p. xvi) ; painted surfaces may herehave served in lieu of incising. On the other hand, it is quite possiblethat these objects had nothing to do with gambling (pi. 35, i, j) .Cut bison hyoid.?One specimen came from mound 6. Both ex-tremities have been carelessly and unevenly detached. The flat sur-faces are moderately well polished and there are rodent toothmarkson one edge. The piece, whose purpose I cannot guess, is 77 mm.long; it may represent an unfinished artifact. A similar but betterfinished specimen, pierced lengthwise, was found by us in ScottCounty State Park (p. 454) . Incised hone strip.?From mound 17, basin 2, pit C, were taken threethinly scraped slightly curved strips of bone, with dressed edges andbroken ends ; the concave surface of each is faintly cancellous. All arefrom 1.5 to 2 mm. thick, 8 or 9 mm. wide, and from 24 to 33 mm. long.The shortest piece has been turned ashy-white by fire ; it has 2 parallelincised lines running lengthwise along one edge. A second fragment,26 mm. long, has 7 tiny nicks on one edge, and deep narrow lineserratically incised on a very well polished surface. The third piece, 262 BUREAU OF AMERICAN ETHNOLOGY [Bull. 17433 mm. long, is also highly polished and has three erratic deeply in-cised lengthwise lines. One of the outer lines is crossed by five shorttransverse cuts, grouped as four near one end and one near the other.Faint but unmistakable traces of red pigment are visible on both thepolished fragments, in the incisions and also in the cancellous tissue.The second and third pieces may be from the same original specimen,the first less probably so ; but it is impossible to determine the length,shape, and purpose of the object or objects represented.Tubular heads.?Small dressed cylinders cut from hollow bone, andnot exceeding 9 mm. in diameter or 63 mm. in length, are here classedas beads (pi. 35, c, d^ /). They were made by sawing a deep groovearound the shaft of the bone near each end, and then snapping offthe articular extremities. Six such extremities, each with a shortstub of cleanly severed shaft, were found ; they range downward insize from that of a turkey, but beyond this the species used are notidentifiable. The cut ends of the cylinders were neatly smoothed ; thesurfaces, undecorated, show varying degrees of polish from use. Oursmallest specimen is 1.5 by 10 mm., whence the others range upwardin diameter and length as shown in the accompanying chart. Lengthvaries from twice to ten times the diameter. The standard formwould seem to be 3 to 6 mm. in diameter by 12 to 38 mm. long. Twospecimens each have a discontinuous encircling groove or series ofcuts near one end, suggesting the beginning of an attempt at severingpart of the tube ; another, 45 mm. long, has a deep encircling grooveabout 12 mm. from one end, and a single short shallow cut about thesame distance from the other end. Scattered specimens came fromnearly every part of our excavations, indicating that they must havebeen in general use. In only one instance, however, did we find anumber of them in what undoubtedly approximated their originalassociations ; this find, a necklace, has been described in another place(p. 296) , and is not included in table 5.Table 5 ^~^-^^ LengthDiameter ^"^^^ Wbdbl] an introduction TO KANSAS ARCHEOLOGY 263Bone tubes.?Two comparatively short thick tubes came from pits1 and 13. One is highly polished, with somewhat irregular ends, anda subtriangular cross section. On the ridge forming the apex of thistriangle are 10 notches, and at one end, 3 incised lines cross the ridgeand extend down each side (pi. 35, e). The specimen measures 14 by45 mm. The other tube, 11 by 62 mm., is smoothed, and has unevenlycut ends, with a few scratches at one end, but no decoration.Very different are several long slender tubes apparently cut from theradii of large wading birds. From pit 1 came two such specimens ex-hibiting complementary curvatures and obviously a matched pair.They are plain, well polished from use, with a length of 125 mm. andmaximum diameter of 5 mm. The articular surfaces are missing, butin size, conformation, and curvature the specimens are identical withradius shafts of the great blue heron (Ardea herodias) ; ^^ very likelythey are fashioned from the radii of a single bird. A similar butslightly smaller (100 by 4 mm.) specimen came from mound 17, pit 8;and shorter but similarly curved and worked radius segments werefound in pits 3 and 8A. Just how these long tubes were used is notclear, but presumably they were for personal adornment (pi. 35, a, 6).There is also one slivered fragment of thick-walled mammal bonetube from pit 1 ; it was partially sawed through and then snapped off.It measures 65 by 12 mm.Miscellaneous objects.?^This includes a varied series of worked spec-imens which on present evidence do not seem to constitute well-definedtypes.An irregular curved slab 105 mm. long by about 16 mm. wide, cutor broken from heavy mammal bone, has had both long fracture edgessmoothed, and the slightly broadened spatulate end ground from bothsides to a cutting or scraping edge 20 mm. wide. It was found inmound 6.Several small fragments of bison scapulae, possibly readapted bitsfrom broken digging tools, have had the broken margins rounded offslightly as though to give a more convenient grip. The cutting edges,sometimes beveled, would have served nicely for scraping, rubbing,or smoothing some moderately soft substance, or for slicing up squashor similar vegetables.A large irregular fragment from the anterior surface of an elkmetatarsal has had the vascular groove artificially deepened by narrowU-shaped scoring. Both lateral surfaces are highly polished and be^irdeeply cut Y, V, A, single line, and other incisions; these run to thebroken edges, and may once have continued onto a larger portion of thebone surface that has since broken away. The fragment measures105 by 30 mm.; it came from mound 6 (pi. 34, I). **! am Indebted to Dr. Herbert Friedman of the National Museum staff for thisIdentification. 264 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174A highly polished cylindrical fragment 58 mm. long by 5.5 mm. indiameter has one end pointed and the other rubbed down after frac-ture. On one side are two groups of transverse cuts, with 4 and 2 cutsper group. Opposite the group of 4 is a small neatly cut X ; besideand just above the group of 2 is another X, and on the opposite sideis a third X. Near the refinished butt are 2 more short cuts. Thespecimen somewhat suggests the bipointed objects or the projectilepoints ( ? ) described above, and is as well finished as any of them.It is unique in having the paired cuts and crosses. It was found inmound 17, pit 8A.The distal extremities of two cervid cannon bones, probably deer,have been cut and snapped off from the shaft. One, from mound 17,pit 8, shows no further modification. The second, from pit 8, showsartificial deepening and lengthening of the vascular groove; on theopposite surface is a deep straight cut that reached the cancelloustissue, with a shallower paralleling groove. Both specimens are un-doubtedly rejectage; and the grooving on the second piece suggestssteps in the splitting of the bone for awl-making.Other cut, smoothed, and shaped fragments are too small and non-distinctive to merit detailed treatment here. They include about 20butt and tip fragments of unclassifiable awls, the midsection of an-other four-sided implement, scapula fragments, etc.OBJECTS OF CHIPPED STONEAmong our collections from the Tobias site, chipped-stone arti-facts rank in abundance immediately below pottery remains. Thevariety of types is not great, but adequate samples are generallyavailable. Workmanship is good, sometimes superior, and many ofthe specimens are excellent examples of the stone-flaker's art. Ingeneral, the projectile points, drills, certain forms of knives, andscrapers are at least the equal of any chipped-stone implements pro-duced at other Plains sites, and they illustrate rather well the degreeof competence attained by native craftsmen before the introductionof more efficient metal products.In part, of coui-se, the raw materials used were responsible for thehigh quality of the finished products. These included brown jasper,apparently not particularly well suited to fine retouching; gray,white, pink, and variegated cherts of uncertain origin ; pink, brown,or gray Florence flint, often attractively banded and characterizedby fusulinid and other fossil inclusions (see also p. 476) ; chalcedonyand moss agate, perhaps from somewhere to the west or northwest;and a little obsidian. This last, inferentially, was imported from theNew Mexico region ; none of the spalls, cores, and flakes that came toour notice exceeded 50 mm. in greatest dimension, and most were Wkdbl] an introduction TO KANSAS ARCHEOLOGY 265 mucli smaller. Thirty-one pieces of obsidian, but no finished arti-facts, came to light, and since they occurred in several pits and allof the mounds we opened, it can be concluded that the material waswell known locally. Because of the distance involved in its acquisi-tion, and also because of the excellent cherts abundantly availablefrom nearer sources, obsidian never could have been utilized to theextent warranted by its inherent superiority.In the following discussion, I have found it convenient to followKidder's (1932, p. 15) outline for Pecos chipped stone, which treatsfirst artifacts with secondary chipping (projectile points, drills, bev-eled knives) ; then artifacts with no secondary chipping on one ormore major faces (end and side scrapers) ; and finally, artifacts withno secondary chipping (axes and other implements) . Pi'ojectile points.?Of 210 specimens classed as projectile points, 178(85 percent) are small triangular forms with or without notches.These occurred everywhere in our excavations and undoubtedly rep-resent the standard type. Unnotched specimens decidedly predomi-nate, there being 134 (64.1 percent of all) of them. The base variesfrom straight to very slightly concave or convex ; maximum width isacross the base, whence the straight or slightly convex edges taperevenly to the tip. The cross section is flattened lenticular. Flakingis generally very good and, save in two or three specimens, coversboth faces. Length is from 14 to 34 mm., width from 9 to 17 mm.,thickness from 2 to 4i^ mm. ; the ratio of width to length varies from1 : 1 to 1 : 3.4 with 1 : 2 probably near the average figure (pi. 37, a-i).Conforming in most particulars of size, shape, and workmanshipto the above, were 38 points (18.1 percent of all) provided with twolateral notches from 4 to 8 mm. above the base. Here again maximumwidth is across the base, which is usually faintly concave; slightlyconvex edges decidedly predominate over straight. Six other speci-mens (2.9 percent of all) have a single basal notch in addition to thelateral pair (pi. 37, j-n) .Less numerous, but still occurring in nearly all parts of the work-ings, were 14 cruder points. These vary from subtriangular withgreatest width at the base (6 specimens) to broadly leaf shaped withmaximum width one-fourth to one-half the distance above the base(8 specimens). They are much thicker (4.5 to 7 mm.) than the fore-going, and have coarser chipping which is usually restricted on oneor even both faces to the margins. The points also tend to be mis-shapen and asymmetrical, often retaining the curvature of the originalflake. They are also larger: width is 13 to 20 mm., length 21 to 38mm. It is possible that these shoddy specimens were not even in-tended as projectile points, though their general shape and size sug-gests some such use. 266 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174The remainmg 18 points comprise two groups of rather distinctivecharacter. The first group includes 1 incomplete and 10 whole speci-mens that are basically triangular, with the greatest width across theusually concave base ; the edges curve evenly from base to tip, and thecross section is lenticular. Each has a pair of deep, narrow, lateralnotches 7-18 mm. above the base. The edges are minutely retouched ; elsewhere, the chipping is coarser than in the small points, but stillof good quality. They are well made and symmetrical, are of theusual materials for the site, and look like large-scale copies of thesmall notched points. Length is from 34 to 94 mm.; width, from21 to 36 mm. ; thickness, from 5 to 8 mm. (pi. 37, o, ^, r; fig. 53, a) . Ten of the above points are from basin 1, mound 17 ; the eleventh,cataloged as from "loose earth at east end" of basin 2, may actuallyhave come from the immediately contiguous part of basin 1. "Whetherall actually served as projectile points, I cannot say; our best ex-ample, of chalcedony (pi. 37, o), seems too fragile to have been soused. The basal half of one and the tip of another have a carbo-naceous adhesion suggesting burnt pitch, resin, or grease.The second group includes 7 stemmed specimens, all from basin 2.Though variable in size and other details, they are somewhat smaller,cruder, and proportionately heavier than the foregoing. Length is32 to 60 mm. ; width, 18 to 34 mm. ; thickness, 5 to 9 mm. They arecharacterized by a relatively broad, thick blade with convex edgesand well-marked shoulders, deep corner notches that produce anexpanding stem, and a base that is usually convex (5 specimens) orslightly concave. The flaking is coarser than on most other specimensfrom the site. All specimens have the blade edges ground down, andin most cases are so blunted as to be useless for cutting or penetrat-ing?a curious feature not noted on any other projectile points fromthe site. The blunting, if not deliberate, probably resulted from useof the blades as scrapers or reamers, i. e., with a sidewise rather thanlengthwise (or cutting) motion. They may have been specializedtools whose resemblance to projectile points is incidental. They are,however, reminiscent of a widespread Plains type (cf. pi. 37, q, s) ofWoodland point and there is a chance they were not made at theTobias site but were picked up on some older campground and re-adapted to another use by the later peoples. I am unable to suggesta plausible explanation for their restricted occurrence in basin 2, orfor the similarly limited occurrence of the large notched triangularsin basin 1, In both basins, as for the site generally, smaller usuallyunnotched forms were characteristic.Drills.?On the basis of form, the 93 objects classed as drills fromthe site may be divided into two major groups, viz, (a) plain-shafteddrills, and ( b ) drills with shafts widened at base. Specimens belong- Wkdel] AN INTRODUCTION TO KANSAS ARCHEOLOGY 267ing to the first group, which comprises two-thirds of the total, havethe maximum diameter at or near the middle, whence they taperevenly in each direction to a rounded (originally pointed?) tip.Large, heavy examples predominate; they are from 51 to 120 mm.long, 8 to 15 mm. in maximum width, and 7 to 10 mm. thick, withrhomboidal or lozenge-shaped cross section. Neither end shows anyspecialization; wear occurs about equally on both ends and all alongboth edges, the latter being generally so blunted by use as to be quiteineffectual today for cutting or scraping. The size, length, andgeneral conformation of a number of these tools is such that they fitnicely into stone pipe bowels of the local varieties, and the verymarked degree of wear on the blade edges argues for just the sortof motion involved in drilling or reaming out pipe cavities (cf.Udden, 1900, p. 47). There is no indication that, or how, these ob-jects were hafted, though the possibility exists that they weremounted, perhaps like projectile points, on wooden shafts. It shouldbe noted that the overwhelming majority of these heavy-duty drillscame from basin 2, mound 17, and in lesser numbers from the pits,and that over half are broken specimens (pi. 38, h-^m; fig. 49, a-c).Nine specimens, including 6 fragments, resemble the above exceptthat they have at one end a thinned stem 7 to 13 mm. long that is a bedFigure 49.?Plain-shafted and T-shapcd drills from Tobias site. Length oi h, 11 cm. 268 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174 set off from the blade by well-marked shoulders. The blade edgesare worn smooth, but the stem edges are not. One specmien has sidenotches instead of a stem. Complete examples are 54 to 88 mm. long,in other dimensions like the foregoing (pi. 38, h-g) .Three smaller drills each have one pointed and one romided end.On the pointed end, 10 to 12 mm. long, the chipping is fresh and un-worn ; elsewhere, the remaining portion has worn edges. These, too,are classed here as stenuned forms on the assumption that the unevenpoint was designed for mounting in a socketed shaft. Length isfrom 35 to 39 mm. ; width, 6 to 9 mm. ; and thickness, 3.5 to 4.5 mm.Drills of the second group, 21 in number, are much more variablein form. In contrast to the basically rodlike build of the precedingvarieties, these consist of a slender shaft with slight, moderate, ormarked flangelike enlargement at the base. They are generally moredelicate and were evidently designed for less strenuous jobs. Fourspecimens, otherwise with little in common, widen gradually from aretouched point to the butt (pi. 39, o). Twenty-one others consti-tute a fairly consistent and well-standardized type (pi. 39, a-h) thatwas represented in practically all parts of our excavations. Theshafts are thin, either tapered or parallel sided, 15 to 26 mm. long,and biconvex or lozenge shaped in cross section. Bases are wide,relatively thin, irregular in shape, and usually have been retouchedonly on edges immediately adjoining the well-made shafts. Onebears the broken stump of a second drill shaft; another has a well-made graverlike comer at one side. In specimens that can be con-sidered complete, overall length is 32 to 52 mm., flange width 16 to36 mm., and shaft length as given above. I do not believe theseobjects were ever hafted; held between the fingers and used withcare, they might have served for boring small holes in wood or shell,or for piercing leather. They are known from related sites in thearea.A single specimen can be classed as abrupt widening with smallflange. It is T-shaped, 70 mm. long, with a shaft rhomboidal incross section and a 21-mm. crutchlike base. The midportion of theshaft, in common with several others from basin 2, mound 17, isstained by resin or other carbonized matter (pi. 39, Z; fig. 49, d).In addition to the above groups, there are 5 drills that may betermed uncommon forms. One (pi. 39, m) has a slender slightlyworn point 19 mm. long, widening abruptly into a 7-mm. thick flangewith oppositely beveled edges and concave base ; total length 60 nmi.Another delicately made specimen (pi. 39, i) with tapered point, rec-tangular flange, and concave base, is 33 mm. long, 10 mm. wide, and2.5 mm. thick. A third specimen, unfortunately incomplete, indi-cates that planoconvex end scrapers were occasionally provided at Wbdbl] an introduction TX) KANSAS ARCHEOLOGY 269the smaller end with a drill point?a combination that I have seenoccasionally in local collections from other central Kansas sites (seealso, Udden, 1900, p. 46).Our investigations indicate that the two drill types most repre-sentative of the Tobias site are the large heavy plain-shafted, oc-casionally stemmed objects for which the term "pipe-drill" may besuggested ; and the abrupt widening, large-flanged form. Both thesevarieties were fomid m several pits, in the mounds, and in the basins.The exceptionally heavy concentration of "pipe-drills" in basin 2 (52percent of all) is curious, but no ready explanation presents itselfunless it be that the structure represented a workshop or communitygathering place for the males or special artisans of the village.Knives with unbeveled edges.?^What must have been, to judge fromthe fragments found, a fairly common artifact type is very in-adequately represented in our collections. There are about 20 mid-or end-section pieces of elliptical, ovate, or otherwise oblong-roundedforms, with the edges evenly thinned and sharpened by retouchingfrom both faces. These fragments rarely exceed 7 to 10 mm. in thick-ness ; and this, relative to their width and presumed length, probablymade them fragile and easily broken. The exact shape and dimen-sions of the knives is, of course, conjectural, but I would suspect thatthey approached in these particulars the oblong round-ended specimendescribed by Udden (1900, p. 39 and fig. 15) as nearly 5 inches inlength, 2 inches in width, not more than a quarter of an inch in thick-ness, with an even, sharp edge all around. That larger blades werealso made is indicated by a complete specimen from basin 1, mound17 (pi. 46) , which measures 198 by 61 by 9 mm. The edges have beencarefully thinned ; at the narrow end they are blunted by grinding toa distance of 45 mm. from the tip (fig. 51, &) . The material is a pinkand gray-banded chert containing fusulinid fossils, and is probablyfrom the Florence flint of southern Kansas. A somewhat smaller in-complete specimen of gray chert from basin 2, mound 17, has a broadrounded end from which the sides converge evenly to a fracture ; it is120+ by 55 by 9 mm. There is no evidence that any of the abovespecimens were hafted.Three incomplete specimens designed for hafting may be notedhere. One, from pit 3A, looks like a large spear point; it has an ex-panding stem with blunted edges and convex-edged blade with tipmissing. The others, both from basm 2, mound 17, are end fragmentsof lanceolate or elliptical objects with lateral notches near the tip,whence the edges of the blade expand. They are 7 to 8 mm. thick, buttheir original length and width are not determinable. Except forthe notches, they are indistinguishable from the knife fragmentsdescribed above. 270 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174Knives with beveled edges.?This large and somewhat variableseries includes elongate, usually pointed, forms in which two con-tiguous edges have been oppositely and abruptly beveled. They in-clude notched and stemmed forms, as well as others not clearly pro-vided with means for hafting. Their use is problematical, and theseveral forms discussed here have been variously identified as spearpoints, knives, scraping tools, and other objects, by different workers(pi. 40).The most common form has a pair of broad shallow notches justabove a flattened, rounded, or pointed base; from the notches, theedges of the blade taper rapidly at first and then more slowly to thetip, thus forming a markedly concave curvature. The blade is usuallyquite slender, approximating 10? to 12? segments of a circle; and theopposite beveling of the edges results in a rhomboidal cross section(pi. 40, a~e). Rare specimens are broader, with convex edges thatattain their maximum breadth well above instead of just at thenotches (fig. 50). Sometimes one notch is missing. All specimens,when held with the tip up, show the beveling on the left edge of theblade whereas that on the right edge is not visible. Complete speci-mens seldom exceed 100 mm. in length, but there is one broken pieceof 130 mm. that may well have been several centimeters longer;width is from 25 to 40 mm., and thickness is usually under 8 mm.There are from the site 11 examples of this form, whole or with onlythe extremities missing, besides 6 basal fragments showing one or bothnotches. Udden (1900, p. 41 and pi. IV, 2, 4) reports similar objectsfrom the Paint Creek site, suggesting that the slender-bladed speci-mens attained this form by repeated sharpening through retoucliingof the beveled edges. I have seen others from the Gray site (proto-historic Pawnee) , near Schuyler, Nebr.Less common are unnotched, narrow-bladed objects with bases thatare simple rounded, tapered truncated, or parallel sided with taperedextremity. There are four of these, of which only one is complete(pi. 40, /, i; figs. 51, a, and 53, & ) . Finally, there are four other unnotched knives with wider beveledblades and a short tapered base that may or may not have bevelededges. The general outline is lozenge shaped to lanceolate; one ortwo suggest, but are not quite identical with, the so-called Haraheyknife (pi. 40, j). All are made of brown jasper; they are 80 to 127mm. long by 27 to 38 mm. wide. There is no evidence that these im-plements were hafted (fig. 51, c) . Thirty-four other fragments, broken from the blades of specimensof one of the foregoing types, cannot be classified other than asbeveled-edge artifacts. They range from very slender drill-like piecesto broad blades, with rhomboidal or lozenge-shaped cross section. Wedbl] AN INTRODUCTION TO KANSAS ARCHEOLOGY 271 I a Figure 50.?Chipped knives with two beveled edges and side notches, from Tobias site.Length of a, 15.5 cm. 484172?59 19 272 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174 a Figure 51.?Knives with unbeveled (b) and beveled {a, c) edges, Tobias site. Length ofb, 19.8 cm. mm aFigure 52.?Stemmed knives with beveled edges, Tobias site. Length of a, 9.5 cm. Wbdel] AN INTRODUCTION TO KANSAS ARCHEOLOGY 273Five are of brown jasper, the others of whitish, gray, or pink-bandedchert.There is, of coui-se, no sharp line between beveled and unbeveledknives, though the bulk of the beveled pieces are rather distinctivein shape as well as in the treatment of blade edges. From basin 2,mound 17, however, came a fire-fractured blade with one end and partof one side missing, and with smoothed lateral notches near the nar-row end. Present length is 155 mm., thickness about 10 mm., andwidth, if a measure of symmetry can be assumed, must have beenabout 55 mm. The shape and dimensions approximate those of thelarger unbeveled pieces. ?:^ ^' ~^?v^? l^ModeFigure S3.?Chipped point and beveled knives, Tobias site. Length of c, 10 cm.ETid scrapers.?Upward of 250 of these ubiquitous little objectswere found ; they came to light wherever we dug on the site. Theyare like those from hundreds of other sites in the Plains region:planoconvex in cross section, prominently ridged dorsally or lesscommonly (ca. 8 percent) low and flattened, varying in outline fromsubtriangular to subelliptical or subovate, and with the edges re-touched all around (pi. 41). The plane underside, representing theunmodified cleavage surface, is usually curved or "twisted" somewhat.Maximum thickness is near the wide end, which is always steeplysloped. Very commonly the edges are blunted. In size, they rangefrom 17 to 97 mm. long, 14 to 40 mm. wide, and 5 to 15 mm. thick.In the longer specimens, width-length ratio is about 1 : 2.5, whereas 274 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174in the shorter examples it is about 3 : 4. Small scrapers predominate.Of 259 specimens, 98 (38 percent) are under 30 mm. long, and 192(74+ percent) are under 38 mm. This compares favorably with thecoeval Leary site (Oneota) where 346 of 437 scrapers, or 79+ percent,were under 38 mm. long. By contrast, the "El Quartelejo" site (Dis-mal River) in Scott County yielded 239 end scrapers, of which only79 (33 percent) lie within this size range. Comparable series fromother sites and horizons are either not available or still await pub-lication, so comparisons with Lower Loup sites, for examf)le, cannotbe made. In earlier horizons, as the Upper Republican, I have theimpression that larger specimens are the rule. Just what signifi-cance, if any, these variations in size may have in terms of functionI cannot say, but should future work bear out the suggested sizedecrease in protohistoric times, some explanation other than merechance would seem to be called for.Side scrapers.?This is an exceedingly varied series of implementswhich I have not attempted to subdivide. They are made from thin,sharp-edged flakes and spalls of all sizes and shapes, converted intocutting or scraping tools through fine retouching of one or moreedges. Retouching is usually confined to the extreme edge, whichin some of the heavier specimens has been chipped back into a steeplypitched flat-bottomed working surface; it is always limited to theconvex surface of the flake (fig. 54) . The wide variety in form, size, <-\ a bFigure 54.?Side scrapers from Tobias site. Length of i, 10.8 cm. Wkdel] an introduction TO KANSAS ARCHEOLOGY 275and material suggests that the flakes were byproducts or improvisa-tions rather than intentionally struck blanks. A high proportion ofthe side scrapers is of banded foraminiferal Florence flint, thoughpractically all other chippable stone used on the site is alsorepresented.Side scrapers vary from small thin slightly curved lanceolate or el-liptical pieces 30 or 40 mm. long by 15 or 20 mm. wide with all aroundretouching (flake knives?) to long subelliptical, subovate, or irregu-lar spalls 120 to 140 mm. by 25 to 60 mm., with only one straight orcurved edge retouched. Single-edge retouching is most common, buta number show two or more distinct working edges. Some are lowand flat in cross section, others have a prominent ridge along themidline or at one side. Most extensively modified are a few narrowcurved and keeled forms somewhat resembling the end scrapers, butwith the long edges only retouched or worn. There is no evidenceof hafting, and most or all were probably held directly in the hand.Approximately 250 specimens are classed as side scrapers ; they wereat least as plentiful as end scrapers, and it is possible that becauseof their less regular shape and limited retouching others were over-looked in the general digging.Axes.?These are large, thick, coarsely chipped ovate to subellipticalobjects, with two broad notches that have been worn or batteredsmooth. The blades are generally sharp, but sometimes show finelychipped or crumbled edges that suggest heavy or long use. They wereshaped solely by percussion flaking, I think, and show none of the finepressure retouching of the projectile points, knives, and scrapers.Length is from 85 to 156 mm., width from 63 to 97 mm., and thick-ness from 20 to 36 mm. There are only 8 examples, most of themfrom various parts of the mound 17 complex. It appears thus thathafted chopping tools were much less important in native technologythan were crushing implements of the hammer type.Cache flints.?Elsewhere the finding of a cache of flints in basin 1,moimd 17, has been noted (p. 221). The great majority were flakesand spalls, of irregular shapes and sizes, from 20 to 100 mm. ingreatest dimension. With exception of perhaps a dozen specimens,none show secondary chipping. The material includes some brownjasper and pink foraminiferal Florence flint, but 95 percent ormore is a blue-gray, often lightly mottled, stone whose exact sourceis undetermined. The pieces are of sizes suitable for manufactureof projectile points, scrapers, knives, and other objects and undoubt-edly represent the raw materials store of some long-dead flintworker.Besides the unworked flakes, the deposit yielded six side scrapers,five low-backed end scrapers, two small bits of beveled implements,and a large ovate notched ax blade. All were of the same blue-graychert as the flakes. 276 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174OBJECTS OF GEOUND AND PECKED STONEOur collections indicate that there was a rather extensive utilizationof various artifact types made by hammer pecking and/or grindingtechniques. Unfortunately, it has not been possible to submit theartifacts for examination by geologists familiar with the immediateregion, but it is evident that both local and nonlocal materials wereemployed. As indicated elsewhere, there are ledges of Dakota sand-stone immediately north and east of the site, and this material wasthus easily available in unlimited quantities. Whether, or to whatextent, quartzite boulders occur in the Rice County area I am notprepared to say, but the presence of Sioux quartzite probably indi-cates contact with the lower Kansas River valley below Manhattan.Limestone was not extensively used ; it may have been obtained fromsoutheastern Rice Comity or elsewhere from Permian outcrops tothe east or southeast a few miles distant from the site. It is clear,at any rate, that the toohnakers of the Tobias site were not restrictedin tlieir use of raw stone to the materials available only at theirdoorsteps. (See table 6.)Grooved, mauls.?There are 19 examples of these objects from theTobias site, if we include two or three spalls too fragmentary to befurther classified as to shape. Thirteen are made of quartzite ; abouthalf of these are of pinkish or purplish-red stone which I suppose isSioux quartzite from glacial deposits in northeastern Kansas. Gran-ite, sandstone, and limestone are each represented by two specimens.There is considerable variation in size, shape, weight, and qualityof workmanship, and the specific uses to which they were put, otherthan the general one of pomiding, evidently varied. In the followingdiscussion, I have divided them on the basis of form rather thanaccording to assumed use.Group I includes three exceptionally well-made specimens of Siouxquartzite, all from the mound 17 complex: one came from each ofthe basins, 1 and 2, and the third from basin 2, pit E (pi. 42, a, 5).All are cylindrical or very nearly so, the sides meeting the circular,flattened striking surfaces in a well-defined rounded angle. A grooveabout 5 mm. deep by 12 to 15 mm. wide encircles the middle; it isusually unmodified, but in one instance is formed by two low parallelridges or flanges rising from the body of the maul. The strikingsurfaces, already described as flattened, actually vary from slightlyconvex to very slightly concave; and the surfaces generally arefinely "dimpled" from the shaping process. They range in lengthfrom 95 to 100 mm., in maximum diameter from 89 to 97 mm., andin weight from 44 to 58 ounces. None shows evidence of heavy use,and the striking surfaces are undamaged. If intended as utilitytools, they cannot have been used long or roughly ; and since similar WBDKr,] AN INTRODUCTION TO KANSAS ARCHEOLOGY 277 ;+Mr-c?5i-l'^ ,^^ I o.a ' '2 ' o-s I 05 I 12 - ?" M ? CO ^2g-Sby^2S5Sgo^'a'ows; ; PI a K g'H fc w ? aaw Pa 2-2< t I I O) bC_;B>^a.g 278 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174well-made symmetrical and undamaged specimens have been fomidat other nearby sites (pp. 276, 337; also Brower, 1899, pi. 11; andUdden, 1900, pi. 5), it is possible that they were nonutilitarian.Group II is represented by a single specimen of vesicular grayquartzite, found in mound 17, basin 2. It is subcylindrical ; the pollsare rounded, and a gi'oove encircles the middle. It is 98 mm. long,86 mm. in diameter, and weighs 32 ounces. In general appearanceand size it approaches specimens in the preceding group, but wasmade with much less care and little regard for symmetry or ap-pearance. Its irregular conformation, even with due allowance fordamage sustained in use, results basically from utilization of aboulder which was only partly dressed between striking surfaces,and so still retains here and there sizable areas of the original boulderface. The type is shown in plate 42, c, e.Group III includes four complete specimens, three of them frommound 17, basin 2, and the fourth from mound 6. Two are of quartz-ite, one of sandstone, and the mound 6 specimen is of granite. Theydiffer markedly from the preceding groups in being somewhat com-pressed laterally with an ovoid or broadly elliptical outline. Thepolls are more strongly rounded, usually diminishing in diameterfrom the margins of the groove so that the side walls are very shortor even indistinguishable. One specimen has the poll margins some-what battered, but in general there is little or no evidence of breakagefrom use. Two have strongly dimpled surfaces, showing that theywere hammer pecked into their final shape. Measurements are:length, 83 to 173 mm. ; width, 90 to 138 mm. ; thickness, 65 to 102 mm.Weight ranges from 28 to 106 ounces (pi. 42, d^g).Two spalls, one of granite from mound 17, basin 2, the other ofquartzite from pit 5, each show traces of a groove and a small sectionof rounded poll. They probably are pieces broken from mauls ofthe same type.Group IV includes one small specimen of fine-grained red quartzitefrom mound 17, basin 2. In general it may be characterized astapered; one poll is circular and flattened, with distinct side walls,whereas the other is short and rounded, with no trace of straightwalls. It was fashioned apparently from an asymmetrical boulder,whose form may have partially dictated the shape of the finishedartifact; the groove is shallow and indistinct on about half the speci-men because of a flaw in the stone. In contrast to the precedinggroups, where the groove encircles the specimens at the middle, thepresent example has the gi'oove nearer the smaller rounded end.^^It is possible that this end was completely covered by a rawhide ^ ResemblluB in this respect and In Its longitudinal aRymmetry numerous northernPlains mauls, large and small, in the Division of Ethnology, U. S. National Museum. IWedbl] an introduction TO KANSAS ARCHEOLOGY 279binding and the flat poll used for pounding pemmican or in somesimilar activity. The piece is 79 mm. long, has a maximum diameterof 89 mm., and weighs 27 ounces (pi. 42, /) . In addition to the foregoing varieties, there are eight specimensso badly battered or split as to resist classification. One or twovaguely suggest group II, but with large fragments split away fromone or both polls; in two cases, this spalling has almost completelydestroyed one poll. Most of the specimens seem to have been some-what flattened and irregular, originally perhaps like group III. Inany case, whether they served in stoneworking, chert quarrying, stakedriving, crushing of heavy bones, or for other purposes, it is clearthat they have been subjected to much rougher or more prolongedduty than any of the classified specimens described above. Five areof quartzite, two of limestone, and one of sandstone. Consideringtheir present fragmentary state, measurements and weights give arather inaccurate picture, though they do afford some indication oftheir proportions relative to the complete specimens. In the lessfragmentary examples, length varies from 116 to 145 mm., maximumdiameter from 77 to 117 mm., and weight from 32 to 59 omices. Onespecimen comes from pit 3A; the others are from various parts ofthe mound 17 complex.To judge from recent ethnological examples, these implementswere mounted on wooden handles 14 to 24 inches long, which werebent around the stone at the groove and held together with a raw-hide covering. This covering, applied when green, commonly encasedthe middle or one-half of the head. It shrank on drying and thussecured the head.Grooved mauls in a variety of shapes and sizes, and doubtlessintended for diverse purposes, were widely used throughout thePlains area in historic and protohistoric times. So far as I knowthere has been no detailed analytical or distributional study of them ; none will be attempted here. Crude fragmentary specimens fromSignal Butte I (Strong, 1935, p. 230 and pi. 25, fig. 2, o), if correctlyidentified, would suggest a measure of antiquity. Curiously enough,they are exceedingly rare, if not indeed absent, from such subsequentprehistoric Central Plains horizons as the Woodland, Upper Repub-lican, and Nebraska Culture, although at Pecos, where they are notcommon, Kidder (1932, p. 55) reports them in pre-Glaze II and ap-parently also Black-on-white associations. I have been unable tofind any published record of the well-shaped cylindrical form (groupI, supra) outside of central Kansas, possibly excepting two fiLne speci-mens from the Hill and Palmer sites, both Pawnee, in Nebraska(Wedel, 1936, p. 77). Less carefully made mauls, often little morethan a suitably shaped or slightly worked boulder provided with a 280 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174groove, occur archeologically from Texas (Sayles, 1935, pi. 15 andMap E; Studer, 1934, p. 91 (?)) through Kansas (Martin, 1909, pi.9, fig. 79 ; Brower, 1899, pi. 11 ; Udden, 1900, pi. 5 ; Jones, 1929, p. 98)and Nebraska (Dunlevy, 1936, p. 196; Hill and Metcalf, 1942, p.193) to the Dakotas (Strong, 1940, p. 375; Will and Hecker, 1944,p. 25), and Montana (Mulloy, 1942, p. 57). As surface finds, thereare innumerable specimens throughout the Plains from Texas to theSaskatchewan Eiver. This brief and no doubt incomplete recitalof records does not pretend to delimit the area of occurrence, butit partially suggests its extent, and it is worth noting that most ofthe specific site records seem to apply to the protohistoric and his-toric periods. Also, they discredit the somewhat ambiguously phrasedand possibly misread suggestion by Over (1940, p. 336) that "SouthDakota appears to be about the center of the area in which they [i. e.,grooved mauls] are foimd . . .," and further, that ". . . Those foundoutside of the above area [i. e., south of Nebraska] were probablyexchange pieces traded with neighboring tribes. ..." I see no rea-son to doubt that most or all of the Tobias site specimens, abovedescribed, were made locally.Hammerstones.?There are but six examples of these in our collec-tions, though from the abundance of pecked and groimd objects atthe site I would suppose that hammerstones must have been presentin much larger numbers. Four are of chert, and two of quartzite.Those of chert, were originally somewhat flattish and subcircular, theothers irregularly oblong. All have battered ridges, angles, andperipheries, and have evidently seen much use against hard substances.One quartzite specimen has shallow smoothed pits suggesting fingerholds. They vary in maximum dimension between 50 and 82 mm.,with both quartzite specimens exceeding in size and bulk those ofchert. Coming from mounds, pits, and basins, they do not differsignificantly from the hammerstones to be foimd in and on camp andvillage sites of all periods in the Plains.Mealing stones and raanos.?That the inhabitants of the site wereaccustomed to grinding their corn on stone mills is attested by thepresence of a number of these specimens. The nether element, forwhich I deliberately choose the term mealing stone in preference tometate, is usually of hard dense Dakota sandstone ; the smaller upperor handstone is of sandstone, quartzite, or rarely limestone. In de-tails of form and in the inferred method of use there appear to havebeen some variations.The largest specimen in our series (USNM 388716; see pi. 43, g)has been carefully worked into an elliptical shape, with the uppersurface and edges smoothed so that no pecking scars remain. Theupper surface has an elliptical depression 35 cm. long by 20 cm. wide Wedbl] AN ESTTRODUCTIOlSr TO KANSAS ARCHEOLOGY 281by 1.2 cm. deep, inferentially from the use of a small handstone witha rotary grinding motion. The slab itself measures 65 by 43 by 9.5cm., and was taken from pit 5. A smaller, thicker, and less care-fully finished mealing stone, also with elliptical grinding cavity, wasfound in pit 3.A third stone (USNM 389013; see pi. 43, /) was evidently used indifferent fashion, to judge from the contours of the grinding surface.This is a rectangular sandstone block, with one end more or lesssquared and the other rounded. The edges have been somewhatundercut in the shaping process but, like the irregularly flattish under-side, have not been smoothed. The grinding surface is concavelengthwise and flat transversely, thus indicating beyond cavil a back-and-forth grinding motion with a long flat-bottomed mano. Thegi'inding surface is generally well smoothed from use, but has nu-merous shallow pits?perhaps from "roughing" to improve a surfacetoo smooth for effective grinding (cf . Kidder, 1932, p. 69) . The stonemeasures 40 by 29 by 8 cm. ; it was found in mound 17, basin 2, pitE, from which came also the largest mano we obtained.Manos are represented by seven complete and five fragmentary speci-mens. The former are generally symmetrical and well shaped; withone exception, they have but a single grinding surface. The largest ex-amples, of which there are three, are elongate-elliptical with nar-rowed rounding ends and planoconvex cross section (pi. 43, b, c) ;they range in length from 19.2 to 23.9 cm., in width from 8.2 to 9.8cm., and in thickness from 4.5 to 5.7 cm. Another group includesthree shorter specimens that are subrectangular with rounded corners,and subquadrilateral or bun shaped in transverse cross section (pi.43, d) ; they are 12.5 to 15.2 cm. long, 8 to 9.2 cm. wide, and 2.6 to 5.3cm. thick, the thinnest specimen being bifaced. The seventh wholespecimen is irregular in shape, but lies nearer the second group insize. Four of the fragments are ends, and they seem to be equallydivided between the two forms found among the complete spexiimens. FiGURE 55.?Mano from Tobias site. Length 12.5 cm. 282 BUREAU OF AlSfERICAN ETHNOLOGY [Bull. 174Eight of the manos are from various parts of the mound 17 com-plex, the others from pits 1, 3, and 5. The whole specimens, it shouldbe noted, are flat or very nearly so on the grinding face, and thereforemust have been used with a nether stone whose surface was trans-versely flattened; neither their shape nor size would work on eitherof the elliptically concave stones described above. For use with thelatter, I would suspect a yet smaller circular or bun-shaped muller,whose diameter would be somewhat less than that of the grindingcavities, but of these, there is no certain evidence in our collections(cf. pi. 43, e).Large sandstone mealing slabs and manos have been reportedfrom several other nearby sites in central Kansas (Udden, 1900, p.50; Brower, 1899, pi. 9; Jones, 1929, p. 98), and it is clear that theywere much used in the protohistoric period. Other finds suggest thatthey were known also in late precontact sites (Brower, 1899, pi. 9 top;Wedel, 1935, p. 229), though north of the State they become muchscarcer.Mortars ( ? ) .?These are oblong, squarish, or round blocks of sand-stone or limestone that have been hammer dressed to shape but werenot smoothed. They are smaller and proportionately thicker than themealing stones, ranging from 17 to about 30 cm. in maximum di-ameter, and from 7.5 to about 12 cm. in thickness. Each has on onesurface a shallow circular concavity 12 to 14 cm. across by about 1cm. deep ; one has a second cavitj^, slightly smaller and shallower, onthe opposite face. These depressions, evidently pecked out with ahammerstone, are evenly contoured but have not been rubbed smooth ; and I see no reason for supposing that any of them were intended toreceive a manolike or muUerlike handstone. It is possible that theywere anvils, set on rawhide, and used in pulverizing meat, berries, andother food with a stone hammer. Two of our specimens (pi. 43, a, h)are from mound 17, basin 2; the third is from pit 3.DisGoidal rubhing stone.?This is a circular or bun-shaped quartz-ite object, with one surface slightly convex and worn, the other mark-edly convex and dressed but not worn. Measuring 10 by 4.8 cm., itfits the hand nicely when held with the worn side down. It is pos-sible that this sort of handstone was used with the mealing slabswhich have an elliptically hollowed surface, or it may have been forrubbing hides. The specimen is from pit B. ^''Sinew-stones.''''?This terai is applied to two specimens from pit 3,but it rests on no documentary or other direct evidence and may be in-correct. Both are of quartzite. One is circular in cross section withone side flattened and the other thicldy convex; it measures 6 by 9cm. On the convex surface is a shallow pit, on either side of which ashallow groove runs part way down the sides of the stone to match, Wedbl] an introduction TO KANSAS ARCHEOLOGY 283but not quite meet, sliallow notches worn into opposite edges of theflattened surface. Suggestions of additional notches occur elsewherearound the edge of this face, all showing rather more evidence of wearthan those on the pitted surface. The second spechnen is somewhatrectanguloid in outline with elliptical cross section and suggests asmall muller, but has not been so used. On the upper outer surfaceis a well-worn notch 7 to 8 mm. in maximum depth by 30 or 35 mm.wide. In both specimens the notches or grooves tend to be broadlyV-shaped and seem to have been produced by the action of a cord,thong, or other soft substance drawn repeatedly through them. Theydiffer markedly from the shaft-smoothers and polishers found on thesite, and must have had a different use.Shaft smoothers.?This widespread Plains artifact type was abun-dantly represented everywhere in our excavations, there being 11 com-plete specimens and 90 or more fragments. All are of yellowish toreddish-brown Dakota sandstone with excellent abrasive qualities.There is, of course, much variability in size and shape, but in generalthey appear to have been somewhat wider than thick, with the endssquared or rounded, a flat, well-finished, longitudinally grooved sur-face, and the sides and bottom somewhat rounded (pi. 44, g). Thegroove varies from a thin straight incision, suggesting a guide line,to a deep U-shaped channel, the deeper examples all being of a sizeto accommodate easily an ordinary round lead pencil (diameter l^inch) . The complete specimens are from 40 to 120 mm. long, 20 to 40mm. wide, and 13 to 26 mm. thick. An occasional fragment has asecond groove on the flattened underside, or else fine deep grooves onone or both of the narrow lateral surfaces.These objects undoubtedly were for the purpose of shaping andsmoothing wooden arrowshafts. They were used in pairs, with theflattened and grooved surfaces held together and the stick rotatedbetween them. La Flesche (1924, p. 113) describes the process amongthe Omaha as follows : ... A good arrowmaker aims to make the shaft as nearly cylindrical aspossible. To accomplish this, he holds the shaft in his left hand between thesandstone polishers, each piece grooved lengthwise, and gives the stick a twirl-ing motion by rolling one end of it back and forth on his thigh with the palmof his right hand. He shifts the polishers along the shaft in order to keep ituniform in size. When one end is polished, he works in the same manner onthe other end, until the full length of the shaft is round, smooth and uniform.All of our finds at the Tobias site were of single specimens, but fromthe nearby Malone site (p. 338) came a shaped sandstone block groovedfor splitting that shows both the method of manufacture and thefact of pairing (USNM 389314; pi. 44, ^). Matched sets and shapedbut unsplit blocks, as well as numerous single examples, have beenfound in Pawnee and Oneota sites in Nebraska (Wedel, 1936, p. 80; 284 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174Hill and Wedel, 1936, p. 45), and the type has a very wide archeo-logical distribution in and about the Plains. Apparently, it extendsback in time at least as far as Signal Butte I (Strong, 1935, p. 230) ; and it is common in practically all known subsequent horizons exceptthe Woodland.Shaft polishers ( ? ) .?Here are included three whole and brokenlimestone blocks, each of which has one grooved surface. One, frommound 17, basin 2 (pi. 42, i) , is of gray crystalline calcite, irregularin outline with two flat smoothed surfaces parallel to one another.Across the larger surface is a straight shallow groove 55 mm. long by9 mm. wide by 3 mm. deep. The specimen itself measures 71 by 59by 49 mm. Another is the half of a broken circular block 78 mm, indiameter by 47 mm. thick. The sides and bottom are roughly dressedto shape, and the top has been smoothed. A groove 13 mm. wide by4 mm. deep has been cut across the upper surface, then hollowedslightly and polished smooth by a rodlike object, except for a narrowshelf along each side where the cutting scars remain. On one side thegroove is bordered by a narrow ridge, beyond which the surface of thestone has been scraped away to the approximate depth of the originalcut groove. This specimen (fig. 56, a) is from pit 3A. The third,from pit 6, is the planoconvex upper half of a horizontally split ob-long cobble. The flat underside has been dressed to an even surfacebut not smoothed; the upper convex surface is smoothed, and istraversed by a narrow, shallow groove showing unmistakable, but notmarked, evidence of rubbing.These fine-grained grooved implements are something quite dif-ferent from the coarse-grained shaft smoothers previously described Figure 56.?Ridged and grcxjved shaft polishers from Tobias (a) and Paint Creek {b) sites.(Jb, Courtesy of J. H. Fries, McPherson, Kans.) Wkdbl] an introduction TO KANSAS ARCHEOLOGY 285 as used in pairs ; and tliey have a vague resemblance to the arrowshaftstraighteners of the Southwest (Kidder, 1932, p. 76). Whether theywere used in the same way as the latter, I am unable to say. Thatspecimens of Southwestern origin were not unknown to the Great Bendpeoples is suggested by a fine piece (pi. 42, h) that came to my atten-tion in 1953 through Mr. J. H. Fries of McPherson, Kans. This isa round flat block of dark-green steatite, 75 to 85 mm. in diameterand about half as thick. The groove is 15 mm. wide by 9 mm. deep ; and 67 notches have been cut into the periphery of the block on itsgrooved surface. A ridge 30 mm. long parallels the middle of thegroove at a little distance, and this ridge also has 6 notches (fig.56, h).According to Mr. Fries (letters of February 6 and 16, 1953), thisobject was picked up on the Paint Creek site, McPherson County, "near the center of the council-circle," which has gone under theplow since I first saw it. It recalls to mind some of the shaft straight-eners reported from Pecos (Kidder, 1932, pp. 76-80) . Only six round-bodied specimens are there reported; and ". . . all the ridged oneswhich could with certainty be dated were taken from Glaze V andlater rubbish." The Paint Creek piece does not exactly match anyof the round blocks figured by Kidder, but it conforms in all partic-ulars to the type represented at Pecos. It is an alien piece in thePlains and was probably carried into central Kansas from the Puebloarea.Sharpening stones {hones).?Several irregular sandstone frag-ments, some of which may be greatly altered bits of shaft smoothers,have evidently been used for sharpening awls, needles, or other sim-ilar slender-pointed objects of bone or wood. They bear grooveswhich, imlike those of the shaft smoothers, are not straight or ofuniform width and depth; and, as often as not, the grooves endabruptly short of the end of the stone. Sometimes the grooves con-verge or criss cross, and they are generally short, narrow, and deep.These could not possibly have served in the shaping or smoothingof rodlike objects, but they could and probably did serve efficientlyfor sharpening pointed implements (pi. 44, 7, A% and fig. 57).Stone pipes.?Six complete pipes, representing at least three dis-tinct varieties, were found, besides which there are perhaps 15 frag-ments probably or possibly from additional specimens. Of the wholepipes, four are from mound 17, basin 2 ; pits 3 and 4 each yielded one.The usual material was catlinite or a fine-grained red to purple-redsandstone; there is one of steatite. It is presumed that all wereprovided with wooden, bone, or reed stems when in use, but of theseno trace was found. 286 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174 .,^^' '.?.,.-s^s^ \ \ ^-K^ X^.ff i Figure 57.?Sandstone sharpening blocks from Tobias site.The prevalent form of pipe was L-shaped, with the bowl exceedingthe stem arm in length (pi. 45, a-c). The under side of the stemarm and the side of the bowl away from the smoker are character-istically straight, or nearly so, and meet at a right angle. The sideof the bowl toward the smoker, on the other hand, is rather markedlyconvex or bulbous, and meets the upper side of the stem arm at anacute angle. Stem and bowl alike are conically bored, probably withflint drills; and the walls, from 1 to 3 mm. thick, are usually wellsmoothed. Decoration, to judge from our specimens and a numberof others seen in local collections, is comparatively rare and simple;when present, it consists of a slight swelling near the top of the bowl,or of a narrow incised line or lines, or of narrow single or doublebeading about the outer lip of the bowl. Dimensions and proportions Wbdel] an introduction TO KANSAS ARCHEOLOGY 287 vary, of course, and our two specimens, though typical in form, donot show the full size range. In the pipe from pit 3 the bowl is 58mm. long, with a maximum diameter of 15 mm., and a maximumbore diameter of 10 mm. ; the stem arm is 21 mm. long and 14 mm.in greatest diameter. The second specimen, from mound 17, basin 2,has a smoke-blackened bowl 37 mm. high by 16 mm. in diameter,and a stem 22 mm. long by 14 mm. in diameter. The largest exampleof the type I have yet seen (pi. 45, a) is said to have been found onthe surface a few hundred yards southwest of the site, and is nowin the private collection of Lowell Peverley of Geneseo. Made offine-grained dark-red sandstone, it has a bowl 83 mm. long by 19 mm.in diameter, and a tapering stem arm 36 mm. long.In addition to the two whole specimens, we have at least 6 frag-ments that are almost certainly from elbow pipes in which the stemand bowl arms met at approximately a 90? angle. Neither arm issufficiently represented, however, to warrant the unqualified assertionthat they are of the foregoing type, though they may well be.Two whole pipes, both from mound 17, basin 2, are characterizedby bulbous bowls set approximately at right angle on stems that pro-ject beyond the bowl (pi. 45, d, e) ; both are of catlinite. In thelarger specimen, the stem is 85 mm. long and tapers to a horizontallyflattened "prow"; there is a laterally perforated crest 13 mm. longon the upper side above the stem opening, perhaps for attachmentby a cord to the erstwhile wooden stem or to a sacred bundle. Theblackened and "caked" bowl is somewhat compressed laterally, witha height of 25 mm. and a maximum diameter of 26 mm. The pieceis well smoothed and polished. The second specimen is much smaller,but is also well finished and polished from use. The stem is 33 mm.long, with a deep encircling groove through which, on the upperside, a small hole has been bored into the stem cavity. There is someevidence that the stem was once longer. The bowl is 20 mm. indiameter.The remaining two whole pipes are tubular in form (pi. 45, A, i).The larger specimen, from pit 4, is of catlinite ; it is 77 mm. long andhas a maximum diameter of 25 mm., whence it diminishes in a gentlecurve to 21 mm. at the bowl end and to 14 mm. at the bit end. Thebore tapers evenly from 15 mm. in diameter at the bowl end to 12 mm.at the stem end. Grinding striations are still visible, and there is nodecoration." The second example, from mound 17, basin 2, is of *5 A similar pipe in the Lowell Peverley collection is said to have been found about one-half mile northwest of the Tobias site, dissociated from any other evidences of aboriginalactivity. It is of dark purplish-red sandstone ; measures 92 by 21 mm. ; and has a lowbeading at the bowl end with 3 parallel encircling incisions. About 25 mm. from thestem end, the drill broke out through the side of the pipe. The specimen Is Illustratedin plate 45, g.484172?5S 20 288 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174 steatite. It tapers very slightly from a circular bowl end 20 mm. indiameter to a diamond-shaped stem 20 mm. wide by 15 mm. thick;total length is 70 mm. The bore is straight-sided, diminishing from15 to 12 mm. in diameter; at the bowl end there are deep unevenlyspaced lengthwise scorings suggesting that the maker was tryingto enlarge the cavity. The method of boring and the implement usedare not indicated. The exterior of the pipe is well polished, thoughsome longitudinal striae are still visible.Nine fragments of catlinite and sandstone appear to have beenbroken either from the bowls of L-shaped pipes or from the sides ofswollen tubular specimens. They vary in length from 20 to 50 mm. ; all are transversely curved and in four cases the longitudinal concav-ity is blackened or "caked" with carbon. One piece from pit 1 in-cludes a short segment of the lip ; it has 7 short nicks spaced at 1-mm.intervals on the lip exterior, below which are vestiges of two encirclingincisions. Another piece from pit lA shows a slight offset 9 mm.below the lip, reminiscent of that on our larger L-shaped specimen.Two fragments show some evidence of grinding or rubbing along thefracture edges, as if for reworking and secondary utilization.There is, unfortunately, no detailed distributional study of pipetypes in the Plains area, despite the fact that a wealth of comparativearcheological material exists in various museums and private collec-tions. Catlinite is exceedingly rare in prehistoric horizons of theCentral Plains. Archeologically, stone tubes appear to have beenuncommon in the area, though there are records of specimens, in-cluding one of steatite or similar material, from Upper Republican(Wedel, 1935, p. 200) and Dismal River (Hill and Metcalf, 1942, p.187) sites in southwestern Nebraska. Elbow pipes of various forms,including bulbous-bowled specimens with short stems that projectslightly beyond the bowl, are much more common, particularly inUpper Republican and related sites (Strong, 1935, pi. 16, fig. 1) . Thelong-stemmed variety with perforated crest appears to be a later, pos-sibly protohistoric and historic, development; whether it antedates thehistoric calumet type of bowl, which it somewhat resembles, I do notknow. The high-bowled short-stemmed L-shaped pipe occurs in con-siderable numbers in the protohistoric sites of central Kansas (Udden,1900, p. 57 ; Jones, 1929, p. 124) , eastward at least to Marion Countyand south to Cowley County. Wliat appear to be virtually identicalspecimens have been reported from the Red River valley in LamarCounty, Tex. (Jackson, 1933, pi. 81 and pi. 15, No. 36; Sanders site ?)the Burkett site in Nebraska (Wedel, 1936, pi. 8, h) ; and Taos, N. Mex.(West, 1934, pt. 2, pi. 195, fig. 9). In the Kansas-Nebraska region,so far as present evidence is concerned, the type is protohistoric, andits greatest concentration is in and just east of the great bend of the Wbdbl] an introduction TO KANSAS ARCHEOLOGY 289Arkansas, where it is associated with an archeological complex similarto that at the Tobias site.Gatlinite ohjects.?Other than pipes and pipe fragments, there arebut two specimens of worked catlinite from the site : one from mound4, the other from mound 6. The first is a small flat subrectangularpiece with slightly convex sides and rounded corners, and is smoothedbut not polished. It measures 30 by 18 by 3 mm. The second is ob-long, with one end somewhat tapered. It is concavoconvex in trans-verse section, and may have been readapted from a pipe fragment.It is 25 by 9 by 2.5 mm. Neither specimen is drilled or otherwisemarked on either surface, excepting a few striations, and the purposeof the objects is conjectural.Turquoise head.?Here may be noted the isolated find of a smallturquoise bead in mound 17, basin 1. It is of a good grade sky-blue stone, well made and polished. The bead is 2.5 mm. in diameter,0.6 mm. thick, and has a perforation 0.7 mm. across. It may havebeen part of the necklace described elsewhere (p. 296), but of this wecannot be sure. Undoubtedly, this and other turquoise beads fromthe site are of southwestern origin, probably traded to the Indians ofcentral Kansas as finished pieces.Sandstone disks.?There are eight of these curious objects (pi.44, a-e), all made of brown Dakota sandstone. Four fragmentaryspecimens were perforated more or less centrally, and in a fifth the bor-ing seems to have been begun but not carried to completion. All thefragments were broken through the perforation, and consist of half orless of the original disk. Thickness varies considerably, and the out-line seems to have approximated, but not very closely, a perfect circle.Eadius varies from 27 to 35 mm., thickness from 10 to 17 mm., andthe biconical perforation was 7 to 13 mm. across. The faces weredressed flat ; the edges were squared or rounded. There are no mark-ings of decorative or otherwise purposeful nature on any of the speci-mens. Two are from pit 1, one is from pit 3, and the fourth is frommound 17, basin 1.The incompletely drilled specimen is subcircular, with the edgesclearly ground down. On one face is the imprint of a fossil Sassa-fras ^* leaf, and the disk has obviously been shaped to include the en-tire leaf (pi. 44, h). On this face is a poorly centered hole about 10by 15 mm. across by 10 mm. deep ; on the opposite face is a shallowpit probably intended ultimately to meet the first hole. The speci-men measures 85 by 78 by 20 mm. ; it was found in pit 1.The three imperforate disks are smaller, but in all the faces havebeen dressed or smoothed and the edges rounded off. Thickness " Identmcatlon by Dr. Roland W. Brown, U. S. Geological Survey. 290 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174Varies from 8 to 12 mm., diameter from approximately 30 to 58 mm.Two of these are from mound 17, pit 8A ; the third is from pit IB.From pit 3A came a small circular sandstone object with biconvexcross section, and measuring 22 mm. in diameter by 12 mm. thick. Ihave not been able to satisfy myself to what extent the form is natural,i. e., concretionary, or artificial.Miscellaneous objects.?From mound 17, pit 8A, was taken a circularbiscuit-shaped object of sandy limestone averaging 44 mm. in diameterby 25 mm. in thickness. The flat surfaces have been smoothed, and oneis crossed by incised lines. A deep, narrow, undulating incision en-circles the specimen roughly midway between the two flat surfaces.I am unable to suggest what use this piece may have served.In the same pit was found a thick flat sandstone ring 36 mm. acrossby 12 to 16 mm. thick, with a 15 mm. cylindrical perforation. Theedges are squared in cross section and, like the surfaces, liave beenrubbed down.From pit IB came an irregular quadrilateral slab of fine-grainedwhite sandstone measuring 95 by 80 by 22 mm. The slightly hollowedand worn upper surface is stained dark red, presumably from thegrinding of pigment.Pit lA yielded a triangular wedge-shaped piece of sandstone, withboth faces moderately smoothed ; the thinned edge, 40 mm. long, sug-gests use as a saw or cutting implement.UNWORKED STONEConcretions.?Two specimens are included here. One, from pit 3,is bilobate with constricted midportion. It is 58 mm. long, and therehas been little, if any, modification of the natural form. The secondspecimen, from mound 17, basin 1, is subspheroidal but seems to havebeen modified somewhat. It is of fine-grained light-gray sandstone,and measures 52 by 45 mm. in diameter.Pseudomorphs.?There are three of these curious objects. One,from mound 6, is flattened subcircular, measures 48 by 35 mm., and isof brown sandstone. The second, from mound 17, basin 2, is ofpurplish sandstone, 24 to 26 mm. in diameter, and has 12 or more facets.It possibly is a pseudomorph after garnet.^^ The third piece, frommound 17, pit 8, is flattened circular in shape ; one half has a suggestionof facets like the preceding, the other half is smoothly convex. Itmeasures 58 by 47 mm.The purpose for which these pseudomorphs were carried into thevillage is, of course, not known. They may have been fetishes orgood-luck charms, but this is sheer guesswork. ? Identification by B. P. Henderson, U. S. National Museum. Wedbl] an introduction TO KANSAS ARCHEOLOGY 291Pigments.?A number of variously shaped irregular lumps of non-metallic hematite are striated and faceted from the removal by grind-ing of red pigment material. With one possible exception, all speci-mens appear to be natural lumps of mineral rather than preparedcakes. The possible exception, from mound 6, is a soft granularfragment with one rounded surface and an irregular fracture face ; itrubs off bright red on the skin. It may be nothing more than highlyferruginous weathered stone, but its dissimilarity to any other speci-mens from the site suggests that it may be a prepared ball or cake ofpaint material. OBJECTS OF SHELLThe shells of fresh-water mussels were surprisingly plentiful every-where in our excavations. Most are unworked, but artifacts made ofshell were rather more plentiful and varied than in most other knownCentral Plains horizons. There is no way of determining to whatextent, if any, shellfish formed a part of the native diet, or whether,as among some later Plains tribes, their use as food was frowned upon.Quite possibly children playing in the nearby stream carried anoccasional bivalve back to the village, but this is hardly an adequateexplanation. Since only a relatively small proportion of the shellshave been worked, it seems not unreasonable to explain the presenceof the remainder as a byproduct of the food quest. All of the formspresent are, or formerly were, locally abundant, and were probablygathered from the Little Arkansas River in the vicinity of the site.In order of their abundance, they include the following 10 species,with an asterisk precedmg those which have been recognized amongour worked specimens : *Proptera alata megaptera (Raf.)Uniomerus tetralasmus (Say)StropMtus rugosus (Swainson)*Quadrula pustulosa prasina (Conrad)*Pleurobema coccineum (Conrad)Quadrula quadrula (Raf.)*AmMema costata (Raf.)Elliptio dilatatus (Raf.)Ligumia subrostrata (Say)Lampsilis ventricosa occidens (Lea)Beads.?This includes all circular centrally perforated specimensunder 15 mm. in diameter. There are but four examples. Two, bothfrom mound 17, basin 1, are saucer-shaped, i. e., have a concavo-convexcross section, are circular or slightly elongate, and the central per-foration is about % the diameter of the bead. They are 7 mm. indiameter, and because of their thinness are quite fragile. The identityof the shell cannot be determined, but ih^ finished specimens closelyresemble certain types of Olivella beads of the Southwest (Kidder, 292 BUREAU OF AMERICAN ETHNOLOGY [Bull. 1741932, p. 185). On present evidence, it is impossible to be certainwhether our specimens are copies, trade pieces, or merely happen toresemble the types cited (pi. 47, c) . Two other beads from momid 17, one from basin 1, the other frombasin 2, are flat or nearly so, with dressed edges, and were evidentlycut from heavier pieces of unidentified shell. They are 12 to 14 mm.in diameter and 2 to 3 mm. thick. One has a 2.5-mm, hole, the otheris mibored. They closely resemble in size and appearance the speci-mens from the "shell bead burials" or ossuaries of southern Nebraska(Strong, 1935, pp. 116-124).Pendants of whole shell.?There are 11 of these, the only modifica-tion of the shell being that acliieved in the process of perforating forsuspension. Characteristically, the perforating was done by rubbingdown the back at, or immediately below, the umbo until a small usuallysomewhat irregular opening 4 to 10 mm. in maximum diameter hadbeen made into the interior (pi. 46, 2). Two specimens differ in thateach has a deep narrow cut transverse to the long axis of the shell, justbelow the umbo (pi. 47, ?, 5) . Shells under 45 to 50 mm. in maximumdiameter seem to have been preferred, though two broken specimensmust have considerably exceeded this figure. All are weathered andchalky, and little or no evidence of string wear can be detected at theedges of the perforations.Among the shells so used are Q. pustulosa prasina (7), P. coccineum(2) , and A. costata (1) . There are 5 specimens from mound 17, basin2 (including 2 from pit D), 3 from mound 17, basin 1, 2 from mound17, pit 8, and 1 from pit IB. Figure 58.?Notched and incised shell from basin 2, mound 17, Tobias site.(USNM 388931.) Actual size. Wbdel] an introduction to KANSAS ARCHEOLOGY 293Pendants of cut shell.?Pendants in which the original contours ofthe shell have been so far altered as to preclude species identificationinclude both circular and noncircular forms. All our examples, ofwhatever form, are from basins 1 and 2 of the mound 17 complex.Two are circular or approximately so (pi. 46, 2). The larger is ap-proximately 43 mm. across, less than 2 mm. thick, and has a slightlyconcavo-convex cross section. From the concave surface have beenbored, slightly off center, two holes 13 mm. apart and about the samedistance from the edge. The piece is well made and nicely finished.The second specimen is somewhat elongate, 31 by 36 mm. in size, andhas a single hole bored near one long side so the pendant would hangwith its long axis horizontal.Noncircular pendants include one flat ovoid specimen, 24 by about16 mm., pierced at the smaller end; a subrectangular piece withrounded corners, measuring 10 by 29 mm., and having a smallhole near each end; and a group of IT small thin variously shapedexamples that clearly were used together. In this latter group, thefew whole pieces (pi. 47, top roio) are ovoid, ovoid with straight base,trapezoidal, or subrectangular; most seem to have, or had, two per-forations near one end, always the smaller in tapered specimens ; onehas tliree perforations ; and complete pieces vary from 16 to 24 mm.long and from 10 to 15 mm. wide. The thinness of the shell suggestssome small fresh-water mussel; and judging from their clusteredoccurrence when found they must have been strung together when lost.Finally, there are two long, narrow, thick, tapered objects (pi. 46, 2)whose manner of use is not clear. Both are straight, with one end cutsquare and the other pointed. Maximum thickness is near, but not at,the blunt end. One is well finished ; the other still has the cut edgesand has not been smoothed down. They are 40 to 43 mm. long, withmaximum diameter of 7 to 8 mm. Neither is perforated, grooved, orotherwise obviously prepared for attachment to a cord. It seems im-probable that they could have been cut from the shell of any speciesidentified from the site, unless much larger examples than those in ourseries were available ; they, or the shell used, may have been imported.Spoons (?)?Ovoid or elliptical scooplike objects were made bygrinding down the entire perimeter of a suitable shell, and also remov-ing the hinge and any other surface irregularities. The only completeexample we have, which is symmetrical and well made (pi. 46, ^),measures 108 by 60 mm. ; it came from mound 17, basin 2. From pit 1came the rounded end fragment of another specimen, with groundedges, probably representing a similar, but larger, implement. Thatthey were actually spoons is, of course, problematical, but they couldhave been so used. I doubt that they were ornaments since neither ofour specimens is pierced or otherwise provided with means for attach-ment. Both are made of the shell of Proptera dlata megaptera. 294 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174PEBISHABLE MATERIALSCoiled hashetry.?Of exceptional interest are the charred remnantsof a coiled basket found inverted on the floor of basin 2, mound 17, atthe Tobias site (pi. 25, c). The upturned bottom had largely dis-integrated and at no point was it possible to trace any of the coilscompletely around the basket. The remains were repeatedly treatedin situ with a fixative and were removed en bloc ; but unfortunatelythey survived the trip to Washington only as fragments. Analysisof construction methods is handicapped by the presence of a heavycoating of scorched organic matter on the inner surface.As measured before their removal, the remains were those of ashallow circular bowl or tray approximately 22 cm. (8i/^ in.) in diam-eter and perhaps 6 or 7 cm. deep. The basket had been constructedon a foundation of single rods, 3 to 4 mm. in diameter and of uniformsize throughout their traceable length. The sewing element con-sisted of splints averaging 1.5 mm. in width. On the outside surface,these splints run diagonally to the foundation at an angle of 40 to50 degrees ; and each stitch is caught under one of the preceding coil.I suspect the coiling proceeded in clockwise fashion. There are 5to 6 coils, and about the same number of stitches, per inch. Theplant materials used are unidentified.In its original size and shape, this basket probably approachedrather closely those used by the Plains Indians in their dice games(Culin, 1907, figs. 22, 24, 32, 34, 108). According to Weltfish (1930,p. 470), there were two types of these coiled gambling baskets;and one of these types made by the Pawnee, Arikara, and perhapsthe Mandan, had "a single willow-rod foundation and [was] sewnwith heavy willow sewing thread." Moreover, the stitches wereusually interlocking. This characterization fits the present archeo-logical specimen quite well; and it may be suggested that we havehere a mid-16th century example of a basketry type widely dis-tributed among the later historic Plains tribes.Textiles.?A short section of twisted grass or shredded corn huskswas found in basin 2, mound 17. It consisted of two elements which,viewed end on, show a counterclockwise twist; each of these ele-ments in turn appears to be made up of smaller fibers with a clock-wise twist. The fragment is 55 mm. long ; it may have been cordageor coarse mat-weaving material. One or two burned lumps of roof-ing clay from this basin bore twisted cord imprints that were aboutthe size of this specimen.Wood.?On the floor of basin 2 in mound 17 was found a smallbundle of straight sticks, preserved by partial charring. They aver-aged 6.5 to 7.5 mm. in diameter, and 35 to 40 mm. long, but theiroriginal length may have been appreciably greater. One end of each Wbdhl] an introduction TO KANSAS ARCHEOLOGY 295was tapered, exposing the pith cavity, and all seem to have beenstripped of their bark. While none shows indisputable proof of cut-ting or other modification, their grouped occurrence near other car-bonized materials was probably not fortuitous. They were of suit-able diameter to have served as arrowshafts, or as rods for a screenor curtain.From basin 1, mound 17, came 4 fiat rectangular slips of wood, eachbroken at one end and apparently cleanly cut at the other. Theedges were straight and the surfaces were scraped smooth. Theyaverage 3 to 4 mm. thick, 7 to 13 mm. wide, and 21 to 69 mm. long.OBJECTS OF EUROPEAN MANtJFACTUBEThat the inhabitants of the Tobias site were in contact with Europe-ans is strongly suggested by the discovery of iron, copper, and glass ob-jects, all under conditions that rule out any possibility of a postoccupa-tional intrusion. Interestingly enough, these objects were found onlyin basins 1 and 2 of the mound 17 complex ; none are recorded from thecache pits or from refuse mounds 4 and 6.Of primary interest in connection with this material is, of course,the question of its source, its period, and the nationality of its pur-veyors. Unfortunately, none of these points can be conclusively estab-lished from our evidence, for there are no visible trader's or maker'snames or marks and I have been only partially successful in locatingcomparable forms from other datable and assumedly more or less con-temporaneously occupied sites. Moreover, so far as I have been ableto determine, none of the trade items is of a type that would not havebeen available at the time that the Tobias site, for reasons to be pre-sented later, is believed to have been inhabited.Iron objects.?These were in every instance heavily oxidized. Frombasin 1 were taken an awllike object and several bits of thin sheet orstrap iron suggestive of knife blade fragments. They occurred nearthe north end of the basin, and 18 to 24 inches underground, i. e., 8 to 14inches above the floor. The awl appears to have been 4-sided, eithersquare or diamond-shaped in cross section, at the middle, whence ittapered to a point at each end ; it was 143 mm. long (pi. 36, 1) . Basin 2 yielded two specimens. One was an iron ax blade, brokenvertically through the eye, with the poll missing. The under side ofthe blade is slightly constricted next to the eye, where it has a width of36 mm. ; thence both upper and lower margins flare to a curved cuttingedge 64 mm. long. From the eye to the cutting edge the specimenmeasures 84 mm., and thickness approximates 8 mm. The size andshape of the eye, and the form of poll, are unknown ; and on the pittedsurface of the blade no markings or stampings are visible. The flaringmargins and small size of the blade are markedly unlike the larger and 296 BUREAU OF AMERICAN ETEGSTOLOGY [Bull. 174more nearly straight-backed axlieads of English or American makefomid on early 19 '''-century Pawnee sites in Nebraska, and theearlier English or Dutch axes supplied to the Iroquois (Beauchamp,1902, p. 59 and figs. 99, 167) resemble the present specimen even less.In the Rio Grande valley, IQ^^- and iT'^'-century Spanish axestend to have longer narrower blades, but I am informed that ablade "almost identical" to our specimen was found at Quarai ^^ in whatis now eastern Valencia County, N. Mex. There is other evidence thatthe inhabitants of the Tobias site were in contact with peoples of theKio Grande, and it seems more likely therefore that our specimen isof Spanish origin than that it reached central Kansas from the east(pi. 36,^).The second specimen from basin 2 is an irregular flat kidney-shaped mass with adhesions of charcoal, twig moulds, and other ex-traneous matter. The mass is slightly magnetic, and tends to split;it measures about 55 by 40 mm. Its original form and the nature ofthe object cannot be ascertained.Copper objects.?Copper occurs only as small strips rolled, or withthe ends folded over, to form tubular beads or beadlike pieces. Thereare three examples, all from the floor of basin 2. The largest iselliptical in cross section, and measures 12 by 9 by 5 mm. It wasmade by folding one end of a 20 by 12 mm. strip over the other, andpressing the overlap tightly together. It may have been fashionedaround a thong or cord as a "dangler," but of such a cord there is notrace. The other two specimens are 10 by 5 mm. and 5 by 3 mm., andshow the same method of construction.Glass objects.?Scattered through the fill of basin 1, at deptlis of 18to 35 inches underground, were five glass beads. All are slightly asym-metrical, but basically are subspheroidal in form with flattened sur-faces at right angles to the perforation. They are of pale-blue glass,slightly iridescent, with fiinely pitted surface, and are unquestionablyof white man's manufacture. Diameter varies from 5 to 7 mm.,thickness from 4.5 to 6 mm., and the perforation is 1 to 1.5 mm. indiameter. They conform in all particulars to those in the necklacedescribed immediatel}^ below, and may once have formed a part of it.Glass, turquoise, and bone necklace.?Among the items of particularinterest from the Tobias site was this assemblage of some 250 beadsand pendants. The objects lay in a cluster (pi. 25, b) on the floor ofbasin 1 in mound 17, beside a large postmold and about a foot fromthe southwest wall of the basin. As found, there was no trace of astring, but painstaking excavation showed that many of the beads stilllay in juxtaposition so that they could be restrung, in part, in their ?* Information by Miss Marjorie F. Tichj-, curator of archeology, Museum of New Mexico,Santa Fe. Weuel] an introduction TO KANSAS ARCHEOLOGY 297 original order. That the objects once formed a necklace there canbe no reasonable doubt, and even the general arrangement of thevarious types of beads as restrung is believed to be substantially cor-rect (pi. 46,i,Ze/i^).The necklace at present includes 144 glass beads; additional frag-ments suggest there were originally between 150 and 160. They aredull whitish, irregular in size, and have pitted surfaces ; but when wet,and while inmaersed in a preservative, they have an attractive light-blue color. Their present appearance may be a result of weathering.In general, they are about the size of a pea and globular in form withflattened surfaces at the ends of the perforations; diameter rangesfrom 4.5 to 7 mm.Bone beads number 38, with pieces of perhaps half a dozen more.They are 7 to 18 mm. long, by 1.5 to 3 mm. in diameter, and bear noincising or other hint of decoration. They served as spacers betweengroups of glass beads.At the lower end of the original necklace both strands came to-gether and passed through a shell cylinder 4.5 nmi. long by 5.5 mm.in diameter. Immediately below this was a polished oblong turquoisependant measuring 15 by 13 by 3 mm., with a perforation at thesmaller end. Below this, in turn, was a short loop with 32 flat circularturquoise beads, 1-2.5 mm. thick by 4 mm. in diameter, and averaging14 to 16 per linear inch. A thin circular centrally pierced shell disk,15 mm. in diameter, divided the turquoise beads into two groups of15 and 17 each. The turquoise is of good quality and beautiful color,and has resisted decay far better than the glass beads.As to the accuracy of the reconstruction, the turquoise beads andthe shell disk divider are unquestionably in correct order. The place-ment of the turquoise pendant and shell tube is less certain, sincethey had been slightly displaced apparently while undergromid.Above this point, to a distance of 27 cm. from the turquoise beads, theglass beads and bone spacers lay in order in the following sequence,wherein the number indicates glass beads and B denotes single bonetubes : 2-B-5-B-5-E-4^B-2-B-4-B-5-B-l-B-5-BThe other side of the string was reconstructed to match this sequence,and the remaining specimens were then arbitrarily strung with groupsof two to five glass beads separated by bone tubes.From the location of the find it may be surmised that the necklacewas perhaps cached away in the ceiling interstices of the structurenear the top of the post, falling to the floor when the roof collapsed.There is no charring or scorching on any of the component elements,and there seems to have been little disturbance of the trinket once itreached the floor. 298 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174It would be interesting to know whether the necklace was broughtin as a finished piece or was made on the site. The turquoise beadsand pendant were very likely imported from the upper Rio Grandearea, and the glass may have come from the same direction, thoughultimately certainly of white man's manufacture. The bone tubes andshell objects could have been made locally.THE THOMPSON SITE (14RC9)In course of our investigations at the Tobias site and elsewhere inRice County, local collectors from time to time brought to our campspecimens dug or plowed out of various sites in the county. Mostlythese objects included raw materials available to, or products madeby, the local Indians and could be readily duplicated by the specimensrecovered in our own researches. There were also occasional exoticitems and materials, including obsidian, bits of turquoise, and, mostintriguing of all, rare potsherds bearing glaze-paint decoration. Theseglaze-paint sherds were for the most part surface finds, but throughthe enthusiastic cooperation of Horace Jones, of the Lyons DailyNews, we were able to learn at last that several had been dug out ofa mound on the C. F. Thompson farm just across the river from theTobias site. In hopes of finding more such fragments, in situ, andthereby getting a cross-date with the Rio Grande pottery sequence,we accordingly undertook limited explorations on the Thompsonproperty.The Thompson site (fig. 59) lies about 450 yards northeast of theTobias site, across the valley of Little Arkansas River. Here therolling uplands from the north end in a series of narrow south-pointing tongues separated by short dry ravines. Atop one of thesesloping tongues, in unbroken sod overlooking the wooded valleybottoms, is the site. It consists of 10 or more artificial mounds, withat least 4 others faintly discernible in cultivated rising ground im-mediately to the north. The site covers considerably less groundthan does Tobias, but, like it, is strung out along a ridge which did notpermit much lateral expansion. Curiously enough, no mounds orother remains were noted on the nearby ridges to east and west,though they would seem equally well suited to occupancy (pi. 48, a).At present the nearest source of surface water is the Little Ar-kansas, nowhere less than 250 or 300 yards distant. The descent to thebottoms is more gentle than that at Tobias, and the sandstone ledgesat the tip of the site ridge are relatively inconspicuous. That springsever issued from their base appears unlikely, though the deeply cutdraw southwest of the site had a small wet spot suggesting a seepat the time of our first visit in mid-June. This was dry a monthlater. Wkdel] AN INTRODUCTION TO KANSAS ARCHEOLOGY 299 2 f3, # , -a . . X " '^..... i / 300 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174hundred yards east of the Tobias site, lay only a little farther fromthe Thompson site. Almost due south across the creek, this circle isscreened from the Thompson site only by the fringe of trees alongthe stream. It may or may not have served the occupants at thissite. The Tobias site, too, is hidden by woods ; if the creek bottomscontained less, smaller, or no timber in Indian days, the two sites andthe isolated "council-circle" would have been easily visible from, andwithin hailing distance of, one another.The mounds on this site showed no obvious superficial differencefrom those on the Tobias site, except that one or two appeared to besomewhat larger. Of the 10 still in sod, 5 were strung along the westrim of the ridge, 4 were on the east slope, and 1 lay on the ridge top.The area covered by this cluster did not exceed 100 by 175 yards,but it is probable that a considerably larger area of occupation to thenorth has been obscured by modern farming operations. For ourpurposes, we assigned an index number to each mound (see fig. 59),and then selected the two largest for trenching. Mound 1, firstpointed out to us as that from which the glaze-paint sherds weretaken, was bisected by an east-west trench 5 feet wide, with a shortertrench at right angles from the center to the south edge. Operationswere shortly transferred to mound 2, upon the strong recommenda-tion of County Commissioner Phil Hohl who is said to have dug outthe puebloan sherds. These two mound tests, together with theopening of five cache pits, constituted the extent of our excavations. Mound 1, on the east slope of the ridge, was low and spreading, witha diameter of about 45 feet, and an apparent height of nearly 3 feet(pi. 48, h). Small sodded depressions near the highest point on thesouth half testified to sporadic test-pitting in the past. To our dis-appointment, the T-shaped trench revealed that the mound had beenbuilt upon a slight rise and that the undisturbed subsoil was nowheremore than 24 inches below the mound surface. Relatively little ad-mixture could be detected at depths greater than 18 or 20 inches, savein the ever-present rodent burrows. The fill generally consisted ofa dark grayish-brown soil containing some ash, bits of charcoal, ani-mal bones, flint chips, sherds, and miscellaneous broken and unbrokenartifacts. Beyond a gradual thinning out numerically as subsoilwas approached, there appeared to be no significant difference in dis-tribution or type of cultural materials. There was no trace of post-molds, fire basins, or other features in our cuts, and nothing to indicatethat the mound was not an accumulation of refuse from nearby habita-tions. This identification is substantiated by the location of the moundon a slope which, though gentle, would almost certainly not have been Wedbl] an INTRODUCl'ION TO KANSAS ARCHEOLOGY 301picked for a lodge site so long as the much more suitable ridge topwas available.At the west edge of the mound, and visible as a shallow surfacedepression, was a cache pit. This, No. 4 in our series, contained in itsupper portion about 30 inches of ashy soil strewn with discarded arti-facts and trash, which blended into the thin midden deposit of thepartly overlying mound. Near the south side of the pit in the un-disturbed lower ash lenses, was a rusted mass of iron rings, sinceidentified as chain mail. Below this was 36 inches of hard almoststerile soil, underlain by nearly 12 inches of finely laminated silt ap-parently laid down while water stood in the open pit. Pottery frag-ments, worked scapulae, scrapers, drills, arrowpoints, and other arti-facts from the upper part of the cache were the same as those foundin the overlying mound.Of particular interest in the mound material was a small cachelikedeposit of household implements on the east side of square 5, at 15inches depth. Included were a crude grooved maul, a large muller,a bone hoe, an awl, and a small mortar. In square 15, at 7 inches depth,was half of a badly rusted iron ring slightly less than a half inch indiameter. This undoubtedly is chain mail. Charred corncob frag-ments and kernels were scattered through the mound fill and in thecache. There were no puebloan sherds.MOUND 2Mound 2 lay approximately 170 yards northwest of the preceding,on the west rim of the ridge. Its diameter was about the same, andthe height was under 2 feet. Most of the central area had been com-pletely dug over in the fairly recent past, so that only the east andwest ends of our 3-foot cross trench yielded any reliable data. Inthese undisturbed portions the fill was very hard, compact, and darkin color, with the expectable small quantities of cast-off sherds, animalbones, chips and stones, and an occasional artifact. This material didnot exceed 16 or 18 inches in depth, except where cache pits directlyunderlay the mound.A scant foot east of the central disturbed area but clearly in un-molested mound fill, at a depth of 15 inches, were found 3 red-slippedsherds with glaze-paint decoration. Under these, and partially under-lying the old diggings, was a cache pit. There was no evidence in ourtrench wall that this pit had ever extended up into or through themound fill, and it seems evident that the cache had been filled in beforemound deposition began. No great time lapse is implied, and it ispossible that the mound accumulation was localized here because theIndians simply continued to dump trash on and over abandoned cachesafter these had been filled up. The puebloan sherds are believed to 302 BUREAU OF AMERICAN ETHNOLOGY [BULL. 174have lain at the bottom of the mound and on top of the older cachepit. Very probably, too, they are a part of the same imported vesselor vessels represented by the similar sherds previously dug from thecentral part of this mound. But in any case, whether in mound fillor cache pit, they were certainly inclusive, and their association withpottery of strictly local manufacture establishes a contact point betweenSouthwest and Central Plains sequences.CACHE PITSThe five cache pits excavated did not differ significantly from thoseopened at the Tobias site. All were large, with capacity of 50 or morebushels each, and at least two showed evidence of a grass-lined floor.All contained strata of hard soil with few ashes or artifacts, suggestingdeposits of earth perhaps dug from other pits and dumped into aban-doned structures. It may be noted that in two or three instances, andparticularly in the lower 12 inches of pit 4, there were flat-lyinglaminae of fine silty material, suggesting the deposition of wind-blownmaterial in standing water. It appears unlikely that the Indians wouldleave 6- to 7-foot holes open for any length of time ; at night, or withchildren about, they would endanger life and limb. Yet it is difficultto see how the laminated silts would be deposited in any other way.The measurements and other data on pits are summarized below(table?). Table 7. ? Data on cache pita at Thompson site (l^RCQ) Pit Wbdbl] an introduction TO KANSAS ARCHEOLOGY 303 occurring exclusively, or nearly so, in the basins of the mound 17 com-plex at Tobias were not found at Thompson. With due allowancefor these factors, and particularly for the last, it is, nevertheless,evident that the material culture inventory at Thompson closelyparallels that found in the pits and refuse mounds at Tobias. Certaindifferences in detail are apparent, but these may well be due to indi-vidual or familial variations in taste and habit or to a probably slightdisparity in time. At any rate, the ensuing descriptions are brieferthan those for the Tobias artifacts, and for additional details onparticular forms or artifact types the interested reader is referredback to the relevant section on similar items from Tobias. Exceptionalor deviant specimens, of course, will be treated in appropriate detail.Food RemainsMaize.?This undoubtedly important item in the native subsistenceeconomy was disappointingly scarce in our workings. Charred ker-nels were noted here and there throughout the mounds and pits, butnowhere in any quantity ; and there are but two small cobs, both in-complete. One, from pit lA, measures 27 mm. long by 12 mm. indiameter, and was apparently 10 rowed. The other, from mound 1,is 22 by 12 mm., and was 8 rowed.No trace of beans, cucurbits, or other vegetal food materials, do-mestic or wild, were noted.Molluscan remains.?These are included here with some reluctance,since there is no evidence whether the natives did or did not eat shell-fish. There were, however, no worked shells in that portion of thedeposits we examined, and it is not certain that the native pottersincorporated crushed or burned shell into their manufactures. Onthe assumption, then, that shellfish may have been consumed occasion-ally or by certain individuals, we may note that five varieties of fresh-water mollusks were represented by 10 specimens. Whether thenatives went beyond the Little Arkansas River at their doorstep togather this material, I do not know. The first three species listedwere also present at the Tobias site. Included in our series are thefollowing specimens : Quadrula pustulosa prasina (Conrad) 1Uniomerus tetralasmus (Say) 3Strophitus rugosus (Swainson) 3Lampsilis anodontoides (Lea) 2BuUmulua dealbatus (Say) 1Turtle remains.?From pit 1 were taken a few skeletal parts of thebox turtle (I'errapene) ; there is no evidence that the carapace hadbeen worked.4&4172?C9 21 304 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174Mammalian reTnains.?Bones of the following mammals were iden-tified from the Thompson site in the numbers indicated : Bison (Bison Mson) 26Dog (Canis familiaris) 9White-tailed deer (Odocoileus virginianus) 6Antelope (Antilocapra americana) 2Dog or wolf 2Elk (Cervus canadensis) 1Jackrabbit (Lepus californicus) 1Wolf iCanis Inpus) 1POTTBEYThere are from the Thompson site no whole or restored vessels;our sample includes 2,387 potsherds, of which 2,118 are body frag-ments and 269 are rim pieces. The number of vessels represented bythis sample, judging from the rim variations, probably exceeds 150.Of the total number of sherds 88 percent, including all rims, aresand-tempered gray wares. The gray ware body sherds can be fur-ther subdivided, on the basis of surface treatment, into the same twovarieties found at Tobias?Geneseo Plain and Geneseo SimpleStamped (p. 233) ; the runs camiot with certainty be allocated to oneor the other of these closely related wares, and so are separatelydiscussed below as a gi'oup. Of the body sherds, 76.3 percent areclassed as Geneseo Plain and 10.4 percent as Geneseo Simple Stamped.These proportions and ratio are not significantly different from thoseat Tobias; and, as shown in table 8, the percentages for the site asa whole are approximated in practically every excavation unit exceptpit lA. It can be safely concluded, I think, that these two graywares were the standard domestic pottery of the site.Absence of whole or restorable vessels and the generally small sizeof the rimsherds make a determination of rim form and positiondifficult. There is evidence, however, of straight vertical to slightlyflaring examples and of others that were gently outward curved;heights of 3 to 6 cm. seem to have prevailed, and decoration is eitheiabsent or occurs only on the lip, which is almost invariably rounded.About one-third of the lips are undecorated, and an approximatelyequal number bear short diagonal cross incisions; "pie-crust" scal-loping, punctating, and herringbone incising occur in diminishingfrequency (fig. 60, 6, c, e) . Seventy-five rimsherds have a single plain(27) or diagonally incised (27) fillet, or two parallel plain (2) fillets,or alternatively a single row of pinched-up nodes or fingernailgougings (19) on the exterior about 10 to 25 mm. below the lip (fig.60, f-4). Fifty-six of these, apparently representing at least 43 ves-sels, came from mound 1, and were further characterized by having,in 36 instances, pie-crust scalloping on the inner upper lip. Fromthe sameness of style and technique I would suspect that most of Wbdbl] A^ INTRODUCTION TO KANSAS ARCHEOLOGY 305 5p o o> 05 I^ t-- 00 ^5* f-H iO t^ >2 2S J2 ?-' Figure 73.?a, b, Beveled knives; c, d, unbeveled knives, from Elliott site. Wedbl] an introduction TO KANSAS ARCHEOLOGY 371Elliott site ; it is of banded fusiilinid chert, 12 mm. thick, 65 mm. wide,and, if symmetrical, had an original length of 135 to 140 mm. A num-ber of other sizable fragments, similarly well worked, closely approxi-mate the contours of corresponding parts of this specimen and suggestthat the type may have been a common one.Knives with heveled edges.?These were about as numerous as thepreceding group, to judge by the number of fragments, and breakageseems to have been rather less. That some, at least, were intended forhafting is indicated by the presence of notched and/or stemmed pieces.In these, maximum width was usually just above the notches or shoul-ders, whence the oppositely beveled convex, straight, or concave bladeedges taper to a point. The base is either rounded or straight, and theblade always has a rhomboidal cross section. Our sample from theCowley County sites is much smaller than that from Rice County, butit indicates that substantially the same forms or varieties were in usein both localities. All of the Cowley County pieces which still showthe base can be duplicated from the Tobias site and neighboring sta-tions (fig. 73, a, h).Among the numerous and somewhat varied beveled blade fragmentsfrom Cowley County it should be noted that all, when held with thesmaller (tip) end up, show the beveling on the left edge. This like-wise tallies with the Rice County findings.Some of the broken pieces may have been originally 4 edged, sincethey suggest a change in the direction of bevel at the broadest part ofthe blade. There are, however, no complete 4-edged or diamond-shaped knives in our series, and I cannot say whether this, the so-called "Harahey" form, was known to, or in use among, the local Indianpopulation. If it was, it must have been relatively uncommon bycomparison with the notched and stemmed 2-edged forms.End scrapers.?Planoconvex end scrapers, usually ovate in outline,were found at all three sites. In most the long edges were rubbeddown, but the broad rounded end was sharp and unworn. Theyvaried in length from 21 to 70 mm., but at every site approximatelytwo-thirds were under 38 mm. long. They conform in all particularsto those from Rice County.Side scrapers.?Irregular spalls of various sizes and shapes withone, or rarely more, edge retouched from one side only, were also pres-ent everywhere. Except in the matter of retouched edges, they dif-fered in no respect from hundreds of other spalls and "rejects." Theywere found rather less commonly at the Elliott site, where fragmentsof bifacially retouched and shaped knives were more plentiful (fig. 74,a, 6,).Chopping tools.?Thick, rudely shaped and coarsely chipped subel-liptic to subovate objects, occasionally notched and with batterededges, were presumably hafted and used as axes. Similar unnotched 372 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174 ^ -Q Wedbl] an introduction TO KANSAS ARCHEOLOGY 373pieces also occur, but whether these represent unfinished examples orwere used in different fashion I do not know. Little care was devotedto the shaping and finishing of these tools, but their size and the bat-tered edges indicate that they saw heavy service. They range fromsmall specimens 65 mm. in greatest dimension to a large piece 225 by105 by 65 mm., and were found, though in small numbers, at all threesites (fig. 74, c; pi. 56, ?, m) . OBJECTS OF GEOUND AND PECKED STONEMealing slab.?A single small fragment of worked limestone fromthe Country Club site is tentatively identified as from the slightlyraised edge of a troughed metate. The slab appears to have been 6 to 8cm. thick, and the presumed upper surface is smoothed. The originalsize and shape of the implement is unknown.Manos.?Manos or handstones occurred at all three sites, suggestingthat the metate-mano complex was much more common than the singlemealing slab (?) fragment noted above would indicate. From ourcomparatively small series of manos it would appear that the preva-lent form was elongate elliptical with rounded ends and planoconvexcross section ; the thick subrectangular type with rounded corners wasmuch scarcer. None of the specimens has more than one grinding sur-face. With one or two exceptions of sandstone, all are of gritty, oftenfossiliferous, limestone of local origin. Complete specimens are 105 to292 mm. long, 84 to 95 mm. wide, and 32 to 65 mm. thick.Grooved mauls.?Three more or less complete mauls, one from eachsite, were found, as well as a spall possibly representing a fourth. Thatfrom Larcom-Haggard (pit 4) has convex sides, moderately convexstriking surfaces, and a well-defined median encircling groove ; it is offossiliferous limestone, measures 129 by 106 mm., and weighs 74 ounces.From the Elliott site came another of sandy limestone, with subrec-tangular convex striking surfaces, straight to slightly convex sides, anda deep groove ; it measures 128 by 105 mm. The one from the CountryClub site is badly battered, with one poll almost entirely spalled away ; it apparently once had straight sides, an oval cross section, and a low-flanged groove, with measurements of 105 by 90 mm., and a weight of44 ounces. The material is quartzite or very hard sandstone. None ofthese three was as well made or as asymmetrical as the finest quartzitespecimens from Rice County. A spall, found with the Country Clubmaul in pit B, mound 1, is wedge shaped, and shows part of a flatsurface separated by a curving edge from a carefully dressed convexand curving surface. Of red quartzite it may possibly represent oneof the better-made forms like those at the Tobias site.SJmft smoothers.?There is but a single complete example of theshaped rectangular longitudinally grooved shaft smoother, but frag-ments evidently of the same type were found at all three sites. The 374 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174complete specimen, from pit A, mound 1, Country Club site, is 102 by 22by 24 mm. The material, in all cases, is a dark-brown Dakota sand-stone. In all respects, these fragments and the whole piece, are likethose from Rice County.Sharpening blocks.?These are small irregular pieces of sandstone,and in some instances re-used fragments of shaft smoothers, whosesurface has short narrow grooves unsuited to shaft smoothing andprobably resulting from the shaping of awls, needles, and similar smallpointed objects. They occurred at all sites.Shaft polisher (?) .?This term is used to designate provisionallyan irregular worked limestone block from pit 8, Elliott site. One sur-face has been flattened by scraping, and is deeply striated from thisprocess ; across it run two grooves, each about 10 mm. wide by 4 mm.deep and deeply striated, and separated from each other by a narrowresidual ridge. The grooves are somewhat angular in cross section,and do not show the polish or smoothing that would result from draw-ing a shaft or rod through them. They are, however, of approximatelythe right size to accommodate such an object, and I suspect represent anunfinished buffing tool. I can suggest no other likely use for it.Rectanguloid block.?This is a carefully shaped, subrectangular ob-ject, with dressed convex edges and smoothed convex surfaces, closelyapproximating in form a small pillow or cushion. One surface hasmarks somewhat suggesting wear from a thong, and on one edge is apecked fingerhole. It is of dark-gray sandstone, and measures 178by 116 by 39 mm. There is nothing to suggest its manner of use; it ismuch more carefully shaped than a mano and cannot be so considered.It was found in a test east of the old Walnut River channel at theLarcom-Haggard site.Pipes.?There are, unfortunately, no whole pipes in our CowleyCounty collections, and the fragments found, though suggestive, aremostly inconclusive regarding the original form and size of the objects.From mound 1, Country Club site, came part of a bowl and short right-angled stem from a small split pipe of soft crinoidal limestone. Thebowl seems to have been rather low and bulbous, but precise recon-struction is impossible.Fragments from the upper rim portion of two pipes were found atthe Elliott site. One, from pit 5, is of catlinite, and seems to havebeen part of a tall somewhat bulbous bowl ; the interior is smoke black-ened. About 2.5 mm. below the lip is an encircling groove, betweenwhich and the lip are faintly incised shigle cheA^rons, pointing alter-nately up and down. The piece conforms closely in size and contoursto the upper bowl of the L-shaped catlinite pipes from Rice County,and possibly is from one of that type, but of this I cannot be certain.The second piece, from pit 7, is of soft gray limestone, and is much Wbdel] an introduction TO KANSAS ARCHEOLOGY 375heavier than the preceding. Below the lip, two deep close-set groovesapparently encircled the bowl. The cavity is about 15 mm. across,and the bowl was at least 25 to 28 mm. in diameter. There are manytiny fragments belonging to this piece, all indicating a thick-walledpipe; but, as with the catlinite specimen above, proof of an elbowshape is wanting.A fragment of thick-walled stone tube was also taken from pit 7,Elliott site. There are longitudinal working striae on the externalsurface, as also on the bored cavity. Thickness is about 11 mm. Thespecimen suggests the thick-walled tubular pipes from Rice County.There is no smoke blackening, but the piece may never have been com-pleted or used.Celts.?Two basalt celts were found in pit 7, Elliott site (pi. 56,A, ^) . The smaller has a nearly circular cross section, a rounded butt,and a thick blade with slightly curved working edge. Most of the sur-face shows the "dimpled" texturing of the hammer-pecking, exceptwhere grinding of the edge has nearly obliterated this. The largerspecimen is less symmetrical, slightly thicker than wide, with a nar-rower blade. Both are slightly chipped at the ends. They measure140 by 48 by 45 mm. and 160 by 45 by 46 mm., respectively.To the best of my knowledge these are the only celts so far knownfrom the three sites we worked in Cowley County, and I recall no in-stance of their occurrence in the Rice County sites. This extremescarcity, or virtual absence, of the type suggests that the present pieceswere not typical and may have been acquired, by trade or otherwise,from another group, perhaps to the southeast.Cwpstone.?This is an irregular chunk of fossiliferous limestonebearing on its upper surface 3 or 4 small hemispheric pits or depres-sions. Best marked of the pits is a symmetrical one 25 to 28 mm. indiameter by 7 to 8 mm. deep ; the others are smaller, and are scatteredat random over the surface. The stone measures about 160 by 90 by75 mm. ; it is from the excavations at the Country Club site.XJNWOBKED STONE ; PIGMENTSIn addition to the stone artifacts, there were many fire-crackedstones and stream-worn boulders of varied sizes and shapes in all ofthe sites. The former may have been used in cooking, as pot rests, orfor other purposes connected with culinary pursuits. Some of thestream-worn boulders could have been used, with little or no modifica-tion, as mullers, rubbing stones, hammerstones, and for other similarpurposes. No attempt was made to collect any of these specimens, orto note specifically their size and other characteristics.Pigments include four lumps of hematite from the Larcom-Haggardsite. All are small and show flat striated facets where red powder wasground off. Two are soft and friable ; the others are hard and compact. 376 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174OBJECTS OF SHELLAbout 100 or more shells of fresh-water mussels, most of them un-worked, were collected from the three Cowley County sites. All are ofspecies presumed to be, or formerly to have been, obtainable in thenearby Walnut and Arkansas Rivers. Of the species identified fromthe sites, only three are represented by artifacts. Wliether the restconstitute refuse from the gathering of shellfish for food is conjectural,but it is probable that the burned shell used in pottery tempering camein large measure from these local species. Table 14 lists the speciesidentified, with the number of shells from each site, and indicates byan asterisk the ones used for artifact manufacture.Table 14 Species Larcom-Haggard EUIott CountryClubAmhlema costaia (Raf)Ouadrula quadrula (Raf)Ouadrula pusiulosa (Lea)*Quadrula pustulosa prasina (Conrad).* Tritigonia verrucosa (Raf)Uniomerus tetralasmus (Say)Pleurobema coccineum (Conrad)*Lampsilis anodontoides (Lea)Lampsilis fallaciosa (Smith)Lampsilis ventricosa occidens (Lea)Lasmigona complanata (Barnes)Proptera data megaptera (Raf)Proptera capax (Green)Ohliquaria reflexa (Raf)Leptodea fragilis (Raf)Truncilla truncata (Raf) *41 16 Wbdbl] an introduction to KANSAS ARCHEOLOGY 377margins of tlie shells. Otherwise unmodified, they may have beenused as scrapers, e. g., in smoothing pots or other surfaces, or as spoons.One specimen is made from the left valve of a Tritigonia verrucosa(Raf ) ; the others are Lampsilis anodontoides (Lea) . B^SUMfi; SUMMARY OF ARTIFACTS FOUNDThat the Indians who once dwelt on the three Cowley County sitesdescribed above shared the same material culture inventory is obvious,assuming that our comparatively limited artifact sample is representa-tive. An item-by-item comparison, as summarized in table 15, merelyemphasizes what was already apparent in the field, namely, that almostwithout exception the specimens from any one site could not be dis-tinguished from those found at either of the others. Items which, asshown in the tabular summary, were found at two sites could withreasonable confidence be expected to occur at the third also, if moreextended work were undertaken ; at least, I doubt strongly that bowls,strap handles on pots, incised sherds, cancellous bone artifacts, pipes,etc., were unknown and unused by the occupants of the site for whichwe have respectively no entry. And, with exception of the mealingslab and hematite, those items which are indicated but once seem gen-erally to be so rare or unusual in the region that they may be regardedeither as products of individual whim or fancy, as in the case of ringbeads, tubes, and tibia shaft straighteners, or else as trade pieces, asfor example, the punctate and engraved sherds, the celts and thecupstone. Pending new evidence to the contrary from future excava-tions, then, we may tentatively conclude that these three sites constitutea cultural unit. For this unit of three sites, to which others may beadded as relationship is shown, I suggest the name Lower WalnutFocus.Among the native culture complexes now known in Kansas, thethree sites here designated the Lower Walnut Focus appear to havetheir closest relationships with the sites in Rice and McPhersonCounties which I have termed the Little River Focus. There arenotable differences, however, and at the moment these seem sufficientto warrant a distinction between the two site groups, which are situ-ated approximately 100 miles apart.We have already noted that the Lower Walnut pottery is almostexclusively shell tempered, whereas the Little River wares are pre-ponderantly sand tempered; that flat vessel bases and handles areproportionately much less plentiful in the Little River sites than inLower Walnut River; that bowl forms are indicated for LowerWalnut but not for Little River; and that simple stamping, repre-sented on 12 to 35 percent of Little River sherds, is wholly absentfrom our Walnut River collections. With respect to bonework, the 378 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174Table 15. ? Summary of artifacts from three sites in Cowley County, Kans. Artifacts Larcom-Haggard,14C01 EUiott,14C02 Pottery:Type Cowley Plain sherds:Jars, restoredBowls, restoredBase flat, circularLip rounded, plain..Lip rounded, diagonal incisedRim incised inner upperHandles, loop (including fragments) .. . Handles, loop, rivetedHandles, loop, decorHandles, loop, angle nodesHandles, strap (iucluding fragments) -Handles, strap, riveted -..Handles, strap, angle nodesIncised sherds, bodySand-tempered sherdsBedware sherdsPunctate sherdsEngraved sherd (see text)Rio Grande glaze sherdsBone artifacts:Scapula digging toolsAwls: split mammal leg boneAwls: neural spine or rib edgeShaft straighteners, deer tibiaScored implements, rib -..Cancellous bone artifactsTubular beadsRing beadsTubesMiscellaneous worked piecesChipped stone:Projectile points:(a) Small, triang. , unnotched -.(b) Small, triang., notched(c) Small, triang. or leaf-shaped(d) Other....Drills (including fragments) : (a) Plain-shafted, large, heavy(b) Expanded base, grad. widened(c) Expanded base, abruptly widened.Knives, with unbeveled edgesKnives, beveled; notched or stemmedKnives, beveled; not notched or stemmed.Beveled fragments, unclassified..End scrapers.Side scrapersChopping toolsGround and pecked stone:Mealing slab fragment (?)...ManosGrooved maulsShaft smoothers (including fragments)Sharpening blocks .--Shaft polisher (?)Rectanguloid blockPipes, incompleteCeltsCupstoneHematite pigmentShell objects:Whole shell pendantNotched shell ?Shell scrapers 850 Ca. 304 2001 12525341042Ca. 155 Lower Walnut Focus lias socketed scapula digging tools while thosefrom Little River sites are usually unsocketed; Little River hassplit-rib awls but Lower Walnut does not; and solid bone projectilepoints are exclusively Little River. In regard to chipped stone, ourcollections suggest that in Lower Walnut sites, expanded base drillspredominate over straight forms, whereas the reverse is true in LittleRiver; and also that, while small triangular projectile points char- Wedbl] an introduction TO IvANSAS ARCHEOLOGY 379 acterize both foci, unnotched forms are proportionately much moreabundant in Lower Walnut than in Little Eiver sites. With respectto the sites themselves, a striking difference is the occurrence of the "comicil-circle" complex at Little River sites and its apparent absencefrom all known Lower Walnut River sites.Along with the above differences, there are many close resemblancesin the inventories from the two localities, as is obvious from an in-spection of the respective sunmiary tables (cf. tables 12 and 15). Theceramic remains, despite the differences manifested in temperingmaterials and in other details, seem to me to be variants of a singletradition; and in the matter of bone and stone industries, the re-semblances impress me as perhaps more basic than are the dissimi-larities. Our Lower Walnut sites compare rather more closely withthe Thompson and Malone sites where, as in Cowley County, ourexcavations were of short duration and limited extent. The Tobiassite, where we opened two unidentified subsurface structures as wellas cache pits and refuse mounds, shows a longer inventory and greatervariety of implement types. It is entirely possible that some of thefocus differences we now see, as above enumerated, would become lessmarked, or even that they would vanish, if excavations on the scalewe carried on at Tobias were conducted at the other sites involved.At the moment, I believe the focus distinctions made here promiseto be useful and valid ; and the many close similarities, which I sus-pect extend to the basic settlement patterns and the subsistenceeconomies as well, further justify the combining of these two foci intoa larger grouping for which I have elsewhere (Wedel, 1947, p. 151)proposed the term Great Bend Aspect. To this larger grouping,either within one of the two proposed foci or in other foci yet to bedefined, can probably be assigned all archeological materials incentral Kansas which have previously been included under the termPaint Creek culture (Wedel, 1940, pp. 332 and 343, fig. 21).The wider relationships of the Great Bend Aspect, and its chrono-logical and cultural implications for Central Plains prehistory, willbe considered in a later section.ARCHEOLOGICAL INVESTIGATIONS IN WESTERNKANSASSITES ON SALT CREEK IN LANE AND SCOTT COUNTIESSalt Creek is a short inconsiderable stream that rises on the HighPlains of northeastern Scott County and flows in a northeasterlydirection across the northwest corner of Lane County to join theSmoky Hill River in south-central Gove County (fig. 75). Thevalley, perhaps 15 miles from source to mouth, is shallow and hasgently sloping sides in its upper portion, with occasional low rocky 380 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174 Wbdbl] an introduction TO KANSAS ARCHEOLOGY 381taining fish, large bullfrogs, and other aquatic forms. The bankswere generally muddy, though clean sand was easily found not farbelow the surface silt.The valley of Salt Creek, like that of most typical Plains streams,is sparsely timbered. Hackberry, cottonwood, and willow are theprincipal larger forms, occurring as clumps wherever the creekcurves against the bluffs, and providing a welcome relief from theshort grass-yucca-cactus assemblage of the surrounding uplands.Lesser growths included chokecherry, wild plums, wild grapes, fra-grant sumac and, in disquieting abundance, extensive thickets ofpoison ivy. Saltgrass, from which the creek is said to have gottenits name, is present in some of the protected draws. Watercressintroduced by the settlers now flourishes about some of the largersprings. Muskrat, raccoon, coyote, badger, skunk, and other smallfur bearers still exist; an otter is said to have been killed on thecreek in the early days, but there is no recollection of beaver. Ofthe surprisingly varied and abundant avifauna, including variousraptores and passerine species, perhaps the most interesting to ourparty was a Carolina wren who persistently endeavored to constructa nest in the turned-up flap of the cook tent door.The archeological possibilities of this little valley were firstbrought to my notice by A. T. Hill, of Lincoln, Nebr. ; and the desir-ability of systematic excavations became apparent when visitors toour camp in Scott County State Park showed us small sample collec-tions of curiously varied materials found intermingled on certainof the terraces. As elsewhere in the western Plains, so here thedevastating dust storms of the 1930's had stripped away much of theloosened topsoil wherever the valley floor and slopes were under cul-tivation. The relative abundance of artifacts on the denuded sur-faces had attracted collectors from a considerable distance in everydirection, though with rare exceptions there seems to have beencomparatively little actual digging. Most of this nonprofessionalactivity on Salt Creek had taken place within a mile or so of the Lane-Scott County line, above as well as below. Here, too, most of ourwork was done, though surface examination indicates that smallamounts of material are to be found almost anywhere in the valleythat is conveniently accessible to live water.THE POTTORFF SITE (14LA1)The principal site on Salt Creek, so far as is known, is in LaneCounty, about three-quarters of a mile east of the Scott County line.It is situated on land which, at the time of our work, was ownedby C. D. Pottorff. Here the stream makes a series of serpentinebends, within each of which is a flat well-defined flood-free terrace 382 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174of 2 to 5 acres extent backed by a low bluff or steep liill. The firstthree terraces formed by the creek after it enters the former Pottorffproperty are of particular interest to us here, though there are alsoindications of habitation on at least two others immediately to thenorth, downstream. Just above the upper edge of the middle terrace,an excellent spring issues from the foot of the east bluff, and thereis another smaller one a few rods to the south. Perhaps 300 yardsto the northwest, across the creek, a short grassy swale cuts thewest bluff line, giving rise to several small springs and seeps. Itmay be presumed that in Indian days all of these were permanentand insured at least a modest flow of water in the creek here at alltimes. At no point would persons residing on any of the upper threeterraces have been over 150 yards from a suitable spring; and ifwater from the creek were usable this distance could have beenhalved. Undoubtedly, the proximity of a dependable supply of goodwater, with its beneficial effects on tree and grass growth and theconsequent attraction of large and small game, was an importantfactor in the selection of the site by native peoples. At any rate, it isclear that the remains, though not indicative of long occupation bylarge numbers of Indians, represent repeated utilization of the spotby distinct groups over a time span of several centuries.As might be expected from the situation of the springs, the heaviestconcentration of Indian remains occurs on the middle terrace andthe nearby north part of the first (south) terrace. Wlien we firstsaw the site, surface traces were not abundant, but the handful ofsherds, projectile points and fragments, and other scraps pickedup bore out our earlier information as to the diversity of typespresent. More intriguing still was the discovery on the gently slop-ing north (downstream) margin of the middle terrace of a verydark, refuse-bearing soil stratum not visible on the flat above theslope. Preliminary tests and borings showed that this stratum,heavily mixed with broken animal bone and up to 12 or 15 inchesthick, imderlay much of the terrace at a depth of approximately 24inches. It was overlain by an approximately equal thickness ofbrown sandy soil, devoid of cultural admixture except where refusehad been redistributed in rodent burrows and man-made features.The uppermost 6 to 10 inches consisted of dark-gray humus, withwhich more bones, charcoal, chipped flints, flakes, sherds, etc., wereintermingled. Unhappily, an unknown but reportedly considerableamount of surface soil has been blown from all of these terracesin recent years, so that the later and seemingly more transitory proto-historic and historic stages in the local sequence have been disturbedand in large part obliterated. Wbdel] an introduction TO KANSAS ARCHEOLOGY 383The middle terrace had a breadth of approximately 250 yards ineach direction. Its south, west, and north sides were skirted by thecreek, with the ground rising steeply on the east to the uplandsbordering the valley (pi. 57, a). On the south the stream flowsdirectly against the foot of the terrace and there is a steep bank 18or 20 feet high, but on the west and north progressively wideninggentler slopes intervene between the flat and the channel. A south-to-north profile through the terrace thus would show an abrupt riseof about 18 feet from the creek bed, then a nearly flat surface some125 to 150 yards wide, and finally a gentle glacislike slope of about100 yards to the creek bank at the north. Our tests showed a some-what heavier accumulation of refuse on the surface of the terracejust above the northward slope, and on the eroding upper slope itselfwas the outcrop of dark bone-bearing soil. Along this "break" fromthe flat to the slope, and slightly east of the north-south midlinethrough the terrace most of our excavating was done.To insure uniformity in recording horizontal relationships of finds,the terrace was staked out on a grid of 30-foot squares. Lines run-ning north and south were lettered A, B, C, . . ., beginning on thewest ; those running east and west were numbered 1, 2, 3 . . ., begin-ning on the north. Each 30-foot square in turn, when excavationtherein began, was further divided into 25 units, each 6 feet square.These units were numbered from 1 to 25, beginning in the northwestcorner of the square and at the west end of each successive row ofunits. Thus, the complete designation for any 6-foot square unit wasa trinomial, e. g., square R 26 : 1, and this designation would appearon all specimen bags containing material from that unit. Bulk of thematerial was segregated vertically by 6-inch levels. Outstandingspecimens or features were more precisely recorded by triangulationfrom two corners of a specified square or unit plus exact indicationof depth below ground surface. All our larger tests and excavationson the middle and south terraces were tied into this one system ofcoordinates, and these in turn with the terrace contours, creek, andsprings.^^ I have found it convenient in the ensuing discussion totreat the excavations by areas.It should be stated at the outset that the stratification detected inour preliminary tests was amply confirmed in the subsequent excava-tion, and that it has important connotations which will be discussedat mxore lenglh elsewhere. At the risk of pointing out the obvious,it should be noted that the various strata briefly characterized abovewere not everywhere of uniform thickness or richness of content. The *? In July 1939, the Scott County engineer was engaged for 3 days to survey the Pottorffsite and also "El Cuartele.lo" In nearby Scott County State Park. The contour and othermaps which were to have been prepared have not been delivered at this writing, and ourown sketch maps only are available.484172?59 26 384 BUREAU OF AMERICAN ETHNOLOGY [BULL. 174top layer particularly was often thin, sometimes being virtually ab-sent when the plowed soil was removed. Likewise, the sterile inter-vening layer thinned considerably from the south edge of the terracenorthward, while the buried dark stratum yielded better at the norththan at the south. Such variations, of course, are to be expectedwhere the vagaries of fluctuating climatic conditions operate betweensuccessive human occupations. Far from detracting from the valid-ity of the observed sequence, they merely emphasize the truismthat at any given point it is the relative position and context of anartifact or feature, rather than its absolute depth from an arbitraryplane, that is significant.AKEA A (HOUSES 1 AND 2)On the basis of soil showings in our preliminary tests, we selectedfor more careful investigation a spot at the upper edge of the bone-bearing outcrop and on the nearby flat just above. Here, instead ofthe customary 5 or 6 inches of mixed topsoil underlain by severalinches to a foot of sterile fill, there occurred a heavy deposit of ashlenses, animal bones, stones, sherds, and flints, in such profusion that arefuse deposit of some size was suspected. Careful dissection of thedeposit disclosed the fact that it had accumulated in two small pit-houses lying close together and, in one case, partly intruding thelower buried soil stratum. The "area" lies in squares K26 and S26,and includes about 828 square feet (fig. 76).House 1, the uppermost and later of the two (pi. 58, a), was sub-circular in plan, with an average diameter of about 12 feet 6 inches.At the center, 17 inches underground, was a circular firepit 19 inches [..-> ^.-v -d),number 68 specimens. They are much more variable in size than theplain triangular points of the protohistoric horizons of central and Wbdel] an introduction TO KANSAS ARCHEOLOGY 403 western Kansas, and include most of the larger and thicker unstemmedpoints (over 36 mm. long) from the Pottorff site.Notched triangular points, in several varieties, also number 68, andinclude most of the better made points from the site (pi. 61, f-k).Thirty-one have a single pair of lateral notches 3 to 6 mm. above thebase; 27 have, in addition to the lateral notches, a single notch inthe base ; 5 have two pairs of lateral notches ; and the remaining 5 havetwo pairs of lateral notches and a single basal notch. In the last twogroups, one pair of notches, usually that nearest the base, tends tobe smaller than the other, sometimes little more than a suggestion.The rare stemmed or corner-notched points from the Upper Repub-lican horizon are generally small; but with breadth and thicknesssomewhat greater in relation to length than among the triangularpoints. They are quite well made, however, and the quality of work-manship displayed is obviously much closer to that on the triangularpoints than to that on the larger rougher points from the underlyingstratum. Representative examples are shown in plate 62, e, Z, m.From the surface of the site came 19 additional points, two ofthem heavy stemmed Woodland-like forms 46 and 59 mm. long. Ofthe remaining 17 triangular specimens, 9 were unnotched, 8 werenotched, and all were well within the range of size and craftsmanshipshown in the excavated specimens.Dnlls.?Chipped drills were surprisingly scarce; the excavationsyielded 2 or 3 examples, and there were 4 from the surface. Thereis a thin well-made T-shaped drill of brown jasper from the diggings(pi. 62, n) ; it is 32 mm. lo^g, 15 mm. across the base, and the tip isbroken. The base of a second similar specimen, 19 mm. across andhaving about 4 mm. of the stem, was also uncovered. A third speci-men has a short point expanding to a wide straight base; it is 18 by10 mm., and suggests a retouched triangular projectile point.Surface finds include three fragmentary drills, of expanded basetype; and one complete example with heavy corner-notched base (pi.62, 6) and short stout point. This piece measures 46 by 31 by 7 mm.,is of brown chert, and is possibly a retouched projectile point ofWoodland type.Three spalls, each with edges retouched to give a gradually widen-ing tip may be reamers or perhaps graving tools.Knives.?Bifacially chipped blade fragments showing evidence ofsome care in shaping were quite numerous; the probable range informs may be judged from the series of whole blades shown in plate63, a-g. Diamond-shaped four-edged knives, with adjoining edgesoppositely beveled, were certainly known, though they were not fash-ioned with very great skill or regularity and the beveling tends to beweak (pi. 63, a, h). In addition to the whole specimens illustrated 404 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174there were less than a dozen subtriangular fragments with two con-verging oppositely beveled edges and the third side obviously frac-tured, all suggesting this same "Harahey" type. The completespecimens are 90 by 31 by 8 mm. and 102 by 48 by 9 mm. ; the largestbeveled fragment is 85 by 53 mm., and if from a 4-bladed implementthis must have been considerably larger than either of the wholepieces.More numerous than the beveled knives, apparently, were elliptical,ovate, or leaf-shaped implements (pi. 63, c-^), with one or both endsrounded and the cutting edges convex. Our collections are too limitedto indicate with finality what form or forms were preferred and theirrespective abundance, but as the illustration suggests, the severalshapes probably intergrade. These knives are 7 to 12 mm. thick, 30to 67 mm. or more wide, and 50 to 90 or more mm. long. Retouchingvaries, of course ; on some specimens it is fine and regular, on others,coarse and uneven.Two leaf-shaped objects of dark-gray plagioclase basalt, each withone end broadly rounded, the other pointed, are either small knives orelse artifact blanks. They are 55 by 33 by 7 mm. and 62 by 40 by 8 mm.End scrapers.?Upward of 200 whole and broken end scrapers werefound. All have in common the planoconvex cross section with a widerounded steeply chipped working end, and a narrow tapering "tail"(pi. 62, p-s). There is the usual variability in details of form, relativethickness, etc. Size ranges from 22 by 17 by 4 mm. to 62 by 25 by11 mm. Of 185 specimens sorted by length, 48 (about 26 percent) areunder 30 mm., 108 (58.4 percent) are between 30 and 45 mm., and 29are over 45 mm. long. Though this breakdown is not strictly com-parable with that for other sites in central and western Kansas, aspresented elsewhere in this report, it can be stated that the scrapersfrom Pottorff are somewhat larger and heavier than those from proto-historic sites of the region.Side scrapers.?These were much less plentiful than end scrapers,there being not more than 50 examples. They were fashioned fromspalls and flakes of varying sizes and shapes ; each has one flat, curved,or "twisted" plane surface, the other being convex and often showingseveral long flake scars. Eetouching was confined to the edge, andwas always on the convex surface. Some of the smaller specimensapproximate the prismatic flake knives in size and shape; others aremuch larger and rudely shaped, none showing the careful workman-ship of the knives. In size they range from 51 by 18 mm. up to about80 by 40 mm.Large chipped implements.?There are about a dozen large heavycoarsely chipped objects, varying from more or less ovoid axlike orceltlike implements (pi. 63, h-l) to broad flat round-ended scraper or Wbdhl] an ESTTRODUCTION to KANSAS ARCHEOLOGY 405hoe forms. They have only primary flaking on both faces; there isno clear evidence that they were hafted, though in a few instancesportions of the edges appear to have been ground down somewhat asif to ease the wear on a lashing or handle. The material is either brownjasper or a gray chert, evidently from a seam not more than 25 or 30mm. thick ; several specimens have remnants of the enclosing calcar-eous matrix on both faces. Most of the implements range from 93 to150 mm. long, 54 to 86 mm. wide, and 19 to 30 mm. thick ; but thereis one large broken specimen 185 by 120 mm.Objects of Stone Othee Than" ChippedMealing slabs (?).?From the refuse-filled house pits came twodoubtful fragments that may have come from grinding slabs. Oneis from the roimded edge of a concavo-convex slab of calcareousquartzite, about 38 mm. thick. Both slab surfaces look worn, butwhether from grinding or through some other agency is not clear.The other specimen, of very coarse sandy conglomerate, shows part ofa concave upper surface that may have been a grinding basin.A third piece from the north slope of the terrace is from a flat sand-stone slab, 20 mm. thick, with smoothed even surfaces. It appears tobe from a much more carefully dressed object than either of the fore-going, but is too small to be conclusive evidence for the presence ofmealing slabs. A surface fragment is unquestionably from such aslab ; however, its associations are uncertain.Manos.?There is one complete mano or hand stone?a thick sub-rectangular quartzite block with rounded angles and one worngrinding surface. The nongrinding surfaces are dimpled from ham-mer-dressing, evidently to shape the object. It is 140 by 85 by 63 mm.The second piece is doubtful. It is the end of a broken quartzitecobble, with one surface apparently smoothed by grinding. Its orig-inal size and shape, like its use, are conjectural.Sinew stones (?).?^A quartzite cobble, 90 by 70 by 40 mm,, hasshallow grooves running lengthwise along the rubbed upper surface,and down onto the rounded edges. The grooves look like the resultof prolonged abrasion by soft yielding substances, such as skin, sinew,or a cord, which were drawn back and forth over the stone. The endsare slightly battered, as from use as a hanmierstone.A small fragment from the end of a second quartzite pebble alsohas a broad shallow groove with subdued striations like the above.Rubbing stone.?This is a planoconvex oblong sandstone block, fireblackened, and showing spotty evidences of abrasion on the some-what uneven flat surface. It fits the hand conveniently, and mayhave been used for hide dressing or otherwise on a yielding surface.It measures 108 by 88 by 28 mm. 406 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174Shaft smoothers.?Less than a dozen shaped sandstone fragmentsare from longitudinally grooved abrading stones. Most have onesurface flattened, the others slightly convex; and the presence ofshallow haphazard grooves on 2 or 3 surfaces suggests re-use of frag-ments for sharpening bone implements, etc. The longest specimen,exact provenience unknown, measures 112 by 26 by 24 mm., and isbroken ojff at one end ; the flattened upper surface has a deep straightfull length groove 8 to 9 mm. in diameter, with sliorter grooves onthe other sides. Other fragments are up to 60 mm. long, 25 to 40 mm.wide, and 15 to 30 mm. thick, thus conforming in most particulars ofsize and form to those found at numerous other Plains sites. Thestone used includes both red-brown and grayish-white varieties, allof good abrasive qualities.There is no direct proof that these implements were used in pairs,as they unquestionably were in historic times, but their general sim-ilarity suggests such a method.Abrading stones.?There were several oblong worn blocks of sand-stone, with flattened sides and rounded ends, that were probablyused as polishing, smoothing, or finishing stones. They were lessregular, but relatively broader and thinner, than the shaft smoothers,and in all cases lacked the grooves. Size varies up to about 80 by 45by 25 mm.Stone pipe.?This is a poorly preserved and fragmentary specimen,of soft friable gray limestone. It appears to have been a simple,heavy, elbow-shaped affair, perhaps 50 mm. high, 35 to 40 mm. thick,with a short square stem-arm not over 40 mm. long. The taperedcavities have been reamed out to maximum diameters of 16 or 17 mm.,and the bowl interior is slightly blackened. There is no decoration.Hammerstones.?About 20 variously shaped but more or lessrounded stones of quartzite or granite have battered ends, corners, oredges. In occasional examples, this battering has reduced the stonesto nearly spherical form. They vary from 65 by 50 by 30 mm. to100 by 60 by 45 mm. To what precise use they were put is notclear, but in view of the relative scarcity of pecked and ground stoneobjects, I would suspect they were employed perhaps in the primaryshaping and roughing out of stone knives and heavier chopping orskin-dressing tools.Stone phallus.?From a small test hole on the soutli slope of themain site terrace was taken a simple carved block of faintly lami-nated metamorphic stone. The block is squarish in cross section withrounded corners, and at one end becomes more bulbous and downcurving. At the extreme tip there is a short deep vertical cut ; and oneach side of the bulbous portion, there is an oblique incision beginniugon the underside and rising to intersect the opposite incision on the Whdel] an introduction TO KANSAS ARCHEOLOGY 407upper side. The piece is an unmistakable representation of the malegenerative organ. It is 136 mm. long, 39 mm. wide, and 35 mm.thick (pi. 56, A;). 'Worked amazonite.?From a small shallow test near the west endof the terrace came a triangular bit of flattened dressed stone, withtwo broken edges and one curved. The material, a pale robin's ^ggblue, was at first mistaken for turquoise, but has since been identifiedas amazonite. It probably is a fragment from a small pendant.Pigments.?A few lumps of red-brown hematite and yellow limon-ite, with striations, grooves, and grinding facets undoubtedly indicatethe source of some of the paints used by the natives.Concretion.?^A single marcasite concretion, flattened circular inshape and measuring 37 by 59 mm., was found. Its surface lookssomewhat worn, but from what sort of usage I am unable to say.These concretions are at present weathering in quantities from theshale beds bordering Salt Creek just below the main terrace, andthis is midoubtedly the source of the specimen we found. They occurrather plentifully in the refuse deposits at the protohistoric ScottCounty Park site ("El Cuartelejo"), as will be noted elsewhere.Objects of ShellPendants.?From area B, east of the house sites, came two ratherirregular pieces of mussel shell drilled for suspension (pi. 61 c>, f).One is elliptical with well-worn rounded edges; it is 40 by 22 mm.,with a 4-mm. perforation. The second piece is roughly triangular,and includes part of the hinge; it measures 36 by 27 mm. Neitherpiece is decorated.Dish head.?A single circular bead, 14 mm. across by 2 mm. thick,biconically bored, was found on the surface. Its relationship to theUpper Republican stratum is not clear. abtifacts from occupation b ( woodland)Food RemainsThere were no traces of domestic or wild plant food from thelower level of the site. It is a safe guess, of course, that these Indians,as well as the later ones, made extensive use of the available berries,fruits, roots, and other food in and about the creek valley. Thatthey also grew corn, beans, or squash seems rather less likely, sincesuitable digging tools have not been found, and the single muUercould as well have served to crush berries. That corn and beanswere known on this general time level is evident from the finding ofcharred corn in a Nebraska Woodland site and also at the Rennersite near Kansas City (Wedel, 1943, p. 26), but that it was actuallygrown and used on the Pottorff site cannot be demonstrated. Thefeeling persists that this was primarily a hunting station. 408 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174Identifiable refuse bone from this level was not very plentiful,though slivered and broken bone was everywhere present. Bonesrecognized included the following: bison, 7; antelope, 16; white-tailed deer, 9 ; kit fox, 1 ; dog, 1 ; wolf, 1 ; coyote,10 ; prairie dog, 25 ; cottontail rabbit, 5; black-tailed jackrabbit, 3; badger, 1. Oursample, of course, is small; but it is interesting to note the relativeprominence of antelope, deer, and other small forms over bison,whereas bison overwhelmingly predominated in the upper level.The possible significance of this difference, observed also in certainNebraska localities, is discussed elsewhere in this paper.Bird bones apparently from this level represent three species : thegolden eagle {Aquila chrysaetos), blue-winged teal (Anas discors)^and red-tailed hawk (Buteo jamaicensis) . All were at depths of12 to 24 inches in area B, but unfortunately where there had beensome disturbance by rodents. The possibility that these bones actu-ally belonged to the later occupation cannot be entirely ruled out.MoUuscan remains were not plentiful, although bits of shell wereencountered rather frequently. Specimens of the following twospecies, both locally obtainable creek forms found also in the upperlevel of the site, have been identified : Ligumia suhrostrata (Say) 6Pleurohema coccineum (Conrad) 1POTTEEYFrom the lower occupation level there are but 95 potsherds, includ-ing 5 rim fragments and 90 body sherds. There are no duplicatesamong the rims, from which it can be concluded that at least fivevessels are represented. Most of the sherds are small, though theoccasional larger ones considerably exceed in size any of those fromthe upper level. Our work does not suggest that these earlier peopleswere so well provided with earthenware as were their successors ; thesherds were widely scattered, and matching or restoration of any ofthe pieces is impossible. Had we been able to locate a refuse depositfor this horizon, as we did for the overlying stratum in the aban-doned house pits, it is possible our labors would have been moregenerously rewarded.The pottery in general is readily distinguished from that of theUpper Republican stratum, and appears to conform closely toKivett's (1953, p. 131) Harlan Cord-Roughened type in Nebraska.Eighty sherds are tempered with coarse angular particles of crushedcalcite, from 1 to 7 mm. in maximum dimension; silicious matter isvery scarce or absent. The calcite inclusions are usually presentin abundance, and can be readily seen with the unaided eye ; occasionalsherds are so thickly tempered that inner surfaces appear to be Wedkl] an introduction TO KANSAS ARCHEOLOGY 409 sprinkled with shell particles, and leaching sometimes imparts avery irregular pitted or "hole-tempered" effect. The fracture surfaceis granular, its coarseness or fineness depending upon the texture andabundance of the aplastic. Hardness very rarely exceeds 2.5(apatite), and the sherds are much softer, more friable, and far lessresistant to water than the Upper Kepublican materials. In colorthe paste varies from light gray or buff to almost black on a freshbreak, with considerable variation in a single sherd. The surfacecolor now, even after washing, is a very light gray but this appearsto be due to a calcareous patina deposited by ground waters (pi.60, ^^).Almost without exception the sherds bear closely spaced cord-impressions on the exterior surface ; the impressions are usually widerand deeper, and often more closely spaced, than those on UpperEepublican sherds (cf. pi. 60, i, j, and l-n) . From the sherd contours,I suspect the impressions were generally vertical on the vessel, andthey seem to have been continuous over large areas. Two sherdsfrom area B have deep single-line fabric (?) imprints separated bynarrow sharp partly overhanging ridges. Casts in clay of these im-prints resemble strings of small beads ; they do not suggest knottingor a wrapped implement or twisted element. There is some resem-blance, too, to twined basketry impressions, though the placement ofthe markings and their depth argue against such an identification.Little evidence exists as to vessel shapes, but there is one fragmentunquestionably from a tapered subconoidal base. The very slightcurvature shown by most of the sherds further suggests jars of con-siderable height and girth, but their proportions remain conjectural.Vessel walls ranged in thickness from 7 to 15 mm., well in excess ofthe figures for our Upper Eepublican sherds. Rims, of which thereare four, appear to have been vertical or slightly flaring, with eitherrounded, or flattened and slightly thickened, lip; cord impressionsoccur on two lip sherds. There is no evidence of handles, lugs, effigies,or other modeled features, or of incising.Fifteen sherds differ somewhat from the foregoing series. Theyare tempered with noncalcitic rounded particles of grit, from 0.3to 2.0 mm. in diameter, and sparingly to moderately abundant; thefreshly fractured surface is finely granular, with a fine, dense, evenpaste. Hardness is 3 (calcite) to 4 (fluorite), and color is light buffto dark gray. Six sherds have deep closely spaced vertical cord im-pressions on the exterior; four have widely spaced (2-3 mm. apart)cord impressions on the exterior with partly obliterated imprints onthe interior as well ; and four have plain somewhat unevenly smoothedsurfaces. Thickness of sherds is 6 to 11 mm. Vessel shapes againare uncertain; the single rimsherd in the series has a plain rounded 410 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174lip, with cord impressions ending 6 to 10 mm. below on the exteriorsurface. . ? , ^ jMention has already been made of the findmg of a rocker-roughenedsherd 9 inches beneath the floor of house 2. This piece (USNM387276) has a fine black paste, tempered with fine sand particles, andwith a surface hardness of about 4. The uneven interior surface washeavily encrusted with carbonized matter; the exterior was unevenlysmoothed. The sherd (pi. 60, k) is crossed by a narrow curving in-cised or trailed line separating a plain portion from a decorated area.The latter consists of two rows of apparently vertical edentate rockerimpressions, each curved impression up to 15 mm. long. The decora-tion is sloppily done and the pattern, of course, is miknown, but thetechnique seems clear enough. None of the vessel rim is present,and the original shape and size of the jar are conjectural. The rockerstamping, though crude, recalls that on Hopewellian sherds from theKansas City district (Wedel, 1943, pp. 29-^1) . The very dark, finelytempered, sooty paste, however, suggests neither Kansas City Hope-weUian nor such local western Plains Woodland wares as I have seen,but rather certain sherds from the Andrew Snyder village site, Cal-hoim County, 111., recently presented to the National Museum by P. 1^ . Titterington. Objects of Bone and ShellWorked bone and shell from the Occupation B level was veryscarce; in fact, only four artifacts made of these materials were found.Just outside the northeast edge of house 1, probably m the refusebelow floor level at a depth of 16 to 22 inches underground, were foundtwo partly pierced bison phalanges. In both, part of the posteriorsurface of the bone has been crudely hacked away to form a largeirregular oblong hole into the marrow cavity. The holes are 16 and25 mm. long; their margins show no evidence of smoothing or finish-ing, and the bones are otherwise unmodified. The specimens closelyparallel those from the upper level, described in a preceding section,as well as a series from a Woodland site in Valley County, Nebr. (Hilland Ki vett, 1941, p. 163) . Their use is unknown (pi. 64, 6) . In area B, at a depth of 16 inches, was found the awllike objectillustrated in plate 61, m. Made of split mammal legbone, it has onerounded and one pointed end, both somewhat chipped. The surface,where not "etched," shows traces of high polish, and the edges havebeen dressed. Along one edge are 18 notches at 2 to 3 mm. intervals.The specimen is 74 by 9 by 4 mm.Also from area B, at 18 to 34 inches depth, came a small fragmentof worked fresh-water mussel shell. The piece is in-egular m shape,measures 26 by 30 mm., and on a short section of the original ventralmar^nn there are 4 notches at 4- to 5-mm. intervals. Excepting the Wbdbl] an introduction TO KANSAS ARCHEOLOGY 411 notches there are no evidences of modification, and I am unable to saywhether the fragment is from a simple serrate ornament or from autilitarian object (pi. 61, n) . Objects of Chipped StoneChipped-stone artifacts were nowhere plentiful in the lower occu-pation stratum. By comparison with the more abundant materialsfrom the upper level. Occupation A, they were, moreover, less skill-fully fashioned and more rouglily finished. With few exceptions,the objects represent spalls, thick chips, and cores, with but an occa-sional piece made from the thin delicate flakes used so extensively bythe later peoples for projectile points, drills, and other small objects.Chipping varies in coarseness, and even on blade edges is never sofine and regular as that on the upper level artifacts. Jasper, dark-colored cherts, and quartzite are the principal materials used; chal-cedony, moss agate, obsidian, and variegated cherts, which lend apleasing variety to the products of the later peoples, are rare or absent.Seven complete and fragmentary projectile points were taken fromthe lower cultural stratum. Five are corner notched or stemmed (pi.64, f-h), with well-marked shoulders, expanding stem, and convex, orin one case concave, base ; blade edges are slightly convex. Completespecimens are 35 to 76 mm. long, 22 to 28 mm. wide, and 5 to 9 mm.thick. One of the specimens, plate 64, A, has a rounding tip which doesnot look like mere retouching of a broken projectile point, thoughotherwise the piece is of the same size and general type as the otherpoints. The two remaining points are different. One is a small thinbroken triangle about 26 mm. long, with convex sides and slightlyconcave base. It was found in the upper part of the lower zone nearthe house pits, but at a point where the two occupation layers touch,and there is a possibility of intrusive entry into the Woodland stratum.Another point, 27 by 16 by 3 mm., has prominent backward-curvingbarbs, and a short straight-sided stem with straight base. Like thepreceding, the teclmique on tliis piece strongly suggests the UpperRepublican industry and the point occurred in an area crisscrossedwith old rodent burrows. Whatever the correct provenience of thesetwo small points, it is clear that a majority of the projectile pointsfrom the Woodland stratum are of a larger stemmed form markedlyunlike the small Upper Republican points. Two additional heavystemmed points, in every respect like those described above, were col-lected from the surface of the main terrace.Eight planoconvex end scrapers are of small to medium size,conforming in all particulars to the familiar "snub-nose" scraper ofthe Plains. They are 30 to 52 mm. long by 19 to 25 mm. wide. Allare unifacially retouched except one, which has been flaked on theflat ventral surface as well (pi. 64, i) . 412 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174Six side scrapers are also planoconvex but lack the careful overallshaping of the end scrapers. They are elongate, in outline followingthe conformations of the original spall, and have a long edge, notthe end, retouched for working. The largest is 55 by 30 by 8 mm.Two thin curved prismatic flakes bear two or three flake scarseach on the back and have the edge or edges retouched for cuttingor scraping. They are 51 to 55 mm. long by 15 to 18 mm. wide.Of shaped cutting, chopping, or scraping tools other than thosegiven above, there is a notable scarcity. There are several thickoblong to subtriangular objects that suggest chopping tools or un-finished implements. They are fasliioned of jasper taken from thinveins in a calcareous matrix. They may be implement blanks; butsince they are virtually the only objects from the lower stratum thatcan be regarded as possibly cutting tools, they should perhaps beclassed as celts (pi. 64, c, e). Rare fragments of thinner pieces,retouched from both sides, may have been knives.Objects op Ground StoneFrom the deep stratitest just south of area B, at 24 to 30 inchesdepth, came an irregular bun-shaped quartzite cobble. The edgesshow some evidence of workmanship ; the flat underside is well worn,the convex surface less so. The stone measures 100 by 94 by 40 mm.,and fits the hand conveniently. It was used, presumably, with arotary motion. There were no mortars or mealing slabs from thishorizon, and I cannot say what substances may have been pulverizedwith the implement.There is also a spall from a lump of vesicular basalt, that showstraces of wear. It is thin, but the diameter is about the same as inthe above piece. Possibly it was used in hide dressing.FUTUBE INVESTIGATIONSThere can be no gainsaying the fact that our very limited excava-tions at the Pottorff site indicate a succession of prehistoric cul-tures?a succession borne out by other nearby findings that are yetto be considered. The significance of this finding will be consideredelsewhere. Meanwhile, it is a safe prediction that other creek valleysin the High Plains region, where permanent springs and shelteredflood-free terraces occur together, can be expected to afford additionalinstances of such culture stratification. Large or spectacular sitesare not to be expected, but there is excellent opportunity for worth-while findings bearing on the adaption of primitive man to a some-what trying and certainly variable natural environment. On thebasis of our experience, I should like to suggest to future studentscertain procedures which, I freely admit, we should have put intopractice. Wbdbl] an introduction TO KANSAS ARCHEOLOGY 413In the first place, there is need for a far more precise and extendedcorrelation of cultural and noncultural strata than is obtainable fromsmall-scale and discontinuous excavations such as ours. A masterprofile trench up to 6 or 8 feet deep, running north and south throughthe main terrace, and provided with laterals at promising pointswould probably have solved several intriguing questions at presentleft open?e. g., the nature of the suggested third and oldest culturalstratum (Occupation C), the relation of the occupation horizons onthe terrace proper to the deposits on the slopes, the attribution ofthe various remains to transitory hunting camps or to semipermanentpartly horticultural villages, and the specific nature of the nonhumanagencies that built up the terrace. For this, obviously, more time anda great deal more manpower than we had will have to be available.Moreover, there should be an interdisciplinary arrangement wherebya qualified geologist, physiographer, or sedimentation expert couldbe called in for consultation. The necessity for such cooperativeendeavors has been voiced before by others; it acquires urgency asthe number of known buried cultural horizons and of stratified sitesincreases. These sites provide tangible proof that primitive man'soccupancy of the region has been linked with fluctuations of climateand resultant modifications of the landscape. Few, if any, arche-ologists have the highly specialized training in geological processesthat is required for the study, interpretation, and correlation ofstream terraces, yet this phase of the evidence is of unquestioned im-portance in reconstruction of past climatic conditions and their influ-ence on human activities. I doubt that we shall have more than avery superficial impression of primitive human ecology in the Plainsuntil a combined attack has been made on the geology of a represen-tative series of the preliistoric stratified sites that we now know toexist. THE RISSTON SITE (14SC4)A little less than a mile southwest of the Pottorff site, where SaltCreek swings from a northerly to a northeasterly course before leav-ing Scott County, three small inconspicuous terraces have been par-tially or wholly isolated by erosion. They are from 50 to 150 yardsnorth and west of the left bank of the creek, at the foot of a longgentle slope that here forms the west margin of the valley. To theeast and southeast are calcareous bluffs up to 50 feet high; at theirbase is the creek, which has running water up to this point but onlyoccasional permanent ponds above. Hackberry and cottonwood fringethe stream at the base of the bluff, with some prairie grasses on theadjacent valley bottom, but otherwise the gently rolling landscapeis dominated by a thin cover of short grass, yucca, and sage (pi. 57, &) . The terraces concerned are much smaller and less sheltered against 414 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174winds than those at the Pottorff site, and seem far less desirable fromthe standpoint of hmnan habitation. Their surfaces, however, fromwhich a foot or more of plow-loosened topsoil is said to have beenblown during the dust storms of 1935 and 1936, are littered withsherds, flints, bone fragments, burned stones, and similar Indianrefuse, and the location is a popular one for local collectors fromnearby towns. Lesser quantities of refuse are to be fomid widelyscattered over the slopes to the west, suggesting by their amount anddiversity the frequent if transitory utilization of the locality byIndians for untold centuries.Of the three terraces, the isolated central one appeared most prom-ising on a preliminary surface examination, and 10 small test pitswere accordingly dug along the midline of its summit. These variedin depth from 22 inches to nearly 6 feet, and were dug a foot or morebelow any sign of cultural admixture. Such admixture was presentin varying amounts in all of the tests, begimiing always at the grassline and ending at a depth of 10 to 18 inches below the surface. Itconsisted of dark-colored soil, mixed with charcoal, bone refuse,burned stones, and rare sherds. Specimens of diagnostic value wereextremely scarce. Test No. 5, near the middle of the terrace, yieldeda fragment of sandstone or quartzite mealing slab, a pecking stone ofblack plagioclase basalt, and a large heavy calcite-tempered sherdbearing deep impressions of a tightly twisted cord. In test No. 8,near the south end of the terrace, there came to light a small circular})it 23 inches in diameter by 22 inches deep; the fill contained muchcharcoal, a tubular bone bead, a thick cord-roughened calcite-tem-pered sherd, and the base of a heavy stemmed or corner-notched pro-jectile point. The purpose of this pit is not clear ; no other suggestionsof aboriginal structures were noted, and the artifact yield generallywas so low that expansion of the tests was not regarded as worthwhile.At the extreme south end of the terrace a small plot of unbrokenprairie sod remained. Here, under circumstances that probablymore closely parallel the soil conditions prior to modern agriculturaldevelopment and accelerated wind erosion, we made our tenth test.The uppermost 8 inches consisted of clean light sandy soil, which wasunderlain by a darker formation containing flint cliips, bone frag-ments, charcoal, and a calcite-tempered cord-impressed sherd. Be-low this occupation zone came more unmixed soil, which continueddownward without break or apparent variation to at least 6 feet. Isuspect that the dark refuse-bearing stratum was once similarly sealedin over the entire terrace by a noncultural deposit that has since beenremoved except in those small spots wliere the sod cover remainsundamaged. Wbdbl] an introduction TO KANSAS ARCHEOLOGY 415Test pits were also dug on the other two terraces, east and westof the central one, with substantially the same results. In addition,a trench 30 feet long was cut across the south end of the east terrace,whose eroding edges showed an outcropping detrital zone. Hereagain mixed soil occurred from the surface to a depth of 16 or 18inches; the fill was very dark in color and contained much charcoal,flint scrap, and bones. Three mullers, two of quartzite, one of verycoarse sandstone, and all with a single worn grinding surface, camefrom the trench; one was bun shaped, one subtriangular, the thirdsubrectangular. Other specimens included the convex base of a brownjasper blade, a calcite-tempered sherd with nearly obliterated cord-roughening, and 10 cord-roughened sherds with rounded gravel orcoarse sand inclusions. This culture-bearing zone faded graduallydownward into clean unmixed soil which, to judge from auger tests,continued to at least 5 feet.The surface materials that we collected on the terraces from timeto time during our stay in the locality strongly suggest at least twodistinct occupations. There are relatively few sherds of the heavycalcite-tempered deeply cord-impressed pottery, and a great many ofa thinner, harder ware containing rounded gravel or sand and havinga cord-roughened exterior. One or two broadly incised or trailedgrit-tempered rim fragments somewhat suggest Lower Loup waresfrom east-central Nebraska, but are too small to be definitive. Pro-jectile points are of two general types : a small, well-made triangularform with or without notches, and a much larger corner-notched orstemmed form. Of less diagnostic value at present are fragments ofmealing slabs and mullers, end scrapers, drills, thick roughly chippedlanceolate implements, and numerous other cutting and scraping tools.Private collections from the site show the same range of artifact formssuggested in our sample, including the two or more distinct potterywares and projectile point types noted above.On the basis of our work at the nearby Pottorff site and otherfindings elsewhere in the Central Plains it seems reasonable to inferthat at least two distinct cultures are represented on the Risston site.To one may be attributed the thinner sand-tempered pottery andsmall notched and unnotched points; to the other belong the thickcalcite-tempered sherds and heavy stemmed points. Stratigraphicevidence to demonstrate their temporal relationship cannot be ad-duced for the Risston site, beyond the partly impressionistic observa-tion that most surface sherds seem to be gravel tempered and thepoints small, whereas the heavier calcite-tempered sherds and largepoints predominate underground. It is tempting to conclude thatthe Risston terraces were once stratified like those at the Pottorffsite; that wind action has carried away the soil of the later occupa-484172?59 28 416 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174tion, precipitating the sherds and other cultural debris on thepresent lowered erosional surface ; and that the cultural deposit foundin all our tests represents only the earliest occupation of the spotby pottery-making groups. This would be a credible sequence ofevents, more credible at least than the assumption that all the remainswere left by one people. Nevertheless, it must be emphasized thatthe suggested sequence is inferential, and cannot be directly sub-stantiated by any evidence now available to me from the Risston site.Local collectors reported the presence of flints, stones, bone refuse,etc., along the creek banks for perhaps a mile upstream from theterraces just discussed?about as far up, that is, as tree clumpsand waterholes may be found. Inspection by members of our partyshowed that these remains were thinly scattered, and revealed noconcentrations that seemed to justify excavation.The rocky bluffs that line the east side of the valley above thePottorff site contain a few small scattered overhangs and niches,generally too small for habitation but of suitable size for burial orcache purposes. From one such overhang opposite the Risston ranchhouse, and roughly equidistant from the Pottorff and Risston sites,a burial is said to have been exhumed some years ago, but we wereunable to secure any details concerning the find. Two members ofour party sifted some of the dust from the floor of the shelter, re-covering a number of small blue, white, and red glass beads. Thisindicates that, as might be expected, historic Indians had visited theoverhang; but no evidence came to light that the prehistoric groupsof the vicinity had used it or any of the other smaller cavities inthe locality. SITE ON WALNUT CREEK, LANE COUNTYTHE WALTER SITE (14LA2)Walnut Creek, which joins the Arkansas River about 4 miles belowGreat Bend, rises nearly 100 miles to the w^est in the Tertiary uplandsof Lane County. The greater part of its course is in the PlainsBorder region, and is characterized by a wide pleasant valley, oftenwith picturesque bluffs, and with a good flow of water. The immedi-ate banks are lined with hardwood forest, and in the old days, thebottoms were covered with prairie grasses. All this should have ap-pealed strongly to the native peoples, hunters as well as farmers, andit is to be expected that numerous camp and village sites will ulti-mately come to light in the valley. At present, however, almost noth-ing is known of its prewhite inhabitants, other than the report ofa site near the mouth and another? the subject of the present dis-cussion?^near its source. Wbdbl] an introduction to KANSAS ARCHEOLOGY 417The Walter site, half a mile south of Alamota in eastern LaneCounty, lies on a small well-concealed terrace on the left bank ofthe south fork of Walnut Creek. The creek, now dry at this pointbut said to have carried a good flow of water 15 or 20 years ago,skirts the north, west, and south sides of the terrace, with moderatelyhigh bluffs on its right bank. At the south edge of the terrace, 25 or 30yards from the formerly inhabited area, are the remains of an oldstone-walled spring. Of the timber lining the creek banks, cotton-wood and hackberry are native, whereas walnut and ash are said tohave been introduced by the early wliite settlers.^^The terrace includes less than 2 acres of ground, much of whichis evidently too low and sandy to have invited occupancy by man.Most of the cultural deposits seem to have lain near the central andhighest point, where an irregular area measuring about 20 by 35 yardshad been completely dug over to a depth of 3 to 5 feet by the formerowner, Mr. E. W. Walter. A considerable collection of broken potteryand stone implements was amassed in the course of this digging, andwas placed in the excavator's private cabinet. Inspection of thismaterial showed that at least two major pottery wares were repre-sented, and that there were also significant variations in the projectilepoint types. These artifact differences, together with the unusualdepth to which the remains occurred and the complete lack of anyrecord as to specimen provenience, prompted our very brief follow-upwork. Unfortunately for us, Mr. Walter's digging seems to havevirtually exhausted the site ; we can offer little more than a suggestionas to the cultural stratigraphy that appears to have once existed onthe terrace.^The Museum party's excavations consisted of an L-shaped cut intothe northeast side of the Walter diggings, which resulted in a 25-foot north-south profile and a 10-foot east-west profile. The cut was5 feet wide, except where the irregular edge of the earlier worknarrowed its width, and the depth varied from 45 to 56 inches. Thesoil was removed in 6-inch strata. Several test pits and auger bor-ings were put down in various parts of the terrace by way of furtherchecking the depth and area of occupation.In the uppermost 4 to 6 inches of topsoil were found several bitsof iron (possibly modern farm refuse), an iron arrowpoint, and one " Lieut. I. C. Woodruff's map of his 1852 reconnaissance of a military route from FortLeavenworth to the Arkansas River (Nat. Arch., Cartographic Div., Record Group 77,Map U. S. 187) notes that Walnut Creek was "wooded from its mouth 75 miles" ; and thatbelow the forks there was "water running." The south fork bears the legend "Buffalo Cr,water in holes." The north and south forks of the Walnut unite a few miles south ofpresent Ness City, presumably about the western limit of timber in Woodruff's day. TheWalter site is some 20 miles by stream farther up the South Fork.**The writer, though he inspected the site and the Walter collection, was not presentduring the excavation here described. The work was in charge of my assistant, M. F.Kivett, aided by J. M. Shlppee and Charles Walter, son of the owner. 418 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174small blue glass bead. Below this was a light-gray silty soil con-taining some sand and mixed with charcoal, bone refuse, flint chips,fragmentary stone implements, and thin cord-impressed sherds tem-pered with rounded gravel particles. There was some slight evidenceof concentration of this material at about the 12- to 16-inch level,though it occurred elsewhere as well to a depth of about 22 inches.Below this was a much darker and more granular soil with fine cal-careous veining and many fossil hackberry seeds ; charcoal was moreabundant than in the overlying stratum, as were also mussel shellsand burned limestone fragments. Artifacts were not plentiful butthere were a number of thick cord-impressed calcite-tempered sherds,stemmed and corner-notched points, end scrapers, miscellaneousretouched spalls, and other odds and ends. This zone, its darkcolor in striking contrast to the formations above and below, varied inthickness from 13 to 22 inches, and was underlain by a compactlight yellow subsoil of loesslike texture and appearance. Augertests indicated that this unmixed subsoil continued at least 5 feetdeeper, becoming increasingly bright yellow in color.No fireplaces, cache pits, lodge sites, or other structures were notedin the main cut, nor is there any way of learning at this late datewhether such features occurred in the area previously dug. Oneof the tests by Kivett's party, 7 yards east of the excavated area, dis-closed an ash-filled fireplace 12 inches in diameter and 8 inches beneaththe surface. An area about 30 inches across and 8 inches deep,centering at this ash bed, had been darkened and discolored by fires.Otherwise, none of the tests proved very productive, suggesting thatso far as subsurface remains are concerned the area of occupationwas small indeed. It seems rather improbable that further excava-tion on a larger scale would be warranted, unless perhaps in the sectionimmediately east of our trench, and even here it is likely that only thefading margin of the older occupations still remains.As might have been expected from a small random and apparentlymarginal excavation such as ours, relatively few artifacts were re-covered. Nevertheless, if they be considered in relation to the soilstratum from which they came, certain stratigraphic differences seemindicated. Thus, all metal and glass objects came from the upper 9inches. They included, in addition to items already mentioned above,a piece of flat strap iron 4-1 mm. long by 20 mm. wide, with one endcut square and the other deeply notched as though a triangular point20 by 29 mm. had been cut out. This may represent scrap left overfrom the manufacture of an iron arrowpoint. Implements of stoneincluded five or six fragments of end scrapers, a crude quartzite knife,a chalcedony flake knife, a thick chert core, and a fragment of sand-stone muller or rubbing stone. Better made than any of these was a Wedbl] an llSrTRODUCTION TO KANSAS ARCHEOLOGY 419triangular brown chert knife measuring 73 by 43 mm., with the longedges oppositely beveled and the short side bifacially chipped. Theonly bone artifact was an awl 10.7 cm. long, made from split anddressed mammal bone, with a romided butt and evenly tapered point.No pottery was found in the upper 12 inches.From the 12- to 24-inch level, lying almost wholly above the darkburied stratum, came a broadly lanceolate curved brown jasper blademeasuring 48 by 26 mm., with the edges retouched; an oblong cuttingor scraping tool of weathered yellow chert 65 by 45 mm. in size ; mis-cellaneous unworked or only slightly retouched flakes and spalls;the joint end of a small cut bird bone bearing several transversescratches that may be flint knife (?) scars ; and six cord-roughenedpotsherds. One of these, from above the 18-inch level, had a fine densegray paste sparingly tempered with moderately fine (mostly under0.25 mm. diameter) rounded particles of sand or gravel. The others,from below the 18-inch level, were tliicker, with worn or weatheredsurfaces, and were heavily tempered with coarse angular fragments ofcrushed calcite.Below the 24-inch level, in the dark buried stratum, stoneworkconsisted mainly of spalls, chips, two projectile points and a fewbroken end scrapers. One of the points, of semitranslucent reddishchalcedony, measured 39 by 22 mm. with well-defined shoulders and anexpanding flat-based stem ; it came from a depth of 24 inches, appar-ently within the dark buried stratum. The second point, 26 inchesdeep, was of black basalt, with strongly marked barbs, a short broadstem, and dimensions of 35 by 35 mm. The sherds, 14 in number,occurred to a maximum depth of 38 inches; all but one were cordroughened, and inclusions consisted of crushed calcite or coarse (up to3 mm. diameter) gravel, or both. These sherds ranged in thicknessfrom 7 to 11 mm.From the various smaller tests away from the old diggings camea crude drill, a unifacially retouched subquadrilateral knife or sidescraper, a crude triangular point, spalls, animal bones, shell frag-ments, and two sherds?all from the upper 24 inches, and presumablyabove the dark stratum. Both sherds were thin (about 4 mm.) , finelycord roughened with the impressions lightly rubbed over, and had acompact dark paste sparingly mixed with moderately fine roundedgravel. One was a rim fragment, with channeled interior and moder-ately collared exterior, the latter showing traces of two carelesslyapplied incised lines which once encircled the vessel just below thelip. These sherds conform to the single specimen found at the 18- to24-inch level, and to a fourth very small fragment taken from thehearth at a depth of 8 inches underground. Significantly, these fourspecimens for which Upper Republican affinities are strongly sug- 420 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174gested, lay above the dark stratum wherein were the thicker calcite-tempered sherds and stemmed projectile points that have equallystrong Woodland implications.No detailed studies were made of the Walter collection, but a seriesof 88 potsherds, said to have come from this site, was presented tothe Museum by Mr. Walter's sons. Though these are unaccompaniedby provenience records and so are of no stratigraphic value, it is inter-esting to note that two distinct wares occur, and that these parallelclosely the two stratigraphically separate sherd types noted in ourexcavations. Thus, 29 of the sherds presented to us have a fine com-pact dark-gray to black paste, sparingly tempered with roundedgravel particles up to 1 mm. in diameter, and with a hardness between4 and 6. In thickness, the sherds vary from 4 to 8 mm., but seldomexceed 5 or 6; surfaces are uneven and poorly smoothed interiorly,with cord impressions that may be crisscrossed and partly subduedby rubbing on the exterior. Eims vary from unthickened, vertical,recurved, or flaring forms to channeled and collar types bearing cordimpressions or parallel incised horizontal lines immediately belowthe lip. In every respect, these sherds recall the Upper Republicanpottery of southern Nebraska, and I think they can be attributedwithout hesitancy to that complex.The remaining 59 sherds, though somewhat variable in appear-ance, are distinguished from the preceding series by their relativesoftness (under 4) , their coarse granular paste, abundant inclusions ofangular crushed calcite and/or moderately coarse to very coarse(up to 3 or 4 mm. diameter) rounded gravel, greater thickness (6 to10 mm., usually over 7 mm.), and rather distinctive surface finish.The surface treatment, except for a few plain sherds, involves tex-turing by means of fabrics or fabric-wrapped implements. Charac-teristically (30 sherds), this texturing consists of boldly impressedparallel cord marks, usually coarser and deeper than those on UpperRepublican pottery, and less often rubbed or polished over. Oneof these sherds, probably from just below the lip of the vessel, showspart of a circular boss punched inward from the exterior surface.A variant of this treatment, seen on seven sherds, differs only in theuse of finer tightly twisted cords ; one rimsherd shows faint horizontalcord impressions extending downward inside the rim to a distanceof about 15 mm. below the lip. Another group of sherds, 8 in number,has been finished in different fashion. Low ridges or ribs run verti-cally down the vessel, at intervals of about 2 to 3 mm., betweenwhich are faint regularly spaced transverse ridges, such as mightresult from pressing a string of beads into the plastic clay. I amconvinced that these imprints were not made with a twisted or knottedcord. The impression of tightly woven twined or of coiled basketry Wbdbl] an rNTRODUCTION TO KANSAS ARCHEOLOGY 421in plasticene most closely simulates the markings on these sherds.Unhappily, the sherds are rather small, and so the overall variationscannot be determined. Since the "ribs" run downward from the lip,the warp of the inferred basket was at right angles to the pot rim,as would be the case with a twined basket. On the other hand, itis possible that a large fragment or perhaps even a specially shapedcoiled basketry form was used paddle fashion in texturing the sur-face. If further findings confirm this suggestion, the regularity andevenness of the "ribbed" or "beaded" surface would imply a ratherhigh quality of craftsmanship in basket making.Concerning the remainder of this group of sherds, little can besaid since they have been badly weathered or otherwise defaced. Twoor three bear imprints that may have been done with a cord-wrappedstick or dowel, and another shows a single broad incised line near abroken edge. Eight others conform in all respects save paste to thelarger group; the paste is fine and dark, with few or no visible in-clusions, and suggests that in the Upper Republican sherds describedabove. For the bulk of these heavy slierds, however, Woodland re-lationships are strongly indicated. There is, so far as our observationsgo, no good evidence of a crossing or mixing of the two pottery types,and the differences between them are generally clear cut and unmis-takable.From the surface of the terrace and from Mr. Walter's back dirtwere collected a number of items that may be noted in passing. Alight-gray sandstone slab 3 to 4 cm. thick had been fashioned into arectangular metate measuring 26 by 30 cm.; the upper surface was =imoothed, with the most pronounced wear in an elliptical area nearthe center, presumably by use of a muller in a circular or rotaryfashion rather than with a to-and-fro motion. A 19-cm. section ofweathered mammal rib was cut off square at one end, broken at theother; there were no other modifications from which its purposemight be guessed. Other specimens included a stemmed iron arrow-point measuring 15 by 35 mm., three blue and white glass beads,scrapers of basalt and jasper, chipped knife and point fragments, abroken muller, and a fragment of another mealing slab. Other thanthe metal and glass, patently of Caucasian origin, none of theseobjects is sufficiently distinctive to be attributable to any specificcultural horizon. Their presence on this station does point up thefact, however, that much more than pottery is involved in the smallstratified sites of western Kansas, and raises the hope that carefuldissection of an undisturbed terrace deposit may some day give us amuch clearer picture of the successive cultural complexes concerned.In view of the meagerness and imprecise nature of the record fromthe Walter site, it is obvious that nothing definitive can be said re- 422 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174garding cultural associations.^" That two distinct pottery wareswere present seems clear, and from our limited excavations it wouldappear probable that one?the calcite-tempered ware?preceded theother. Probably associated with this older calcite-tempered potteryare medium to large stemmed or corner-notched projectile points, ofwhich we unearthed two and of which the Walter collection includesseveral. Otherwise, there are only retouched spalls, a few scrapers,and one small piece of a rubbing or lapstone that can be certainlyascribed to this earliest period of occupation. From the thickness ofthe deposit a rather prolonged occupation, or more likely a series ofintermittent (seasonal?) reoccupations, is probable, and the greatabundance of broken animal bones strongly suggests a primary re-liance on hunting.The second, and inferentially more recent, ware seems to have beenassociated with more skillfully fashioned knives and scrapers. TheWalter collection has a number of small triangular notched projectilepoints that may also have come from this horizon, but our work gaveno direct proof of such an association. It would be interesting toknow just where the mealing slabs occurred, and whether they wereused for grinding corn or for pulverizing wild seeds, berries, orfruits.A still more recent use of the terrace is indicated by the ironarrowpoints and glass beads. These, obAdously, date from the lasttwo or three centuries when the western Plains constituted the hunt-ing territory of a score or more of tribes. Similar objects, to whichno time or tribal label can be assigned, may be found on almost anysuitable camping or watering place throughout the Plains area.The cultural associations suggested above, and perhaps also thetemporal distinctions, must be viewed with reserve; by themselves,the data from the Walter site are tantalizingly inadequate. For-tunately, there are other sites in the area that corroborate the aboveinterpretations, and one of them at least?the Pottorff site?providesclear-cut evidence of sequence involving the same two prehistoriccomplexes implied at the Walter site.SITES ON LADDER CREEK, SCOTT COUNTYDuring June and early July of 1939, 4 weeks were devoted to anexamination of antiquities in and near Scott County State Park, 12miles due north of Scott City (fig. 75). Scott County lies on the dryflat High Plains between Arkansas and Smoky Hill Rivers, 50 miles ^ Very few faunal remains were recovered ; they are insufiQeient to give reliable Indicationof possible stratigrapbic variations. Species represented, with the number of bones orfragments ascribed to each, include: bison (14), prairie dog (4), jackrabbit (2), antelope(1), white-tailed deer (1), and sliunk (1). The antelope and jackrabbit bones were foundat depths of 24 inches or more, presumably in the Woodland-like horizon (cf. occurrenceof these forms at Pottorff, p. 408) ; bison bones were distributed through all levels. Wedkl] an introduction TO KANSAS ARCHEOLOGY 423 east of the Colorado State line. Along its northern edge is a stripof rough hilly ground where the upland margin is being dissectedby ravines and short watercourses draining northward into theSmoky Hill Eiver, Ladder (or Beaver?) Creek rises in easternColorado and flows in a shallow valley east by south to a point about6 miles northwest of Scott City, whence it turns northward to emptyinto the Smoky Hill after another 12 or 15 miles. Though commonlyrepresented on maps as an intermittent stream, live water is to befound along most of the lower 20 miles or more of its course, andthere is or once was much excellent pasturage. An artificial lakehas been created behind a concrete dam thrown across the creek halfa mile south of the county line, and 1,200 acres of the adjacent valleyare now set aside as a State park. Near its upper end, on the eastside of the creek, is a 100-acre preserve containing a small herd ofbison (pi. 65, ?).Ladder Creek, as stated, runs nearly due north through the park,in a pleasant little canyon varying from less than half a mile tonearly a mile wide. On either side are high irregular bluffs ofTertiary materials, overlying a Cretaceous chalk formation intowhich the immediate creek valley has been cut. At the contact be-tween the porous Tertiary deposits and the underlying imperviouschalk, ground water comes into the valley in considerable volumethrough a series of excellent springs. These are perennial, and theircomxbined flow gives rise to the small but never-failing creek whichmeanders placidly down the valley between wide flat or gently slopingflood-free terraces. The banks are overhung with tall grass andbushes, relieved here and there by an occasional clump of cotton-wood or willow. Buffalo and other short grasses dominate most ofthe uncultivated valley floor as well as the flat uplands beyond thebluff line. Ravines and short, deep box canyons, some of considerablesize and ruggedness, open off the main valley, and bear stragglingstands of timber interspersed with chokecherry, wild currant, buffalo-berry, and similar small growth. In early days, the grassy valleyand its broken margins gave food and shelter to a variety of animalforms, including bison, deer, antelope, wildcat, wolves, an occasionalmomitain lion or bear, and smaller species; and it can be inferredthat prairie chicken, quail, and perhaps waterfowl were also avail-able.Today, the larger native mammals, and of course the Indians, aregone. The jackrabbit and coyote remain; there is a small prairiedog town near the south end of the park ; and owls and turkey vulturesnest in the crevices of the soft rocky ledges. The place, however,had not lost its charm altogether in 1939. The all but treeless valley,opening suddenly and unexpectedly at the feet of the traveler from 424 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174the east, south, or west, "with its yucca- and cactus-studded slopes,rocky bhiffs, narrow steep-sided ridges, flat-topped buttes, isolatedcones and mounds, immediately impressed one as a parcel of theSouthwest; and its appeal was heightened by the sharp contrastbetween its green meadows and cool waters, on the one hand, and thedry sun-drenched seemingly endless surrounding plains. With itstraditions of Pueblo Indians finding here a temporary escape fromSpanish oppression on the Rio Grande, and of Dull Knife's embattledCheyennes on their way northward to a nonexistent haven in thesandliills of Nebraska, the peaceful little canyon seemed like Indiancountry still. "el CUARTELEJO" (14SC1)Worked flints, occasional sherds, and other traces of aboriginalactivity may be found at various points in, above, and below thepresent park area. The principal site?a ruined 7-room stone struc-ture lying about midway between the north and south park bound-aries?has long been known to archeologists through the researchesin 1898 of S. W. WiUiston and H. T. Martin for the University ofKansas. It is situated on a slight knoll, about 75 yards west of theformer creek channel and perhaps 250 or 300 yards from the west edgeof the valley. The creek here follows a wide arc curving to theeast, within which is a terrace some 500 yards long by 250 yardswide. There is a fine spring near the upper end of the terrace about300 yards southwest of the ruin, from which may be traced the courseof an old irrigation ditch that once watered the flat. An abandonedrailroad grade, a scenic drive, and the impounded lake waters havepartially obscured details of the terrain as it was in the days ofIndian occupation.Fortunately, there are extant several published accounts of theuncovering of the ruin (Williston, 1899; WiUiston and Martin, 1900;Martin, 1909). Briefly, it appears from these that walls 18 to 24inches thick and about 2i/^ feet high, made of coarse-textured chalkyboulders from the neighboring bluff, formed an east-west rectanglemeasuring 53 by 35 feet (fig. 78). This contained 3 rooms on thesouth and 4 on the north, no 2 of the same size (Martin, 1909, p. 14).There was no evidence of doors or other openings. Walls and floorswere mud plastered. Fireplaces were rectangular, lined with thinstone slabs set on edge; one room contained two such hearths, andone that was unplastered also had no fireplace. Into the comer ofone room had been built a small oven; another contained a typicalpueblo grinding trough of clay, on a platform about 6 inches high.Most of the rooms had a pair of small postholes 12 to 18 inches apart,usually in one corner or near a wall, possibly indicating the positionof ladders for entry and exit through a roof hatchway. Burnt Wedbl] an introduction TO KANSAS ARCHEOLOGY 425 FEET -^^ N Figure 78?Plan pf Scott County pueblo ruin (after Martin, 1909). A, Oven in room 5;B, grinding basin in room 1; stippling, raised areas; solid circles, postholes; small rec-tangles, slab-lined fireplaces. 426 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174 adobe, charcoal, masses of carbonized corn, and the scorched ap-pearance of the artifacts all indicate destruction of tlie building byfire. By the side of the building were "two large hollowed out places,which had probably been used for the puddling and mixing of theadobe employed in the construction ..." A hundred yards to thesouth was evidence of "several other smaller buildings, all of whichmust have been of adobe alone, since no rock remains"; these, too,seemed to have been burned out. About 25 yards to the north therewere "three or four small structures, each separated a small distancein an east and west line parallel with the main building. These struc-tures were apparently circular in outline, and were perhaps teepees."The investigators concluded that the ruin had been built by Taos andPicuries fugitives from the upper Rio Grande in the late I7th orearly 18th century, and that it represented the "fortified" outpostthereafter known to the Spanish as El Cuartelejo. Concerning thelatter point, we shall have more to say in another place ; but there canbe little quarrel with the estimated dating and none at all with theascription of the ruin to a puebloan group. It is clear, too, that theoccupants of the structure were in contact with, if not actually visitedby, white men. A granite shaft at the southwest corner of the ruinnow rises over the obscure traces of this, the northeasternmost knownconstruction of the Pueblo Indians.Our 1940 investigations were directed toward the finding of datablepuebloan remains in association with a definable Plains culture hori-zon, in hopes of establishing thereby a time datum that could beextended eastward and northward to other recognized archeologicalcomplexes. We worked mainly in the previously untested refusedeposits and certain other nearby features not mentioned by Willistonand Martin. Beyond relocating the corners of the stone-walled ruinas bases for our surveys, we did not examine the main structure. Ef-forts to locate and reexamine the lesser buildings reported north andsouth of it were unsuccessful and I suspect all traces of them wereobliterated while the terrace was being farmed before it was set asideas park land. TEST NO. 1Small exploratory pits were put down at a number of points on themain terrace, as well as on several nearby flats upstream and acrossthe creek. In many of these, the topmost few inches yielded charcoal,flints, and other materials in meager quantity. About 70 yards duesouth of the monument, not far from the vanislied outbuildings men-tioned by Williston and Martin, the soil was mixed with bones, sherds,flints, and charcoal to a depth of 16 inches or more, well below any levelever reached here in plowing. A 50-foot north-south base line, dividedinto 5-foot units, was accordingly laid out, and systematic removalof all mixed soil begun (fig. 79) . Wedbl] an introduction TO KANSAS ARCPIEOLOGY 427 v////// El OuartelejoMonument % 50 ,25 Test No. 2 tHSSS!A?3'.'S,w;-S:.j Pueblo Ruin ? 25 Test No 3 Test No. I Figure 79.?Map of excavations in vicinity of Scott County pueblo ruin and monument,14SC1. 428 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174The first feature disclosed was an elliptical basin 8 feet wide by 16feet long, with the long axis running north-south. At the center thefloor lay 24 inches below ground surface, sloping evenly upward on allsides to a depth at the walls of 16 inches. The fill included much ani-mal bone and charcoal, with lesser quantities of flint chips, potsherds,projectile points, scrapers, and the like. This material ended abruptlyat the basin walls, wliich were easily traced on all sides. Five feet fromthe north end, midway between the east and west walls, was what ap-peared to be a posthole 7 inches in diameter by 12 inches deep. A simi-lar but slightly smaller hole lay a few inches from the south end, alsoon the midline. There was no indication of a fireplace, unless a fire-blackened area 13 by 16 inches across near the northwest wall be soconsidered. The floor as a whole had not been hardened or blackenedby fire, but the overlying mixed fill everywhere ended abruptly at thislevel and gave way to clean undisturbed subsoil except where the twopostholes (?) reached greater depths. Among the artifacts may benoted the finding of an iron awl on the floor near the center of thebasin. The purpose of the basin is not clear, but the postholes, ifsuch they were, suggest that a structure may once have stood on thespot. The absence of a firepit argues against use as a domicile. It ispossible, though I think rather unlikely, that the basin represents aborrow pit or that it was dug for the disposal of refuse. My guessis that some sort of temporary habitation or perhaps a summer shelteronce stood here, the refuse having been thrown into the depressionafter abandonment of its original purpose (pi. 66, a).Outside and near the south end of the basin, 3 feet from the eastwall, was a circular bell-shaped pit 39 inches across the top, 53 inchesacross the bottom, and 39 inches deep. The walls and slightly dishedfloor were well defined, though not burned or lined, and the fill wascomparatively soft and ashy, with some charcoal. Artifacts includedsherds, scrapers, a ball of yellow ocherous clay, a tubular bone bead,two scapula hoes, a fragment of clay pipe bowl, a bone awl, threeslender unidentified bone objects, a tubular bone bead, and a shell diskbead. In form, size, fiill, and all other respects, the pit suggests a cachesecondarily used for deposition of rubbish. It should be added thatcharred and rotted stick fragments lay over the top and about themouth of the pit. These may represent remains of a cover, but theywere within the zone of possible disturbance by the plow and couldhave been intrusive.A few yards north of the basin and nearby cache pit was anotherarea of refuse-filled soil. This was irregularly crescentic in outline,with an east-west extent of nearly 45 feet, but a width of only 3 to 8feet (pi. 66, h). The fill was heavily mixed with ashes, char-coal, broken animal bones, stones, and other rubbish, so concentrated Wbdbl] an introduction TO KANSAS ARCHEOLOGY 429in some spots as to suggest the dumping of floor sweepings, fireplaceashes, and other materials from an all-out house cleaning. Depth ofthe fill varied from 14 to 20 inches. There was no evidence whateverof construction in connection with this deposit, and its irregularityof outline does not suggest a purposeful excavation to me. Two alter-natives present themselves: one, that there was here originally ashallow swale or minor drainage chamiel that the Indians filled in withtheir refuse ; the other, that earth had been dug from the strip by theIndians for some unknown purpose and the resulting depression sub-sequently utilized for a trash dump. Unfortunately, we obtained noevidence that would seem to balance the scale in favor of either ofthese possibilities, and the question remains undecided.TEST NO. 2The principal midden area in the vicinity of the pueblo ruin ap-parently lies immediately north of the structure. There were noobvious surface indications in this spot, and the presence of subsurfacedeposits was disclosed only after a series of small exploratory pits hadbeen opened. We dug a 60-foot test trench north from the base of thepueblo, crossing this with a 45-foot east-west trench about 35 feet fromthe wall. These cuts were subsequently widened in the most promisingportions, so that a total area of nearly 900 square feet was worked out.The midden was by no means exhausted, but there is no way of judg-ing what proportion of the deposit remains unworked (fig. 79).The composition of the midden varied somewhat from point to point,but the profile along our north-south baseline seems to have been fairlyrepresentative (pi. 65, 5) . The pueblo wall itself, here represented bya single course of stones and covered by light-gray windblown soil,rested on clean grayish to brownish sandy soil containing small cal-careous pebbles. From the wall northward the present ground surfacerose a few inches for the first 5 or 6 feet, then gradually descended14 to 18 inches in the next 40 feet, and finally rose slightly again. Theuppermost 10 inches throughout most of this lateral extent consistedof fine gray material with some admixture of charcoal, animal bone,stones, and rubbish. Within this horizon, at a depth of 4 or 5 inches,traces of a cleavage surface were noted, the soil below being rathermore compact than that above. The cleavage surface was accentuatedby short thin lenses of fine sandy or silty material, nowhere exceedingone-half or three-fourths of an inch in thickness by 6 or 8 inches long.These occurred at intervals of 16 to 18 inches, or in what appeared tobe approximate multiples of that interval. Here and there at thesame level were thin seams of carbonized vegetal matter. I suspectthat the cleavage plane represents plowsole, the charred material per-haps being burned grass or other surface growth turned under in 430 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174fanning. The regularly spaced silt lenses suggest wind- or water-laidmaterial deposited on a surface of alternating ridges and furrows suchas a cornfield might ojffer, the silted furrows having been finally cov-ered over and leveled off by natural or mechanical agencies to givethe present comparatively smooth surface.From the wall to a distance of about 12 feet northward, this silt-lensed formation had been obscured by a disorderly bed of orange-brown burnt earth, charcoal, and other debris, with few or no artifacts.This bed, 3 to 8 inches thick, was traced west at least 10 feet and anequal or greater distance eastward from our cut ; nearly everywhere,it seemed thickest near the wall, and contrasted strongly with theunderlying deposits. I am inclined to think that this material wasnot originally deposited by the Indians as midden, but is part of thefill removed from the rooms of the burned-out pueblo during Martin'sexcavations.Below the level reached by the plow, the soil contained an increas-ing proportion of cultural admixture, the greatest concentration every-where occurring between 10 and 20 inches underground. This zone,of course, was somewhat variable in content and thickness, but gen-erally yielded an abundance of broken animal bones, fire-blackenedand often heat-fractured stones, bits of burnt earth, carbonaceoussoil, flint chips, sherds, and a variety of complete and incompleteartifacts. With a few possible exceptions scattered at the north endof our diggings, there were no post molds,*" though occasional shallowpits and burned hearthlike spots were noted. Below 20 or 22 inches,admixture thinned out rapidly ; at no point, except where pits had beendug by the Indians, was cultural material found below 28 or 30 inches.The underlying sterile soil was generally brown or gray brown incolor, sandy textured, with small rounded calcareous pebbles scatteredthrough it. The undulating and somewhat pitted surface of thissubsoil rose gradually toward the pueblo, more or less paralleling therise of the present ground line. Five feet from the pueblo wall, itcurved sharply upward, so that the refuse strata pinched out in thewindblown gray dust piled to a depth of about a foot against and overthe stone foundation. Thus, stripping of the midden to unmixedsubsoil would leave the pueblo wall foundations standing on a slightknoll of similarly unmixed material 1 to 2 feet high.The special features encountered during the midden excavationsmay be briefly noted ; all seem to have rested on, or been dug from,approximately the 20-inch level. Beginning 15 feet from the wall,and running almost continuously northward for about 12 feet, was ali/^-inch layer of very dark carbonaceous earth, containing corn, *o Two or three probable post molds were partly filled with broken longbones of bisonInserted upright, as If to wedge a post or standard. Wkdbl] an introduction TO KANSAS ARCHEOLOGY 431 cobs, squash rind, bones, stone, and other refuse. At the same level,nearer the pueblo, the soil was lighter in color, but likewise con-tained much trash, including a scrap of iron and one or two glassbeads. Beneath the north end of the black charred layer, whoseeast-west spread was not worked out but certainly exceeded 5 feet,a shallow basin 24 inches across had been scooped 10 inches into thesubsoil and filled with bones and stones. Ten feet north, and at thesame level, was a second charred area approximately 10 feet long by5 feet wide. At its east edge was a shallow circular basin 30 inchesacross by 6 or 7 inches deep, filled with irregular lumps of orange-brown and black burnt clay, charred sticks, and ashes. The wallsof this basin were baked hard to a thickness of 1 or 2 inches ; in andabout it, besides the materials just noted, were found carbonized corn,cobs, and bits of squash or gourd. The surface about the firepit, onwhich the burned materials were scattered, seemed rather well com-pacted, though there was no evidence of a covered structure or definitehabitational unit.Near the north end of the trench, and evidently dug originally fromthe 20-inch level, was an elliptical refuse-filled basin measuring 10feet by 7 feet 3 inches. Its long axis lay slightly east of north andwest of south, and the bottom was 35 inches below present groundsurface. The sides of the basin curved evenly upward in all direc-tions from the deepest point; they were covered by a 3- to 5-inchlayer of bone, charcoal, and other debris, in turn overlain by 2 to 3inches of ashes. Removal of this burnt fill disclosed the bottomsof two older pits, each approximately circular in plan with onepartly overlapping the other. Both appear to havi? averaged 42 to48 inches in diameter. Pit lA, undoubtedly the earlier, reached atotal depth of nearly 50 inches below the present gromid level, or30 inches below the plane into which the basin had been sunk; pitIB was about 7 inches shallower. Both contained numerous animalbones, stones, sherds, and carbonized material. Their original depthcan only be guessed at ; if the 20-inch level from which the overlyingbasin was excavated represents the surface at time of habitation, thetwo earlier pits would have been about 30 (pit lA) and 23 (pit IB)inches deep. It is possible that the basin itself was not purposefullydug as such, but instead resulted from collapse of the upper part ofthe pit walls, though this view is admittedly debatable. At the north-east edge of the basin, and also at the 20-inch level, was a flat-bottomed depression measuring 35 by 24 inches across and 3 inchesdeep. Filled with very dark burnt soil, charcoal, and ashes, thisprobably represents a fireplace, but its exact relation to the adjacentrefuse basin and the two earlier (cache?) pits is not clear.484172?59?29 432 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174The burned areas, hearth complexes, and other concentrations ofrefuse at or very near the 18- to 20-inch level throughout test No, 2presumably mean that this and the immediately overlying fill repre-sent the period of most intensive occupancy of the site. They appearto have been outdoor cooking areas, perhaps used by the inhabitantsof the pueblo during the warm summer months. There is no evidencethat they antedate construction of the pueblo ; on the contrary, theyseem to have been placed in a slightly depressed area that was grad-ually filled and its surface raised by prolonged accumulation of refuse.A considerable proportion of the overlying fill is soil and sand, prob-ably of aeolian origin, but this material is intermixed everywhere withhousehold rubbish, and there is no indication of periods of abandon-ment. The 20 inches or more of refuse-mixed soil certainly indicatesan occupation of some duration, but there is no way of judging itsleng-th in terms of years. The deposit is evidently wholly, or almostwholly, post-Caucasian, since iron and glass beads occurred verysparingly but inclusively to within a few inches above undisturbedsubsoil.Curiously enough, material distinctively puebloan in origin wasexceedingly scarce; other than an occasional incised clay-pipe frag-ment, there is almost nothing in our collections that would not be ex-pectable in an already defined nonpuebloan Plains culture horizon.Yet, in nearly all respects, the artifacts we recovered from the middenclosely parallel those found by Williston and Martin (Martin, 1909)in the ruin, and there is scant doubt in my mind that the midden intest No. 2 is to be attributed to the inhabitants of the nearby pueblo.We shall return to this problem elsewhere after the artifacts have beenfully described. TEST NO. 3Twenty yards due south of the "El Cuartelejo" monument a thirdmidden area was partially worked out. As with those described above,surface indications were virtually nonexistent, though the refuse oc-curred in a rather well-defined, possibly dug, vertical-walled basinsomewhat like that in test No. 1. The area involved was irregularlyelliptical in outline, with the long axis running roughly east-west.The maximum width was 12 feet, and the length of the area weopened was about 25 feet ; test holes indicate that refuse-mixed earthcontinued for some little distance eastward beyond the limit of ourdigging (fig. 79; pi. 66, c).The uppermost 10 inches of basin fill included much fine silty ma-terial and on exposure dried rapidly to an extraordinary hardness;artifacts and refuse were comparatively scarce, suggesting that mostof this zone had been blown, washed, or otherwise deposited since thetime of the Indians. Below the 10-inch level, the fill was very much Wedel] an introduction TO KANSAS ARCHEOLOGY 433 softer and contained much ash; in places, this could be removed bythe unaided hand alone. Between 12 and 18 inches depth over mostof the basin were heavy concentrations of ash mixed with charcoal,these tending to form lenses up to 6 inches thick. Here and there, inthis same stratum, were seams of burnt grass, bits of corncobs andstalks, and other fragmentary vegetal material. Near the south wall,at a depth of 17 inches, lay six bison vertebrae which had evidentlybeen deposited in articulation. Scattered through the underlying fill,and most plentiful in the 4 or 5 inches immediately above the basinfloor, were sherds, worked flints and chips, bone tools, animal bones,and stones. The floor itself, at a depth of 26 or 27 inches, was coveredwith a layer of fine black silt, apparently washed in or perhaps de-posited in ponded rain waters ; above this, in turn, was a thin stratumof carbonized material. Except at the east end, the walls of the basineverywhere were of light, unmixed, sandy soil markedly unlike thefill and in consequence easily definable.Careful scrutiny of the cleared floor revealed no evidence of postmolds, but near the center was a very slightly depressed subcirculararea 42 by 36 inches in diameter containing much burned earth andvegetal matter. A hearth is suggested, or possibly the floor of anearlier, i. e., prebasin, cache or cooking pit.Just outside the refuse basin, a foot from its west end, strippingaway of the 6 or 8 inches of plow-disturbed topsoil disclosed a nearlycircular dark mottled area just under 4 feet in diameter and con-trasting strongly with the surrounding light-gray soil. This provedto be a well-defined pit (pi. 66, c), whose walls curved outwardslightly at a depth of about 7 inches and then converged graduallyto a faintly concave floor 42 by 38 inches across. Maximum depthfrom present ground surface to floor was 34 inches. In all these di-mensions and in its general form, the pit fell within the range of vari-ation noted for prehistoric caches of the Central Plains. The fill,however, suggested neither primary storage nor secondary utilizationfor dumping of household rubbish. Below the plow-disturbed top-soil there was first a 6-inch deposit of dark soil mixed with charcoal,and under this a layer of irregular burnt limestone fragments. Thencame an 8-inch stratum of fire-reddened earth, underlain by a 10-inch thick bed of charred sticks and twigs. On the pit floor werescattered a few more stones intermixed with charcoal and ashes. Thewalls at a depth of 22 inches, corresponding approximately to theburned earth stratum within, had been baked hard and discolored toa thickness of 2 to 4 inches. All this unquestionably means that veryhot or prolonged fires had been maintained in the pit; and, in thelight of certain other fuids to be described shortly, it seems reason-able to infer that this and similar pits (fig. 80) were dug for a 434 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174 Jc?1 PIT I, TEST 3Disturbed by plow, etc. Dark soil, some charcoalBurned stones @^^Fire-reddened soilCharred sticks, twigs,etc.Stones on floor PIT 3 Ashy fill,sonie charcoal astonesCharcoal 8 stones on floor Wbdbl] an introduction to KANSAS ARCHEOLOGY 435 specific purpose other than food storage and do not represent burned-out caches. Artifacts, we may add, were absent except for a fewsmall sherds and scraps of animal bone.BOASTING PIT (PIT 3)On the park ballgroimd, 110 yards north-northwest of the monu-ment, a small area of soil discoloration was noted on the groundsurface by a member of our party. The topsoil was carefully strippedaway to a depth of 9 inches and to a distance of 2 or 3 feet inall directions. At this level the discoloration showed up as burnedearth and carbonized material surrounded by a well-defined ring oforange-yellow fire-hardened soil averaging 2 inches in width and 19inches in inside diameter. Outside the ring was dark-gray unmixedloam. A vertical section bisecting the ring showed that it was thestrongly constricted neck of a jug-shaped pit, the walls of which flaredoutward and downward to a maximum diameter of 44 inches, retain-ing to the bottom their orange-yellow color and bricklike hardness.The fire-blackened floor was 28 inches underground at the center,rising slightly to 25 inches depth at the perimeter. A thin bed ofcharcoal and ashes covered the floor, and on this was an almost con-tinuous layer of irregular, heavily burned limestone fragments from2 or 3 to 8 or 10 inches in maximum diameter. Above the stones, thefill was light and ashy, and contained a heavy admixture of burnedsticks, twigs, and other organic matter. A single flint scraper wasthe only artifact found in the fill (fig. 80, hottom^ and pi. 67).The virtual absence of artifacts in the pit, together with theheavily burned walls, blackened floor, high ash and charcoal con-tent, and calcined rocks, all argue against its use for the concealmentof foodstuffs or other household possessions. Prolonged and in-tensely hot fires are certainly indicated, and these can hardly havebeen other than intentional. I think there can be little doubt thatthe structure represents a roasting oven used in the processing ofcertain food materials. Noteworthy in this connection was the ab-sence of any evidence whatever of a draft hole.OTHEB EXCAVATIONSNumerous small exploratory pits were sunk by members of our partyat a number of habitable spots upstream and downstream from theterrace on which the principal ruin lies. Most of these disclosed tracesof charcoal, flint chips, and other materials in the topsoil, but usuallynot in sufficient profusion to invite extended excavation. The mostpromising were on a terrace east of the creek, 4 to 5 hundredyards east of Big Spring, immediately west of the "Buffalo Park,"and about 1,000 yards due south of the monument. The abandoned 436 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174 railroad grade which traverses the entire north-south extent of thepark, crosses the south part of this terrace in a cut.As elsewhere in the park, surface indications of aboriginal activitywere exceedingly scanty. Near the south-central part of the terrace,however, between the railroad grade and the creek, there was a pitwhich had been partially dug out not long before our arrival. More-over many years ago another pit closely resembling the roasting ovenopened by us on the ball ground, had been excavated by Mr. A. T.Hill of the Nebraska Historical Society in the west face of the railroadcut. All this raised hopes on our part that careful search might dis-close the presence of habitational units, puebloan or otherwise?hopeswhich, however, were not fulfilled.The newly opened pit mentioned above had been dug to a depth ofabout 2 feet and at no point had either the floor or the wall beenreached. We completed its excavation, delineating a circular cavity48 inches across the mouth, 60 inches across the bottom, and 41 inchesdeep. The walls had been burned to a bright red color, in places asmuch as 3 inches thick. The bottom was flat; on and immediatelyabove it were a number of burned limestone fragments, broken musselshells, and numerous large bones of bison. A 5-inch layer of charredsticks and twigs extended entirely across the pit, 3 feet below theground surface. The exact character of the fill above the ash couldnot be ascertained, but it seems to have contained many animal bonesas well as charcoal and ash. Near the top, to judge from the narrowring of fill left for our observation, there were some thin laminae ofsilt, perhaps deposited in standing waters before the cavity had beenentirely filled in. This pit, I believe, represents another roastingoven, used as a receptacle for refuse after its abandonment for cooking.Thirty-five feet north of this oven was found a shallow ellipticalbasin measuring 54 by 30 inches, and 14 inches deep. From this weretaken broken bones, flint chips, charcoal, and burnt earth. At thenorth side of the basin was a cluster of limestone rocks 14 inches indiameter and 8 inches below the present ground surface. A singlesmooth grit-tempered sherd came from the rock cluster.West of the drive which crosses the terrace, and about 30 yardssouthwest of the oven, was another cluster of burned stones. Theselay 12 inches underground; the overlying fill was somewhat darkerin color than the surrounding soil, and contained a few scraps of boneand a quartzite knife. The stones covered an area 24 by 39 inches,slightly depressed at the center ; mixed soil extended but a few inchesbeyond in all directions, including an area 35 by 45 inches.About 30 yards northeast of the oven burnt earth and charcoal weredetected at a depth of 6 inches. At 7i/^ inches this mixture gradedinto a mass of charred twigs filling a basin 22 inches in diameter and Wbdbl] an introduction TO KANSAS ARCHEOLOGY 4375 inches deep. The basin walls were burned a rusty brown color toa thickness of about an inch. Two stones lay in the basin, near thesouth wall. Trenches run out from the charcoal basin disclosed notthe slightest indication of a habitation unit, though burrowing rodentshad transported bits of charcoal 5 and 6 feet in all directions and todepths of 2 and 3 feet at several places (pi. 66, d) . The pit investigated by Mr. Hill *^ lay at the edge of the terrace ; one side had been removed by the railroad cut, so that a full crosssection of the structure was exposed. In shape and dimensions itclosely resembled the roasting pit opened by us on the ball ground,with a circular floor, walls incurving at the top, and a narrow mouth.The floor diameter was 42 inches, the depth 23 inches, and top diameterless than 22 inches. The walls were fire reddened to a thickness ofseveral inches ; and large chunks of this baked earth, evidently brokenor caved off from the pit mouth, were found in the fill below. Hillsuggests that the original opening was perhaps nearer 15 or 18 inchesacross. To one side of the opening, lying just beneath the sod line,was a large stone slab which may have been used as a covering forthe pit. Leading out toward the terrace front from one side of thepit, at floor level, were traces of what Hill believes was a draft hole8 or 10 inches in diameter. In this particular, the pit seems to beunique among those opened from time to time in this locality. Severalinches of fine wood ash and some charcoal covered the floor ; no arti-facts are mentioned.Despite our inability to locate definable lodge sites, there can be noquestion as to the utilization of this terrace by Indians. In all prob-ability the stone- and/or charcoal-filled basins represent fireplaces,used from the present ground surface or from a dust-covered levelonly a few inches lower. The pits dug by us and by Hill can hardlyhave been other than roasting ovens, similar to those near the puebloruin. Few artifacts were found ; they were mostly of stone and notparticularly diagnostic. Recognizable puebloan objects were notpresent. I see no reason for attributing the remains we found on thisterrace to Indians culturally other than those who left the refusedeposits and other traces about the pueblo ruin, or to earlier or laterpeoples. IRRIGATION DITCHESBrief mention has already been made above of the obscure tracesof an old irrigation ditch that once led spring waters onto the terraceoccupied by the pueblo. The springs lie a few yards north of the oldSteele home, residence of the owner of the property at the time ofMartin's excavations, and just below the edge of the present roadway. ? Letter of July 26, 1945, and photographs KSc 1-7, 13, and 16 of the Nebraska StateHistorical Society's archeology files. 438 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174A short stretch of the canal, perhaps 2 feet deep by 4 or 5 feet wide,is still used to water a vegetable and flower garden immediately tothe south. Most of its former course, however, has been more or lesscompletely filled in, and is further obscured by the old railroad gradeand the present scenic drive. It is faintly discernible just east of therailroad embankment as it curves around a shoulder of the hill torun northward. Its northern terminus we were not able to ascertain,nor do Williston and Martin give any clue to the original length. Pre-sumably, it could have supplied water for several acres of corn in theimmediate vicinity of the pueblo.About half a mile to the south there is a second group of springsknown today as Big Springs. From these we were able to trace aditch, perhaps 5 feet wide by 4 to 6 inches deep, that ran northwardaround the head of a short draw and then northeastward onto a flatjust south of the park superintendent's residence. The draw hasapparently been widened or lengthened slightly since the ditch wasin use, as there is a break at this point. The end of the trace appearsto be some 250 or 300 yards from the spring. This is apparentlythe ditch alluded to by Williston (Williston and Martin, 1900, p. 125)as follows : About a half mile above the site of the present ruins, the tertiary underflowcomes to the surface along the side of a hill in such perpetual abundance that itis utilized in the irrigation of a considerable tract of land.We made no attempt to cross-section either of these two canals dur-ing the 1940 work, and therefore can throw no new light on theirstructure or possible antiquity. Local informants assured us that thefollowing statement by Martin (Williston and Martin, 1900, p. 130)applies to both : Mr. H. H. Hathaway informs me that the earliest settlers here utilized whatwere undoubtedly the remains of an old irrigating ditch in digging their ownditches in the vicinity of the present residence of Mr. Steele, and which ditcheshe now uses in the irrigation of his garden.LATEB INVESTIGATIONSDuring the winter of 1944-45, purposeful lowering of the lake levellaid bare considerable areas of terrace that had been inaccessible tous in 1939. Mr. Hill made several visits to the spot at this time, andinformed me *" that some distance north of the "El Cuartelejo" monu-ment in the vicinity of the present bath house there is "evidence of avery large camp ground, and several fireplaces show." Pottery andother surface remains appear to be identical with those excavated byus about the pueblo ruin. A notable exception is a painted sherdretrieved from the mud near a cache pit, and since identified by Mera *^ *' Letter of July 23, 1945." Letter to Wedel, August 9, 1945. Wbdbl] an introduction to KANSAS ARCHEOLOGY 439 as Tewa Polychrome. Nearby, a "bumed out cache" was opened;it was 14 inches across the top, 36 inches deep, and 36 inches across thebottom. The walls from mouth to floor had been burned red to a depthof 3 inches or more, and, as with the roasting pits described above,they met the floor in a curve rather than at an angle. Much ash andcharcoal, but apparently no artifacts, were present. I have no doubtfrom the description given by Hill that this is another roasting pit,and that the remains here can be safely ascribed to the same peopleas those who dwelt about the pueblo. It is to be regretted that cir-cumstances apparently precluded a thorough-gomg survey and ex-cavation of the site or sites temporarily exposed by the lowered lakewaters. FOOD REMAINSVegetal materials.?Charred cob fragments and kernels of cornwere found in tests 2 and 3, but nowhere in any quantity. Cob frag-ments are 12 to 18 mm. in diameter ; most seem to have been 8-rowed,with a lesser number of 10-rowed pieces. Several include short sec-tions of the ear stalk or bits of husk about the base of the cob. Fromthe roasting pit on the ball ground, associated with husk fragmentsand with 13 mm. of the stalk still attached, came a small 8-rowednubbin about 30 mm. long, with the husk still covering 2 or 3 rowsof kernels. Another cob from the same pit is 65 by 17 mm. in sizeand had 8, or less probably 10, rows of kernels. The kernels generallyseem short and fat, measuring about 7.5 mm. long, 8 mm. wide, and5 mm. thick.In contrast to the scarcity of corn in the middens we tested, is thegreat amount reported in the pueblo ruin (Martin, 1909, p. 15). Ofthis, Williston (1899, p. Ill) previously observed that ? The most interesting fact, however, connected with the ruins is that the cornis not the modern white man's corn, but the primitive aboriginal corn. Frag-ments of the cobs, as I found them, were not larger than one's finger, and hadonly a few rows of kernels upon them. The kernels were larger than those ofpopcorn, smaller than our field corn, with rounded surfaces, each individualkernel but little compressed by its mates. This corn is so very distinctly differ-ent from any now grown in the state, that it at once proves the antiquity ofthe structure. The corn, if not grown by the aborigines, must have come fromseed furnished by them. This corn is found in considerable abundance, quartsor even pecks of it may be obtained with a little labor. . . . This corn hasalso been reported from other places in the vicinity. . . . The corn evidentlyformed the staple food-product of the inhabitants, and I have no doubt what-ever but that it had been grown in the immediate vicinity.I have the impression that samples of this charred corn are pre-served in the Museum of Natural History at the University of Kansas.Since it appears to differ in size of cob, row number, and perhaps inother particulars, from the older corn recovered by us in the KiceCounty sites, an examination by maize experts and plant geographers 440 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174may be in order. It would be interesting to know particularly whetherits closest relationships are with southwestern varieties or with thosedeveloped by the semihorticultural Indians of the eastern Plains.Fragments apparently of squash or gourd rind were found in thehearth in test 2; and Martin (1909, p. 18) mentions finding of squashseeds in one of the rooms in the pueblo. We found no beans or sun-flower seeds.A single fragmentary pit of wild plum was taken from our test 2.Turtle remains.?Of these, at least three, and possibly four, formswere represented. Most common were bones of the cooter or pondterrapin {Pseudemys). Leg bones and miscellaneous fragments ofthe conunon snapper {Ghelydra serpentina) were present, some ofthem from rather large individuals. There were a few carapace frag-ments of the box turtle or land terrapin {Terrapene). Less certainis the identification of several bones of the painted turtle ( Ghrysemys) . None of the turtle bones or carapace frgaments bore any tool marksor other evidence of modification for the making of artifacts. As afood item, turtles cannot have been very important, if used at all, be-cause their remains comprise but a very small part of the bone refuse.Bird hones.?These, too, were scarce, there being not more than 12identifiable specimens in our series. They include the following forms,of which the first three were presumably taken for their feathers ; theothers for their flesh and/or feathers : Ferruginous rough-legged hawk {Buteo regalis)Prairie falcon (Falco mexicamis)Great horned owl {Bul)o virginianus)Coot (FuUca americana)Teal (Anas sp.)Anatidae (Duck)Whooping crane {Orus americana)Mammalian remains.?For the most part, the animal bones encoun-tered were broken up and splintered so badly that their identificationwas impossible. A few were weathered as if they had lain on the sur-face for a long time before being covered, but the majority had asmoother fresher appearance. Fire blackening was noted in a fewinstances, and etching by rootlets or soil acids was quite common.None of the bones showed tool marks.Identifiable bones represent the following forms: bison (Bison) ^ 51; deer (Odocoileics), 32; dog (Canis), 17; prairie dog {Cynomys)^2 ; badger ( Taxidea taxus) , 2 ; and pocket gopher ( Geomys) , 1. Bisonbones, as might be expected, were most abundant, and this species,together with deer, evidently furnished the great bulk of the meatconsumed by the natives. The endless grasslands surrounding thevalley, and even more, the permanent springs and lush grass of thevalley itself, probably brought the herds within easy reach of the Wedel] an introduction TO KANSAS ARCHEOLOGY 441 village; and local traditions speak of some of the side canyons asformer bison traps, for which purpose they would have served admi-rably. Deer may have inhabited the brushy side canyons, as well asthe Smoky Hill valley not far to the north. The absence of anteloperemains seems strange, since the terrain looks like good antelope range ; but perhaps deer were easier to get, more palatable, and their hornsand hides of greater usefulness to the Indians.The dog remains were all those of small to medium-sized animals, nolarger than a coyote; there was no evidence of the large, powerfuldraft-type beasts indicated at the Kice County sites.The last three forms may or may not belong to the native food listall are common Plains species and the broad grassy flat on which thesite lies is just such a location as would attract rodents and burrowinganimals. There is still a sizable prairie-dog town a mile or two up thevalley, though no burrows were observed on the village flat. In anycase, these small forms evidently did not figure very heavily in thenative economy. POTTEBYOur collections from the site include 3,810 sherds, of which 3,382are body fragments and 428 are rim pieces; there are no whole orrestorable vessels, other than 1 or 2 miniature pots. Almost withoutexception, the sherds are small, and rarely is it possible to even guessat the vessel forms or sizes originally present. With the exceptionof four shell-tempered sherds of foreign origin, all of the materialis probably of local manufacture. Two wares are indicated.The majority ware, for which the name Scott Plain is suggested,is represented by 3,208 body sherds and 414 rimsherds, or 95 percentof the total. It is dark in color, and contains rounded quartz in-clusions ranging in diameter from 0.25 to 3.0 mm. and averagingbetween 0.5 and 1.0 mm.; the inclusions are usually moderatelyabundant, but are smaller than those in Upper Republican ware.Paste is fine and even in texture, with a straight granular fractureand some tendency to crumble on a freshly broken surface. Surfacehardness varies from 3 to 4.5, and occasionally reaches 5. Color,usually very dark gray or nearly black, is sometimes light buff orreddish brown. Occasional pieces show a light-colored core, withone or both surfaces dark to a depth of 0.5 to 1.0 mm. The darkestpieces often have a carbonaceous crust on the exterior surface, andwhen crushed have a sooty or "greasy" look and "feel." Surfacesare generally uneven, occasionally imperfectly smoothed, but neverpolished, slipped, or painted. On some pieces the aplastic protrudeson the interior to give a rough, gritty surface, but this is not char-acteristic. 442 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174Vessel shapes are uncertain,''* but there are pieces suggestive ofjars with recurved or flaring rims, constricted necks, and roundedshoulders, and none indicative of bowls, bottles, or other forms.Fragments of a low conoidal base from a squatty jar were found intest 2. Rimsherds are nearly always simple, direct and unthickened,with an uneven rounded or sharpish (never flattened) lip, and theyseem generally to have been outcurved or flaring with reference tothe constricted neck. Sherd thickness may be as much as 11 mm.,but the vast majority are less than 6 mm. thick.Decorative treatments of any sort, as well as appendages, arealmost absent. Thirty-one sherds, less than 1 percent of the seriesunder consideration, bear simple stamping made with a groovedpaddle. This includes several rimsherds in which the stampingextends up to the lip. Two rimsherds, out of a series of 414, havediagonal incisions on the lip, and one of these has in addition asmall lug on the outer upper edge of the rim flush with the lip. Twobody sherds each have two or three parallel incised lines that mayonce have formed part of a larger pattern.The minority ware includes 171 body and 13 rim sherds, just under5 percent of the total. These are distinguished from the precedingby their thinness (about 3 to 4 mm.), by the great abundance ofminute particles of mica, by their hardness (approaching 6), and bytheir high-pitched metallic ringing when dropped on a sheet metalsurface. Microscopic examination of crushed fragments shows ahigh mica content throughout, with a few siliceous particles that seemrather more angular than the quartz inclusions in the majority ware.E. P. Henderson, of the Division of Mineralogy, United States Na-tional Museum, to whom representative sherds of both wares weresubmitted, observes that the minority sherds differ very markedlyfrom the others, having been intensely burned, often to the point ofsintering. The paste is black, fine, sooty, and friable. Sherd surfacesare uneven and often somewhat granular, but are smooth to the touch.Most show the presence of mica on one or both surfaces. One sherdhas a suggestion of simple stamping; four or five show striations at ** This statement requires some modification. When A. T. Hill in the winter of 1944-45examined caches exposed by lowering of the water in Lake McBride (p. 438), he re-covered numerous sherds, including parts of at least two large jars. One of these, restored,Is shown in plate 68, b. It measures approximately 36 cm. in height ; diameter is Irregular,varying between 30 and 35 cm. The lip is undecorated. The piece can probably beclassed as Scott Plain, although it is considerably larger than the majority of these vesselspresumably were. Except for its greater size, this plainware jar seems to conform to thelargest vessel figured by Hill and Metcalf (1942, p. 181 and pi. 6, fig. 2) from 25CH1. aDismal River culture site in Chase County, Nebr.Restoration of the second piece has not been possible, but it appears to have been a Jarwith somewhat subconoidal body terminating in a small flat disk base about 6 cm. indiameter. The paste and finish seem alien to the local complex, suggesting a trade piecefrom the Great Bend Aspect?or perhaps a local product made in direct imitation of GreatBend pottery. WbdelJ AN INTRODUCTION TO KANSAS ARCHEOLOGY 443jEirst thought to be from grass-brushing but in one case a plasticenecast looks like fine tightly twisted cord impressions. Otherwise, thesherds are plain, and the rims, except for their thinness, do not differfrom those of the majority ware. There are no appendages and thevessel shapes are unknow^n. I am not certain whether this ware can becalled mica tempered; it may be that the mica and quartz particleswere derived from some weathered granitic rock and thus were partof the clay used rather than an added aplastic. In any case, thename Scott Micaceous seems appropriate as a designation for theware.Two miniature vessels are represented in our collections. Thelarger, represented by about one-fourth of the original pot, seems tohave had a globular or slightly ovate body, constricted neck, and lowslightly flaring rim. Its height was probably not over 70 mm., witha diameter slightly smaller and a wall thickness of 3 to 8 mm. It isgravel tempered, contains some mica, and there is a suggestion ofsimple stamping on the lower body area (pi. 68, a) . The second piece has a sharpish lip, somewhat constricted sides, abroadly pointed base, and a more or less conical cavity. The pasteis buff colored, and contains no visible aplastic. Maximum diameteris 36 mm., height about 30 mm. (fig. 81) . Figure 81.?Miniature pottery vessel (USNM 387097) from Scott County site, 14SC1.Actual size.Among the several pottery wares represented in our Kansas collec-tions, as described in preceding sections of this report, there are nonethat are likely to be confused with Scott Plain or Scott Micaceous.Sherds of the first were picked up by members of our party here andthere in the Scott County region, at w^iat were probably hunting orotherwise temporary camp sites. Both wares, on present evidence,appear to have a definitely western or High Plains occurrence in theState. Outside the State, presumably similar types have been de-scribed (Hill and Metcalf, 1942, pp. 179-185) from the Lovitt site inChase County, Nebr., and elsewhere nearby. The majority ware in the 444 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174Nebraska sites is sand tempered, and corresponds to our Scott Plain ; the mica-tempered minority ware is evidently analogous to ScottMicaceous. Noteworthy differences are apparent in certain details,however. Thus, about 30 percent of the Lovitt site sherds are simplestamped, as against less than 1 percent so treated in the Scott Countysite. Furthermore, vessel lips include flattened and everted examples,and about one-seventh of the lip sherds bear incised decoration. Thesedifferences, I suspect, are attributable to the relative proximity ofthe Nebraska sites to other archeological cultures, as for example, theprotohistoric Pawnee, among whom simple stamping and lip incisingwere characteristic procedures. Despite such differences in mattersof detail, I think there can be no doubt that the Scott County waresare closely and directly related to the more fully described potteryof the Lovitt site. Exotic SheedsOnly four sherds in our collections, all from test No. 2 north of thepueblo, can be certainly regarded as of nonlocal origin. All are thicklyshell tempered, have a light-gray to buff paste with a chalky "feel,"and are softer (about 2-2.5) than the local wares. One is a rimsherd,with diagonally incised or "ticked" lip ; another is a section of thickloop handle with a sort of sleeve at one end from being "riveted" intothe vessel wall; the other two are undecorated and unmodified bodyfragments. In every respect they are indistinguishable from the shell-tempered sherds we found almost exclusively in sites on the lower Wal-nut River east of Arkansas City, and they are probably identifiable asCowley Plain ware. Similar shell-tempered sherds occur as a smallminority group in sites in Rice County, a nearer but perhaps ratherless likely source.Somewhat to our surprise, our excavations about the pueblo ruinproduced no sherds of puebloan ware. Williston (1899, p. 112) wasinformed by a resident of Scott City that he had found "a good deal ofcolored pottery" in the ruin. Here also according to Martin ( 1909, p.18), "Coiled as well as smooth pottery was found, but only a singlepiece that showed evidence of decoration. Some of this pottery hasbeen submitted to Professor Hewitt, of Las Vegas, N. M., who . . . wasof the opinion that all this pottery had been introduced from NewMexico, and had not been made in the vicinity of the building or vil-lage." His observation that "... the pottery found was in part com-posed of plaster of Paris, possibly obtained from the crystals of selenitescattered over the chalk exposures in the vicinity ..." applies to noneof the sherds we recovered, and we found neither coiled nor smoothdecorated (painted?) fragments. There is no indication as to thenumber of puebloan sherds taken from the ruin, or the presence ofother types comparable to Scott Plain and Scott Micaceous. If pueb- Wedbl] an introduction TO KANSAS ARCHEOLOGY 445loan pottery in the ruin was scarce, its absence from our diggings in themiddens is not so surprising. There is thus, unfortunately, no way ofjudging to what extent our midden findings parallel those in thepueblo, as regards ceramic remains.Elsewhere, I have mentioned the recovery by A. T. Hill of a TewaPolychrome sherd in the vicinity of roasting pits and caches exposedby lowered lake waters a few hundred yards north of our diggings.The material associated with the pits and caches is indistinguishablefrom our midden findings, and another camp or village site of the samepeoples as those about the monument seems indicated. With dueregard for the surface provenience, from a normally lake-covered site,of this sherd (pi. 69, a), it should be noted that Tewa Polychrome isa protohistoric ware. According to Mera (letter to Wedel, August9, 1945) : ... A time range for this type runs from about the last two decades of the 17th,up to the early years of the 18th, centuries. The villages where it occurs, in con-centrations suflScient to indicate manufacture, are situated along the Rio Grande,roughly from about the mouth of the Santa Fe River on the south to that ofthe Santa Cruz River on the north, a distance of some thirty miles ....We shall return, in a later section, to a discussion of the chronologi-cal implications of this association. It would be interesting to knowhow this sherd compares in time and provenience with the puebloanmaterials taken from the ruin by Martin and by Williston's informant.Exotic SherdsOwing probably to the fact that most of our work was in refuse de-posits, we found no unbroken pipes of any kind. Fragments, how-ever, were relatively plentiful ; as with the other ceramic remains, theywere mostly reduced to tantalizingly small pieces, and so it is a verysketchy, though suggestive, glimpse we have of this particular artifactcomplex. Of some 33 fragments believed to be parts of pipes, 30 areof clay and probably represent upward of 25 separate objects orig-inally. The other three pieces are stone, and come evidently from twopipes.Among the clay-pipe fragments, six are characterized by a ratherdark moderately coarse granular paste, with sand grains and com-monly some mica visible on fresh breaks. Surfaces are unevenlysmoothed, and the pieces give the impression of being the product oflocal craftsmen. Many of the remaining fragments have a fine evenpaste with few or no visible inclusions, appear to have been much betterfired, and show a better surface finish. Some, at least, of these appearto have been imported, or at any rate to have been made by personswith a much higher technical ability than those who produced the pre-ceding pieces. 446 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174As to form, three of the clay specimens are certainly of the taperedstraight tubular "cloud-blower" type. One longitudinally split ex-ample (USNM 387123; pi. 69, k) is 95 mm. long with a maximumdiameter of 25 mm. The stem bore is 7 by 40 mm., the bowl cavity16 by 55 mm. In cross section the pipe is, or was, evidently circularthroughout its entire length. Another specimen (USNM 386609; pi.69, i) has a present length of 51 mm., with a diameter at the bit of 18mm. and at the broken end of the bowl of 24 mm. Tlie bit was evi-dently flattened ; the bowl cavity, 17 mm. in diameter, is heavily caked.The third fragment (USNM 386733) is 65 mm. long, and also showsa flattened bit ; the stem bore is 5 by 15 mm., whence the caked bowlcavity enlarges to at least 15 mm. diameter. Outside diameters are13 mm. at the bit and 22 mm. at the broken end of the bowl (pi. 69, ^) . All these pieces are undecorated, and I think must be of local make.Two other fragments (USNM 386587, 386850) show an obtuse angleor bend between the straight stem bore and the expanding bowl cavity,and are surely from bent-tubular pipes. They were at least 20 to 30mm. in diameter but their original length and exact conformationsare unknown. The first has a broken protuberance on the outside ofthe bend, somewhat reminiscent of the spur on early European tradepipes (pi. 69, i).One other feature concerning pipe shapes must be noted. Of 12 or13 pieces showing all or most of the bit, only 3 or 4 have a circular crosssection. The other 9 show a flattened section, or one that is more orless lozenge shaped with rounded angles. Examples intergrade fromthe circular to the nearly flat form, with a rather striking differencebetween the two extremes of the series. It should be noted that in atleast two instances, the flattening is associated with a sudden terminalwidening (pi. 69, e; USNM 386679) or with an elaborate modelingor serration of the edges (fig. 82, e; pi. 69, d\ USNM 386842) of thestem. Another small piece (USNM 386637, not shown), whose posi-tion in the pipe is uncertain, has a prominent longitudinal shouldermodeled on an angle of nearly 90?. These last three specimens areall of hard well-fired ware, light gray in color, with well-smoothedsurfaces, and do not conform to the local pottery tradition.Perhaps the most interesting pipe fragments from the site are 5 or6 that bear incised decoration, and several with elaborations of format the bit and bowl ends. The incising consists of fine lines enclosingor bordering small areas filled with minute punctations or very shortstroked elements done with a very fine-pointed implement. Thesemarkings were evidently arranged in patterns, the overall appearanceof which cannot be determined from the fragments available. Severalline and dot design remnants are shown in figure 82, a, &, c, along withthree instances of cross-hatched incising. The piece shown in plate Wedhl] an introduction to KANSAS ARCHEOLOGY 447 Figure 82.?Pottery pipe fragments from Scott County site, 14SC1, showing Rio GrandePueblo influence in decoration or bit elaboration (^r). Actual size.69, 6, (USNM 386878) is of interest further because of the well-madeoffset of about 4 mm. at the end of the bowl. The elaborate stem inplate 69, d, (USNM 386842) has already been noted. Not shown isa heavily caked bowl fragment (USNM 386972) with about 20 closelyset but unevenly spaced parallel lines that apparently encircled thepipe; and a smaller piece (USNM 386808) with 5 parallel encirclinglines.The three stone-pipe fragments, we may note in passing, appear tohave come from two pipes. One, of soft light-gray limestone, wasapparently of the tapered tubular form. The other fragments are of areddish stone flecked with white calcareous particles that dissolve inhydrochloric acid; the form of the pipe from which they came isindeterminate.The pipes indicated by our series of fragmentary specimens musthave resembled the more complete series from the Lovitt site in Ne-braska (Hill and Metcalf, 1942, pp. 185-188), where tapered tubularclay forms with flaring bit, and bearing incised and jfinely punctateor pricked decoration, are said to be characteristic. As Hill and Met-4,84172?59 30 448 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174 calf have indicated, this is not a typical Plains form, and it has notbeen reported from any other archeological horizon in the Kansas-Nebraska area. As a matter of fact, the only comparable pieces Iknow from this region are those collected by Martin (1909, pi. VI,13-18, and pi. VIII, 62) from the Scott County pueblo ruin. Thata strong, and probably direct, puebloan influence must be recognizedin these specimens is clear, I think. Some of our fragments certainly,and perhaps most of them, are from pipes that closely parallel Kid-der's (1932, p. 156 ff.) Class I pipes from Pecos, either Type A("round slim types") or Type C ("heavy fat type"). The "fishtail"form of mouthpiece, flattened and expanded laterally, is a characteris-tic of Kidder's Class II, many examples of which are further describedas having "round collar-like bowl-ends of smaller diameter than thebody proper." One such bowl end (USKM 386878) from our diggingshas already been noted. The fine-line incising, punctating, and prick-ing, as well as the cross hatching, noted on Scott County pieces occuron Class II, Type A pipes at Pecos, as do rectilinear design blocksfilled alternately with parallel lines, punctates, and other motifs (cf.Kidder, 1932, fig. 141 g and Martin, 1909, pi. VIII, 62). Finally, iffurther evidence of southwestern contacts in this matter is needed, Ihave an opinion from Mera, to whom several of our Scott Countypieces were submitted for examination. Concerning the serrate-edgedstem, the flared bit, and the "shouldered" fragment, all noted above,Mera (letter of August 9, 1945) observes that "All of the pipe frag-ments, No. 386842, No. 386679, and No. 386637 are unquestionably oftypical Rio Grande style and would correlate nicely with the sherd[of Tewa Polychrome] in age." Two other fragments (USNM 386609and 386778, the latter with cross-hatched and pricked decoration),according to the same observer, "do not conform to the prevailingfashions in early historic times as do the others. The character ofthe paste is also megascopically quite different from most Rio Grandeexamples familiar to me. Perhaps manufacture at the place of dis-covery might be conjectured if an intrusion of people from the westcan be considered." OBJECTS OF ANTLEB AND BONEAntler objects.?Two antler-tip fragments show modification byman. One suggests the flattened tip of a scraper haft, such as thosedescribed by Hill and Metcalf (1942, p. 200 and pi. IX, fig. 1) from25CH1 and 25FT9. The other has a nicked and scarred tip (fig.83). Another section of antler shaft has been cut off at one end,broken diagonally at the other; it may be rejectage. A fourth pieceseems to be thinly scraped antler plate, apparently cut from the shaftwhere a branch tine occurred. The edges are cut, the ends broken,but striations suggest attempts at scraping or rubbing down the Wedel] an introduction TO KANSAS ARCHEOLOGY 449 Figure 8J.?Worked and polished antler tip from Scott County site i4SCl.original surface. Both deer and elk antler are represented by thesespecimens.Scapula digging tools.?None of these are complete, but there area great many dressed and worn fragments that without doubt canbe so identified. The fragments show sharpening of the vertebralborder, removal of the scapular spine, trimming of angles andborders, and partial or complete removal of the head of the bone.The sharpened vertebral border is usually beveled and worn to ahigh polish. The scar left in removal of the scapular spine is nearlyalways jagged and irregular, in no case ground smooth as in theCowley County specimens noted elsewhere. None have any trace ofthe socket or groove at the head that characterizes the Rice andCowley County implements. In one case, where the head was onlypartially removed, it was whittled down from both sides to aboutone-third of its normal thickness, and the adjacent dorsal and ventralsurfaces of the bone have been flattened. The range in size and form,and the exact manner of hafting, remain uncertain.Awls.?Of 65 awls and awl fragments, 14 were fashioned frommammal leg bone and the remainder from mammal ribs. Those ofleg bone include 2 where the head of the bone has not been modifiedexcept by the original splitting; they are 109 and 120 mm. long,much worn, and may be of deer metapodial (pi. 70, Z, m). Anotherspecimen, 77 mm. long, has the head partly worked do^vn. Six awls,from 47 to 157 mm. long, were improvised from irregular splintersmodified only by the rubbing down of one end to make a piercinginstrument; the shaft and butt are always ragged and unfinished(pi. 70, 6>, p). Five other awl tips are too fragmentary to beclassified.Awls of mammal rib include one rounded butt fragment, care-fully dressed on all surfaces but with the cancellous tissue exposedin the initial splitting only partly smoothed. Nineteen others aremade from face splinters of split rib, pointed at one end but, unlikethe preceding piece, not otherwise finished on the sides, edges, orbutt. Varying in length from 45 to 162 mm. this group includes theheaviest piercing tools from the site (pi. 70, n) . Twenty-nine awls, and possibly two additional fragments, compris-ing in all nearly half the awls from the site, were made from the edge 450 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174 of a bison rib or the anterior margin of the neural process of the sameanimal. They are triangular, or nearly so, in cross section, withrounded corners, and one flat side nearly always shows a faint trace ofcancellous tissue. The butt is rounded or rounded conical, and theshaft tapers quite evenly to the tip. Complete examples are 60 to98 mm. long, but there are broken pieces unquestionably of the sametype that were not less than 168 mm. long. Maximum diameter is12 mm., but few exceed 8 mm. Two or three have somewhat flattenedtips, and possibly should be classed as polishing tools. These awlsare about as well made as any from the site, and though not so heavyas the split rib specimens, they must have been a serviceable and ef-ficient instrument. Their near identity with the most common awltype at Rice County sites should be noted (pi. 70, q^ r) .In the overwhelming preponderance of awls made of split ribs andneural spines as contrasted to leg bones, the Scott County site againconforms closely to the Lovitt site (Hill and Metcalf, 1942, p. 196).Both of these types also occur at Lower Loup Focus sites in east-cen-tral Nebraska, but in what proportions, relative to one another and tosplit leg-bone awls, I am not able to say (cf. Dunlevy, 1936, p. 197).Fleshing tools.?One complete, or nearly so, specimen, and fivefragments were found ; all were apparently made from bison metatar-sals, originally with the ankle bones attached. In most cases, theselatter have become detached through decay of the ligaments and arelost, so that only the modified metatarsal or a fragment remains.The most nearly complete specimen (pi. 71, a-, USNM 387009) in-cludes the metatarsal and some of the ankle bones, and has an over-all length of 208 mm. The metatarsal has been shaped by cuttingdiagonally downward from the anterior surface so as to remove thedistal end and produce on the posterior wall a rounded gougelike edgefurther sharpened by beveling from the interior surface. The pos-terior face, just above the working edge, has fine striations whosemeaning is not clear. The blade is somewhat battered, but shows notrace of serrations.Of the remaining five specimens, two consist of the metatarsal onlyand three of short sections of metatarsal including the working edge.One of the former (pi. 71, h) has 17 narrow notches cut into theworking edge, the other is unnotched ; and in both the blade is beveledfrom the posterior surface. They are 163 and 158 mm. long. Thethree fragments, from 82 to 103 mm. long, have 13 to 14 serrationseach on the sharpened end (pi. 71, c, d) . Similar implements, usually serrate-edged and often with ironblades, were widely used in historic times by the Plains and othertribes (Wissler, 1912, p. 58; Mason, 1891, p. 589; Fletcher and LaFlesche, 1911, p. 343 and fig. 70) for removing bits of flesh and fat Whdbl] an introduction TO KANSAS ARCHEOLOGY 451from animal hides. Archeologically, they have been reported fromseveral historic, and at least one protohistoric, Pawnee sites in Ne-braska (Wedel, 1936, p. 84, and pi. 10, e-h; Dimlevy, 1936, p. 198 andpi. 14D) ; from the Lovitt site (Hill and Metcalf, 1942, p. 198) ; fromArikara (Strong, 1940, p. 370), Mandan (Will and Spinden, 1906, p.169), and Cheyenne (Strong, 1940, p. 375) sites; and from the Hagensite in Montana (Mulloy, 1942, p. 71). Curiously enough, we didnot find them at any of the protohistoric sites in Rice and CowleyCounties, Kans., though I did pick up one small fragment on a cul-turally related site near Larned at the mouth of Pawnee Creek. AtPecos several specimens were found (Kidder, 1932, p. 233) in roomsand kivas "... of post-Columbian date, and most of them . . .were apparently not deserted before the latter part of the eighteenthcentury." Earlier occurrences should be noted, such as Jenks' (1932,p. 461) Arvilla gravel pit specimen and two notched deer metapodialsfrom a Woodland site in Valley County, Nebr. (Hill and Kivett, 1941,p. 166). No comparable serrate-edged pieces have yet been reportedin the Upper Republican and contemporary Central Plains complexes,though these intervene temporally between the Woodland and theabove-listed protohistoric and historic occurrences.Rib implements^ striated.?Sixteen specimens have been fashionedfrom sections of bison rib, varying from 114 to 176 mm. in lengthby 22 to 33 mm. wide. In each, one end of the rib has been neatlycut off; the other was either cut and irregularly broken off, orelse shows only a ragged fracture, though in either case this brokenend shows some evidence of smoothing. The cut end is squarish orsomewhat rounded, and in most cases the external surface of the bonehas been beveled or sloped back by oblique grinding of the cut end(fig. 84, a). The most distinctive feature of these objects is a seriesof fine transverse striations (pi. 71, /-;'), beginning just above the cutend, running entirely across the bone, and covering from 10 to 40mm. of its length. The striations are straight or nearly so, closelybunched, and in some specimens have worn the bone surface down asmuch as a millimeter or more. Implements with both ends cut andfinished have a striated zone at each extremity (pi. 71, /, g). Invari-ably, the striations occur on the internal surface of the bone oppositethe beveled facet where the latter is present; the adjacent cut end isusually smoothed or rubbed down, but not as if from extended wear.The scratches look as if they had been made with a sharp flake orflint knife used with a sawing or cutting motion, but they are notefforts at cutting through the bone nor do I see how they could haveresulted from flint chipping.Whatever their use, these objects differ widely from the scored ribsections, variously designated as musical rasps, tallies, etc., that occur 452 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174 if ' 1^1 Figure 84.?Finished and transversely striated rib implements from Scott County site14SC1. Length of a, 13.5 cm.in such numbers in protohistoric and historic Plains sites; the stria-tions are much finer, shallower, and more closely bunched than thebold transverse scorings on the "rasps", and the implements showrather more care in shaping. Martin reports none of the striated speci-mens from the pueblo ruin, though he does figure (Martin, 1909, pi.VII, 27) a section of scored rib; we, on our part, found none of thescored ribs in the midden excavations. The striated implements seem-ingly have not been reported from other Plains sites, and I have beenunable to find a description of comparable objects elsewhere.Needle fragments.?Four narrow dressed strips of bone appear tobe from needles. Two are 55 and 62 mm. long, slightly curved, andabout 5 mm. wide, with sides and edges carefully smoothed. One endof each is broken ; the other end is cut off square, and there is a 2 mm.perforation 4 to 6 mm. from the butt. The other fragments, 23 and40 mm. long, are broken at both ends and have no perforations, butin width, thickness, and finish are like the foregoing specimens (pi.70,^).Bone arrowpoint ( ? ) .?This is 69 mm. long by 5 mm. in diameter,in form straight, tapered, and circular in cross section. One end is Wboel] AN INTRODUCTION TO KANSAS ARCHEOLOGY 453pointed; about 8 mm. from the other end, the taper becomes moreabrupt, developing into what suggests a stem for insertion into ahollow shaft or foreshaft. This latter feature seems to rule out thepossibility of an awl, and may mean that a projectile point is repre-sented. The piece is well finished but not polished (pi. 70, h).Flaking tools ( ?).?Four objects made from the edge of a mammalrib or neural spine somewhat resemble a common form of awl from thesite. However, they are generally heavier and not as well finished,and the rounded to conical butts are nicked and roughened. Theiridentification as flaking tools seems more plausible than the view thatthey represent crude awls. They vary in length from 55 to 115 mm.(pl.70,e).Objects of cancellous hone.?There are two of these. One is irregu-larly wedge shaped, with smoothed surfaces and thin edges ; it meas-ures 34 by 31 by 6 mm. Despite the absence of pigment residue, thepiece suggests in shape and size the bone paint applicators of thehistoric Plains tribes. A second piece, 38 by 35 by 15 mm., is ellipticin form, thinning slightly to a moderately thick edge. As with thepreceding specimen, this retains no trace of the original compact sur-face bone. Both were doubtless fashioned from pieces of the innomi-nate, femur head, or other large bones of the bison (pi. 70, d).Tubular beads.?There are 56 complete tubular bone beads, besidesperhaps a dozen additional fragments. All were made of bird orsmall mammal bone, from which the articular extremities have beencut. Two bones showing this process were found, with one extremitydetached, the other grooved but not gone. Beads vary in length from14 to nearly 90 mm. Nearly all are plain and undecorated but oftenwith a use-polish, and many show a slight curvature. The only dec-orative attempts are faint encircling incisions about 3 to 6 mm. fromeach end of two or three specimens. These objects were found in allparts of our diggings and, because they could be easily made frommaterials readily available, were no doubt popular among the natives.The table below shows their range in length and diameter (pi. 70, b, c) . Table 17 Diameter 454 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174Bone whistle.?This specimen, incomplete, is made of bird bone likethat used in the larger beads. It is now 79 mm. long, with one endfinished, the other broken. About 10 mm. from the finished end is ahole 4.5 mm. across, rather ineptly made by sawing or whittling theshaft with a flint flake. I cannot say whether this was actually awhistle with a single aperture, as seems likely, or alternatively, wasthe end of a flageolet with a series of stops. Martin (1909, pi. VII, 25)figures a "bone piccolo" with a large stop at one end and 2, or possibly3, small stops at the other. We found no evidence of this latter typein our work, but its presence in Martin's collections is, of course, onemore bit of evidence for southwestern influences (cf. Kidder, 1932,p. 249) at the Scott County pueblo site (pi. TO, a) .Worked bison hyoids.?Two flattened objects with cut ends weremade from the hyoid bone or stylohyal of the bison ; neither is perfo-rated or decorated, though on one the ends have been smoothed. Theymeasure 28 and 48 mm. in length; their purpose is not known(pi. 70, /).The end of a third hyoid is evidently rejectage from the manufactureof pieces like the above. The shaft was cut from both surfaces untilthe central cavity was reached, and the extremity was then snapped olf . The remaining portion of the shaft above the cut end is smoothed,suggesting that the bone had been rubbed down and polished beforethe shaft segment was detached.Bracelet fragment ( ?) .?This is a thin plate of scraped bone, meas-uring about 50 mm. long by 32 mm. wide by 1.5 mm. thick. The edgesare rounded and smoothed, the ends broken ; the piece is transverselyflat and longitudinally curved. The convex upper surface has beenwell smoothed, whereas the concave face shows striations suggestingtraces of cancellous structure. There is no ornamentation. Purpose ofthe piece is unknown ; it may be part of a bracelet or bow guard.Bone knife.?A thin flat subrectangular piece, measuring 28 by 65mm., has been cut from the blade of a scapula digging tool. One longedge shows the thin beveling of the original tool blade; the otheredges have been rounded and smoothed, beyond doubt purposely. Thepiece may have been a "squash knife" or something for similar use.Perforated objects.?There are two of these. One (pi. 70, e) is froma thin rectangular object 28 mm. wide with straight edges and onesquared end. Near this end are two small holes about 15 mm. apart.The other end is broken. The second piece is longer, narrower, and lessregular, with a single small hole near one long edge.OBJECTS OF CHIPPED STONEChipped stone artifacts were present in considerable numbers,though generally speaking, they showed less care or skill in manufac-ture than did those from the Kice and Cowley County sites. There Wbdbl] an mTRODUCTION TO KANSAS ARCHEOLOGY 455 was, moreover, a rather more limited range in artifact forms at theScott County site. Raw materials included brown jasper, chalcedony,moss agate, variously colored cherts, and a very little obsidian. Fos-siliferous Florence flint was not found, and if known locally, musthave been very rare.Projectile points.?There are 99 of these. Characteristically, theyare small, thin, well made, with bifacial retouching, a length varyingfrom 15 to 33 mm. and averaging between 18 and 25 mm., and a widthof 9 to 16 mm. In form, 78 (78+ percent) are triangular and un-notched, with base slightly convex or straight and edges straight orslightly convex. Eleven others (11+ percent) differ only in havinga pair of small lateral notches about 3 to 6 mm. above the base. Sixhave a rather markedly convex base, as though the comers of a simpletriangle had been cut off (pi. 72, a,h).The remaining four points are somewhat larger than the majorityof the triangular specimens, from which they diverge chiefly in beingthicker, relatively broader, more coarsely flaked, and in having barbsor shoulders and an expanding stem with convex base. The mostnearly complete example measures 29 by 19 by 4.8 mm. ; the otherswere apparently of comparable size. I suspect that these were not oflocal manufacture; similar pieces occur on mixed hunting sitesthroughout the region, and our specimens may have been picked up byhunting parties from the Scott Comity site and carried back to theirvillage. I am not able to suggest, however, the specific culturalhorizon to which points of this type should be referred (pi. 72, d).Not included in the above count are three tiny chipped objectssomewhat suggesting reworked triangular points (pi. 72, c) but possi-bly intended for some other use. They range in base width from 11 to13 mm., and in length from 10 to 15 mm. ; in two, width slightly exceedslength. Two have a slight shoulder between the retouched point andthe basal flange ; in the third the point is off-center. It is possible theywere mounted in wooden or bone handles and used as reamers, scar-ifiers, gravers, or for other delicate incising or cutting.Drills.?Thirty-four complete and nearly complete drills, and frag-ments of 6 or 7 others, came to light. The prevalent form has a slendercarefully chipped shaft 10-30 mm. long, elliptical to quadrilateral incross section, which expands abruptly into a large flat base represent-ing the original flake with edges slightly or not at all retouched. Thesize and shape of the basal flange varies according to the conformationsof the original spall. The two largest examples measure 60 mm. inoverall length, though most are under 50 mm. Retouching of the baseedges, where present, was just enough to remove the sharp edges andangles. There were 28 drills of this type (fig. 85, c; pi. 72, f-h). 456 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174 Figure 85.?Straight-shafted and expanded base drills from Scott County site, 14SC1.Actual size.Three specimens resemble the above in having the abruptly widenedbase, but differ in that the base is much smaller and has been retouchedall around to give a subrectangular to oblong outline. In these, theretouching is somewhat inferior to that on the preceding ; shafts are13 to 27 mm. long, overall length 29 to 53 mm.There are two plain-shafted drills. One, with both ends damaged,is 61 mm. long; it has a thick lenticular cross section, and is rudelyflaked. It somewhat suggests the large heavy plain-shafted "pipe"drills from Kice County but is smaller, less well made, and lacks thecharacteristic blunting of the long edges (fig. 85, a). A somewhatsimilar but smaller drill, 39 mm. long, is better made ; it has a trian-gular cross section and shows wear at both ends (fig. 85, &). The re-maining specimen, 30 mm. long, has a quadrilateral cross section ; it iswidest about one-quarter of the distance up from the butt, which isrounded and shows no wear.Knives.?No complete knives were recovered in our work, but ap-proximately 134 fragments showing one edge retouched from bothfaces are tentatively classed as cutting tools (pi. 73, d-f). Because oftheir invariably broken condition the size and shape of the knives re-main uncertain. Some may have been improvised from suitable spallsof generally elongate outline, but there are hints of lanceolate, subrec-tangular, or ovate forms; and an occasional thin carefully workedfragment suggests that there must have been some very fine examplesof flint-working among these tools. There is evidence, too, that beveledknives were occasionally used, but of these there are not more than a WedblJ an introduction TO KANSAS ARCHEOLOGY 457dozen fragments; they indicate a tapered blade, and there is one in-complete piece that suggests the four-edged beveled "Harahey" type(cf. Hill and Metcalf, 1942, p. 190). There were also some relativelylarge rough specimens that seem to have been upward of 12 to 15 cm.long by 6 mm. wide, with well sharpened edges. It is possible thatthese ought to be classed as choppers or perhaps as hide scrapers,though in any case their use is largely conjectural.Martin (1909, pi. VII, 32 and 36; pi. VIII, 57 and 58; and pi. IX,84) illustrates a number of cutting, chopping, and/or scraping toolsthat probably do not differ much from the pieces represented by ourseries of fragments. I am unable to explain the extraordinary abun-dance of broken chipped artifacts and the very few whole pieces wefound. The fact that nearly all of our material is from middens maybe a partial explanation, yet even here one would expect rather moreintact pieces than we obtained. One has the impression that artifactswere either subjected to very heavy use and unusually complete break-age or else that they were deliberately smashed up before being finallycast away.End scrapers.?These, too, occured in relatively large numbers, anddiffered in no important particular from those at other Plains campand village sites. The under side is flat or slightly curved, withoutretouching, and represents the unmodified original surface of theflake ; commonly, the bulb of percussion is still visible at the smallerend. In form, the scrapers are subovate to subelliptic in outline, butwith a number of specimens quite irregular in shape. The wider endin all cases has been chipped back to an angle of 45?-80?, and the maxi-mum width and thickness usually occur here. The long edges are re-touched, the narrow end less often so. The dorsal surface is variable ; sometimes it shows only two or three large broad flake scars, and islow and flattish; in other specimens it is strongly ridged. Severalspecimens have a suggestion of a small protuberance or graverlikepoint, this usually occurring where the broad working end meets alateral edge (cf. Champe, 1949, p. 289; Gunnerson, 1959). Lengthranges from 20 to 80 mm. ; but of 281 specimens only 79 (28.1 percent)are under 38 mm. and 129 (45.6 percent) are between 38 and 57 mm.They thus average appreciably larger than the end scrapers from RiceCounty protohistoric sites, and also show less care in shaping or skillin flaking (pi. 72, z, j).Side scrapers.?^This group, comprising nearly 200 specimens, in-cludes unshaped pieces with the edge, or edges, retouched from oneside only. Most are irregularly shaped spalls and flakes, up to 100mm. long and 20 mm. thick. One surface is flat or nearly so, the otheris ridged or otherwise convex, and the edge retouching is always doneon this convex side. In some cases the working edge is blunted from 458 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174 use. For the most part they suggest improvisations, and contrastmarkedly with the carefully shaped end scrapers. They approach theknives rather closely, however, except in the manner of retouchingthe edge, and the distinction here made between knives and sidescrapers may not reflect a similar dichotomy of use among the natives.Probably cutting and scraping could be done about equally well witheither tool.Hide scrapers.?There are at least 10 or 12 heavy edged implementsthat almost certainly represent the large hide scrapers characteristicof the historic Plains tribes. Most are broken, but they were evidentlymore or less elliptical, and measured up to at least 100 by 70 by 20mm. One long edge was retouched, and the ends were partially sotreated ; the other long edge, constituting the back, was left blunt forconvenience in grasping (pi. 73, a). Similar objects, but usually ofquartzite rather than of brown jasper as here, are common at Pawneesites in Nebraska (Strong, 1935, pi. 1, fig. 2, a, d; Wedel, 1936, p. 76) ; and in 1933 I saw identical implements still in use among theComanche near Walters, Okla.OBJECTS OF GKOUND AND PECKED STONEMealing slabs and manos.?Only one mealing slab was found byus, and it is apparently incomplete. Of limestone, it measures 235by 135 by 48 mm., and there is some evidence of shaping by percussionor spalling at each end. The upper surface is worn in very slightlyconcave manner.Manos include three fragments and one whole specimen (pi. 73, i).The latter is of pinkish quartzite, measures 103 by 65 by 37 mm.,and has one worn working face (fig. 86). One fragment is from amuch larger stone that may once have been as much as 200 mm. long Figure 86.?Single face mano of quartzite from Scott County site, 14SC1. Actual size. Wedel] .^t introduction TO KANSAS ARCHEOLOGY 459it measures at present 100 mm. long, with both ends broken, and is 110mm. wide by 45 mm. thick; the surfaces seem to be little worn. Theremaining fragments are unquestionably from long-used specimens,prism shaped or triangular in cross section, with one broad and twonarrower grinding surfaces. Tlie larger of these manos was at least95 mm. wide by 35 mm. thick.The previous excavations at the pueblo netted a larger series ofgrinding implements than did our work; Martin (1909, pi. VIII, 60and pi. IX) recovered several flat-surfaced metates and a number ofmanos, the latter including both short and long round-ended forms.A puebloan corn-grinding complex seems indicated by the availableevidence.Mortars or anvil stones.?There are two of these, one of diorite,the other of quartzite. Both are irregular, unshaped boulders notexceeding 140 mm. in greatest diameter and 65 mm. in thickness. Oneor both of the larger surfaces are slightly concave, either from grind-ing or from pounding. The specimens resemble the small mortars,used on a rawhide basin, on which pemmican was pounded by theDakota and other historic Plains tribes.Shaft straightener.?This is a flat-bottomed elliptical block of fine-textured limestone, with planoconvex cross section transversely andlongitudinally. It is incomplete, but assuming a general symmetry ofoutline the original length and width were each approximately 55mm., thickness 28 mm. Across the highest part of the convex surfaceis a smoothly worn transverse groove 15 mm. wide by 5 mm. deep.There is a hint of an attempt to mark out by pecking a longitudinalgroove on the flat base. This is not a typical Plains implement, butbasically it is like the southwestern shaft straighteners (cf. Kidder,1932, and p. 284) and may have been used for a similar purpose.Shaft smoothers.?Fragments of elongate-shaped sandstone blocks,bearing a longitudinal groove, were common everywhere in the de-posits. None of the pieces are large enough to show the size and shapeof the original objects, but they do have the groove on a flat surface,with rounding sides and bottom, while a few pieces show the boatlikeends of the common paired buffers. It is probable, though not proved,that they were of that type. In size, the fragments suggest smaller,more slender forms than those we found on Rice County sites, and thereis perhaps a somewhat greater variability in shape (cf. Martin, 1909,pi. VTI, 19-24). The material is usually a fine-grained, light-colored,very friable stone, with an occasional dark red-brown piece; and asMartin (op. cit.) points out, it was probably obtained from outcrop-pings of the Dakota sandstone more than 100 miles to the east of thesite. 460 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174Sharpening hlochs.?There are a few other irregular, and appar-ently not intentionally shaped, sandstone blocks bearing short, oftenconvergent or even crisscross grooves of varying width, depth, andlength. These, I suppose were for sharpening awls, needles, andsimilar objects.Pipe blank (?).?This is a carefully shaped block of smooth fine-grained limestone, nearly square in cross section ; it measures 108 mm.long by 22 to 25 mm. in width and thickness. All four sides and bothends have been dressed flat, and are unmarked except by fine grindingstriations. Its purpose, is unknown, but the size and proportions sug-gest that it may have been a blank for a tubular pipe (pi. 73, j).OBJECTS OF UNWOBKED STONEHammerstone.?There is but a single example in our collections.It is of gray quartzite, rounded, and measures about 50 by 60 mm.,with the former edges and angles rounded off from long-continuedhammering or pecking. This single occurrence is probably a veryunreliable indicator of the extent to which such objects were oncepresent, and numbers of the stones may have been overlooked in thedigging.Hematite.?Several small bits of soft friable hematite were notedat various points in our work. There were, in addition, two muchharder and more metallic pieces, both with well-marked wear facets.One of these was triangular, 24 by 15 by 8 mm. ; the other was quad-rilateral, with all surfaces ground flat, and measured 52 by 38 by 16mm. Both give a bright red-brown streak, and were undoubtedlyused as sources of pigment.Marcasite concretions.?No less than 24 of these were found. Mostare flattened circular or subcircular in outline, biconvex in cross sec-tion, and vary from 12 by 20 mm. to 25 by 55 mm. There is onereniform and one pear-shaped specimen. About half have a rusty redfriable surface; the others are dull gray and hard. None show anyevidence whatever of alteration by man, and they seem to be just asthey were when foimd by the natives. For what purpose they werebrought into the village I do not know, but their comparative abun-dance suggests something more than idle curiosity. As to their sourceI am also uncertain; we found none in the adjacent valley during ourscouting trips, though the broken terrain outside the Park to thenorth was not very closely examined. We did find numerous concre-tions of the same sizes, shapes, and material, weathering out of softBhale beds along Salt Creek just below the Pottorff site about 14miles east of Scott County pueblo. Wliether this or some nearer de-posit was the source of our archeological specimens, I cannot say(pi. 73,/). Wbdhl] an introduction TO KANSAS ARCHEOLOGY 461Miscellaneous objects.?From test 2 came a large unworked shell ofa Cretaceous oyster, perhaps brought in as an oddity ; it could havecome from a comparatively short distance.A bifurcate object of rough-surfaced stone with rounded ends, andmeasuring 115 by 67 by 21 mm., is apparently inorganic and of con-cretionary origin. It, too, is unworked.From test 2 came a flat rhomboidal crystal of gypsum; it looksweathered but is unworked. It measures 55 by 35 by 13 mm.OBJECTS OF SHF.IJ,Molluscan remains and worked shell were exceedingly scarce. Lessthan a dozen pieces have certainly, or probably, been worked. Most ofthese appear to be from bivalves, perhaps thin-shelled fresh-watermussels taken out of the nearby creek ; but they are too fragmentary ortoo much worked to permit species identification.Four specimens show drilled perforations. Two of them are rec-tangular, measuring 14 by 25 mm. and 11 by 18 mm. ; each has a smallhole near the cut end and is broken at the other. A third piece, 10 by18 mm., is pierced near one end and has another hole partly brokenaway at the other. The fourth piece is subtriangular and incomplete,with one hole. In all these examples, the drilling was mainly from theconcave surface of the shell, with just enough work on the externalsurface to prevent the drill from breaking out a ragged hole (pi. 70, j) .Four other pieces, cut and shaped to a rectanguloid form, show nodrilling.Of interest is the presence of at least three specimens certainly offoreign provenience. One is a broken univalve, identified as Olivelladama Mawe, which inhabits the Gulf of California off the west coastof Mexico. It is well worn and polished, but has undoubtedly beenspire-lopped for stringing on a cord. Two disk beads, 6 and 9 mm. indiameter, are apparently also of Olivella shell; the larger one issaucer-shaped and both have a 2.5 mm. central perforation (pi. 10, k).These disk beads, as also the spire-lopped piece, are all common typesin the Southwestern United States ; and I think it very probable thatnot only the shell but probably the finished beads themselves were im-ported into Scott County from the pueblo area.OBJECTS OF EUROPEAN MANUFACTUREEuropean trade or gift items were scarce, but since we, as well asMartin, encountered them inclusively at several points there can be nodoubt of their association with the archeological complex here repre-sented. That a French or Spanish trading post ever existed on thisspot, as was once supposed, is exceedingly doubtful; the metal andglass so far found is no more than would be expected on a site whoseoccupants doubtless roamed widely on hunting trips and occasionally 462 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174came into contact with European traders, adventurers, or colonists.All of our metallic specimens are too heavily oxidized to be recogniz-able as to their make and source, and in most cases even the identityof the object represented is not certain.Objects of iron include three specimens from test 2, all at depths of17 to 21 inches, one from test 3, and one with no provenience stated.From test 2 came a flat piece measuring 95 by 25 mm., with one thickstraight edge and the other curved; one end tapers to a point, theother suggests a shank for insertion into a handle. The piece ratherstrongly suggests the blade of a sheath knife (pi 73, g). Anothersmaller flat piece with one curving edge and a small hole as if forriveting may also be from a knife blade. The third piece, 60 mm, long,is slender and tapered, and possibly is an awl. The object from test 3is very heavily oxidized but it, too, appears tapered, and may be an-other awl fragment; it is 44 rmn. long. The fifth piece is square at oneend, about 2.5 to 3 mm. in diameter, and tapers to the other end ; it is47 mm. long, and suggests still another awl tip.From test 1, at a depth of 18 inches, was taken half of a much weath-ered, rather crumbly, pale blue glass bead. It has a diameter of 6 mm.,a length along the perforation of 5 mm., and the perforation is 1.5 rmn.in diameter. So far as I can judge, it conforms rather closely to theblue glass beads from the Tobias site in Rice County, and representsa trade type widely used over a considerable period of time.We found no objects of copper.In addition to the above finds in the middens, it should be noted thatduring the clearing of the pueblo ruin, Martin (1909, pp. 16 and 17,and pi. VIII, 59) recovered a broken iron axhead from room II, andin room V found "half of a clam shell, which had been sawed length-wise with a toothed saw, the tooth marks being very plainly apparent."This latter specimen I have not examined, and so I cannot judge thevalidity of the observation. Writing prior to Martin's excavations,Williston (1899, p. 112) quotes from a letter an unconfirmed reportthat "beads and crosses" had been found here, though the precise spotis not specified. In any event, the evidence from excavation is clearand incontrovertible, and there can be no doubt that the Indians whoinhabited this spot had been in contact, directly or indirectly, withEuropeans. In view of the evident contacts with puebloan peoples, Iwould suspect further that the metal and glass goods came from thesame region and were probably of Spanish origin, though the possi-bility of French provenience cannot be ruled out.DISCUSSION AND CORRELATIONSTo students of Plains archeology, it will be readily apparent thatthe artifact materials described in the foregoing pages comprise afairly representative Plains culture complex of early liistoric (proto- Weuki.J an introduction TO KANSAS ARCHEOLOGY 463historic) times. There are a few items, such as the incised and flare-bitted clay pipes and Olivella shell beads, that are not ordinarilyfound in Plains sites; but such pieces constitute such a relativelysmall fraction of the total present that they can readily be explainedon the basis of trade contracts with, or cultural influences from, theEio Grande area. This heavy preponderance of typical late Plainsartifact types is not altogether what I would have expected in viewof the proximity of our diggings to what is apparently a ruined struc-ture of puebloan, not Plains Indian, origin.Comparison of our artifact sample with that reported by Willistonand Martin from their excavations in the pueblo ruin indicates thattheir collections parallel in many ways our findings. Puebloan ma-terials are clearly suggested in the coiled pottery briefly described byMartin (1909, p. 18), in the incised tubular clay pipes, and perhapsin the flageolet and the metates, all presumably taken from the ruinitself. Otherwise, the chipped projectile points, end scrapers, ex-panded base drills, knives and choppers, such ground stone objects aslongitudinally grooved shafts smoothers and perhaps small grinders(anvil stones?), the bone awls, toothed flesher, scapula digging toolfragments, and, of course, charred corn, are in line with our findingsin the refuse deposits outside the ruin. All this suggests that whilethe architecture and some of the furniture, such as the grindingtrough and rectangular slab-lined hearths, and probably the nearbyirrigation ditches, were undoubtedly of puebloan origin, the materialculture in possession of the inhabitants was generally far less pueb-loan than it was Plains Indian in character.Among the described and named culture complexes now recognizedin the Central Plains, the Scott County site (14SC1) evidently findsits closest counterpart in the Dismal Eiver Aspect sites of westernNebraska. I have already noted the close resemblances in ceramics,pottery pipes, and certain other artifact categories between 14SC1and the Lovitt site, 25CHI, on Stinking Water creek in Chase Coun-ty, Nebr. The Lovitt site lies approximately 140 miles, airline, almostdue north of Scott County. The scene of extensive excavations bythe Nebraska State Historical Society in 1939, the Lovitt site isthe first Dismal River culture site to be fully described in print (Hilland Metcalf, 1942). As the type site for what is now termed theStinking Water Focus of the Dismal Eiver aspect (Gunnerson, 1959) , to which have been assigned other sites in Dundy, Frontier, Harlanand Hooker Counties, Nebr., we may here carry further our compari-sons of the complex just detailed from 14SC1.First, with respect to the sites themselves. Each occupies a terracelocation beside a small perennial creek and at some distance fromthe larger streams of its area. The Lovitt site covers some 75 acres484172?59 31 464 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174 of ground and is probably the larger; but the remains at 14SC1 areknown to extend imder Lake McBride and onto several terraces southof the flat on which most of our digging was done, and their totalextent has not been determined. At 25CH1 were found the remainsof two house structures?one marked by a circular posthole arrange-ment some 20 feet in diameter, the other by a five-posthole groupingabout 11 feet in diameter, and both with central fireplaces. Thepentagonal pattern is reminiscent, of course, of the similar structuresidentified subsequently by Champe (1949, pp. 286-288 and fig. 70), atWhite Cat Village, 25HN37, on Prairie Dog Creek near Alma, Nebr.At 14:SC1, no posthole patterns suggesting structures were uncovered,though I suspect that stripping of larger areas than we uncoveredmight well disclose their presence. Irregular basins and pits occur-red at both the Lovitt and Scott County sites, though the bell-shapedcache pits so characteristic of other protohistoric Central Plains hori-zons were rare or absent. The roasting pits at 14SC1 have no knowncounterpart at 25CH1, but they do occur at other Dismal River sitesin the southwest Nebraska area (Hill and Metcalf, 1942, pi. 3, fig. 2;Champe, 1949 a, p. 288 and fig. 70, d) . Both sites yielded some indi-cations of burned corn and considerable quantities of animal bonerefuse. At neither was there any indication of defensive works, ofthe refuse mounds characteristically scattered over the surfaces ofmany protohistoric village sites farther east, or of associated burials.The characteristic pottery at both 14SC1 and 25CH1 was a relativelythin, hard, dark-colored ware, generally sand tempered; and micatempered or micaceous sherds constituted a minority ware. At14SC1, 95 percent of the sherds were plain surfaced and less than1 percent bore simple stamping; at 25CH1, the proportions wererespectively 69 and 30 percent. About 5 percent of the Scott Countysherds were classed as micaceous, as against less than 1 percent ofmica-tempered sherds at Lovitt. The plainware sherds (LovittPlain) at 25CH1 were usually better smoothed, "with a well polished,often shiny, exterior" that is rare on Scott County sherds. At 25CH1,62 (about 15 percent) of the 425 rimsherds carried incised, punctate,or other decorative elements on the lip, as against 2 instances in 414rims at 14SC1. Traces of incised body decoration occurred on a veryfew sherds from each site. Fi-om both sites there are straight claypipes, sometimes incised and with clear indication of form elaboration.At both 14SC1 and 25Cril, the characteristic projectile point wasa small triangular shape, unnotched forms predominated stronglyover notched, and stemmed points are decidedly in the minority. Endscrapers and side scrapers were very plentiful at both sites, drills lessso; at 14SC1, expanded base drills decidedly outnumber the straightor cigar-shaped forms, whereas at 25CH1 the two types occurred in Wedel] an introduction TO KANSAS ARCHEOLOGY 465about equal proportions, and there is another type?straight, with "a short thick protuberance on the edge, apparently about midwaybetween point and center" that has not been reported from 14801.At both sites were found milling slabs, manos, anvil stones, groovedshaft smoothers, sandstone sharpening blocks, tubular stone pipes orfragments thereof, grooved mauls, and obsidian. Turquoise beadsare reported from 25CH1 and from 25DN1, but have not been foundat 14SC1.With respect to bonework, the two sites again shared a number oftraits. These included bison scapula hoes; bison metapodial fleshers,some with serrate edge, others smooth bladed ; cancellous bone paintapplicators ; stemmed projectile points ; and tubular beads. Also, atboth there was a variety of awls, including flat split-rib specimens,others made from the edge of a rib or neural spine, round to triangularin cross section, and still others fashioned by merely sharpening bonesplinters. Items found at 14SC1 but scarce or not reported from25CH1 include transversely striated rib segments with one finishedend, scored ribs, eyed needles, and awls made from split mammalleg bones. Conversely, epiphyseal hide dressers, rib shaft wrenches,pierced bison phalanges, antler scraper hafts, and incised bone tubeswere found at25CHl, but not at 14SC1.Work in shell occurred at both sites, but very sparingly.It should be obvious by this time that the cultural similaritiesbetween 14SC1 and 25CH1 are much stronger than are the differences.As we shall point out later, many of the items just listed for the twosites are by no means unique to them or to the Dismal River culturethey represent, but are shared by such other protohistoric Plainscomplexes as Great Bend Aspect and Lower Loup Focus. Some ofthem, like metapodial fleshers, bone paint applicators, and certain awltypes, occur also in late levels at Pecos and in later historic Plainshorizons. In the Dismal River sites, however, they are associatedwith more or less distinctive potterywares, decorated tubular claypipes, and roasting pits, and thus constitute a characteristic assemblagethat can be delimited in time and space. The propriety of allocating14SC1 to the Dismal River Aspect, as all students will recognize, isclear and indisputable.As to the period of the occupation represented by 14SC1, there arealso clear leads. The artifact assemblage in general immediately sug-gests to the experienced observer a protohistoric dating, as does thefact that White trade goods were present in very limited amounts.Our finding of Rio Grande pipe types in direct association with thelocal complex, and their linkage in respect to time of origin withTewa Polychrome pottery from the surface of the site, pomt, as Merahas noted, to a period aroimd the last two decades of the 17th and 466 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174the early years of the 18th century. Interestingly enough, among theceramic materials gathered earlier by Williston and Martin, otherSouthwestern pottery types have recently been identified by Tichy(Smith, 1949 a, p. 295), including "Tewa and Pojoaque Polychrome, 'Kapo' black ware, late red wares, and Eio Grande culinary wares.Taken as a group, these varieties indicate a late 17th to early 18thcentury date for the site." All this, of course, dovetails nicely withseveral dendrochronological dates from Dismal River sites in Ne-braska, including one of 1706 for 25CH1 (Hill and Metcalf, 1942,p. 205), another of 1709 for 25DN1 (Hill and Metcalf, 1942, p. 205),and a third of 1723 for 25HN37 (Gunnerson, 1959) . About the sameorder of time is suggested for the Dismal River stratum at Ash Hol-low Cave (Champe, 1946, pp. 27, 47), partly on dendrochronologicalevidence.The dating, on archeological and dendrochronological grounds, ofthe Scott County site circa 1700, and its allocation to the Dismal RiverAspect, in turn now generally identified with the Plains Apache orLipanans, has some ethnohistorical implications worth noting here.Neither the evidence assembled by Williston and Martin, nor that re-sulting from our more recent work, gives any indication that morethan one culture period or time level are represented at 14SC1.Rather, the data strongly support the inference that the pueblo ruinand its appurtenances (slab-lined hearths, grinding trough, oven,etc.), architecturally anomalous in the western Kansas plains, weredirectly associated with the Plains Indian material culture complexmanifested so strongly and consistently m all of our excavationsaround the ruin and on other nearby terraces along Ladder Creek.We have here, in short, the site of a Plains Apache conmiunity ofcirca A. D. 1700 that included a multiroomed stone structure, irriga-tion works, and other features clearly inspired by, if not the actualhandiwork of, Pueblo Indians. The relative scarcity of recognizablepuebloan traits in the artifact inventorj'^ generally may reflect eithera short stay by the pueblo Indians or else their ready adoption, hereor previously, of the everyday implements and utensils of the localresidents.This close association of puebloan with Plains Apache remains im-mediately recalls to mind certain late I7th and early 18th centurySpanish documents relating to Pueblo-Plains relationships. As wehave noted elsewhere, on at least two recorded occasions Indians fromTaos and Picuris forsook their villages on the Rio Grande and fledeastward into the buffalo plains. One such fugitive group that leftTaos circa 1664 was subsequently brought back by Archuleta, butfrom what point in the Plains is not certain. In another break in1696, a considerable party from Picuris eluded the pursuing Span- Wbdel] an introduction TO KANSAS ARCHEOLOGY 467iards and spent some ten years among the Plains Apache. These mal-contents were led by Don Lorenzo, a Picuris chief ; and when rescuedfrom the Cuartelejo Apache in 1706, at their own request, by Ulibarri,they numbered more than 60 persons. From the itinerary in Ulibar-ri 's diary of 1706, plus that in Valverde's report of his visit in 1719to the Cuartelejos, Thomas identifies the location of El Cuartelejo as "approximately in present Otero or Kiowa County, Colo. . . . prob-ably' [at] the junction of Mustang and Adobe creeks" (Thomas,1935, p. 264 n. 23). And this, if we follow the contemporary docu-ments, seems indeed to be the locality to which they lead us.Ulibarri's diary, in addition to the route traveled, records variousdetails regarding the land of the Cuartelejos and their manner ofliving. It is perhaps worth noting that when Ulibarri's rescue partyarrived at the "great settlement of Santa Domingo of El Cuartelejo,"Don Lorenzo and the rest of the Picuris came out of some "huts, orlittle houses;" others of the refugees, described as completely desti-tute, were scattered about among other Apache rancherias (Thomas,1935, p. 68). A detaclmient sent out to gather up these distantPueblos collected 18 persons from the Rancheria of Sanasesli, 40leagues distant (ibid., pp. 70-71). Thomas quotes from an 18th-century account of Ulibarri's journey, which says inter alia that theApaches of La Jicarilla conducted the expedition "to El Cuartelejo,where had been restored the houses which the fugitive Taos Indiansbuilt in the past century" (Thomas, 1935, p. 262 n. 6). Have we, inthese recurring remarks about the houses of the fugitives, a hint thatthey were utilizing habitations of more substantial or permanent na-ture than those of the Apaches?stone huts or small pueblos, perhaps ?Unfortunately for Thomas' case, there is no shred of archeologicalevidence, so far as I have been able to determine, for the former exist-ence of an Apache-Pueblo community at the locality in easternColorado which he identifies as El Cuartelejo. That such evidencemay yet come to light, I cheerfully concede; but as of this writing,there has been no independent confirmation of the "pueblo construc-tions, built by fugitives from New Mexico in the late seventeenthcentury [which] existed at El Cuartelejo in eastern Colorado."(Thomas, 1935, p. 268, n. 60). Nor, for that matter, have such con-structions yet come to light at any other point in the Plains north ofthe Arkansas River, except in northern Scott County, Kans. Andhere the archeological findings directly contradict Thomas' furtherassertion (ibid.) that the ruined pueblo is "of greater antiquity" thanthe documented flight of the Picuris into the Apacheria. Quite thecontrary ; the site falls neatly into the proper time interval, as Willis-ton (Martin, 1909, pp. 13, 18-22) argued half a century ago. 468 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174As I see it, then, the case for El Cuartelejo in eastern Colorado restssolely on the testimony of certain historical documents. That forCuartelejo in Scott County rests on archeological evidence, includingparticularly the unique association of a pueblo ruin with PlainsApache cultural remains. If Scott County pueblo and its associatedarcheological materials is not the very Cuartelejo rancheria fromwhich Ulibarri rescued Don Lorenzo and his Picuris compatriots (cf . Champe 1949 a, p. 291), then we must conclude that it was a simulta-neously occupied conmiunity (Sanasesli?) in which pueblo Indiansfrom the upper Rio Grande and Plains Apaches were residing to-gether in the late I7th and early 18th centuries.YOUNG BURIAL SITE (14SC2)So far as we were able to learn, no burials have ever been found inthe iimnediate vicinity of the pueblo ruin, nor are there any mounds orother special features that might suggest their presence nearby. Someyears before our investigations, however, in course of road construc-tion about the north gate of the park, one or more graves were gradedout. The drive that skirts the west shore of the lake here curvesaround the lower shoulder of a small but conspicuous butte knownlocally as Buzzard Knob, and leaves the park just beyond the spillway.On the left of the road, as it passes through the gate, is a sloping sandybench bearing a sparse stand of yucca, sage, prickly pear, and coarsebunchgrass, and overlooked by modest cliffs of calcareous material.A few notched arrowpoints and a scraper were collected from thisbench, but otherwise there was nothing to suggest that the spot hadever been of particular interest to the Indians. The burial area was onthe east slope of the Knob, a few yards outside the park, and about 1^miles, airline, north-northwest of the pueblo ruin.Our excavations began on a 40-foot front, at the top of the cut bankbordering the road, and were carried westward up the slope to amaximum distance of 35 feet. The area dug slightly exceeded 800square feet, within which was found evidence of at least five separateinterments. Except where disturbance by man could be detected thesoil profile presented a comparatively simple picture. The topmost4 to 8 inches were a light to moderately dark-gray sand, doubtlesscolored by decaying vegetation. Flint chips occurred in some nimibersin this zone, and there were also bits of charcoal. Below this was alight-yellow to nearly white clean sand, containing rounded pebblesand small calcareous rocks. Abandoned and filled rodent burrowstraversed this deposit in many places, introducing charcoal fragmentsand occasional small rocks to depths of 2 or 3 feet below the groundsurface. This sand continues to an undetermined depth, which inplaces certainly exceeds 6 feet; its lower levels showed little colorchange but were more compact. At the south edge of our excavations Wedel] an introduction TO KANSAS ARCHEOLOGY 469there were a few larger limestone fragments below topsoil, but at nopoint did we reach bedrock or any suggestion of it.Burial No. 1.?This lay near the north edge of the excavated area,6 or 7 feet back from the edge of the bank, under a layer of irregularboulders some of which protruded slightly above the ground surface(pi. 75, a). The bones were badly broken and scattered ; it is possiblethere were originally two bodies in the grave, but most of the remainsseem to have belonged to an adult female. The grave proper, about 22inches deep, appears to have measured approximately 48 inches northto south by 36 inches, but bone fragments had been scattered, or per-haps dragged by rodents, as much as 2 feet to the east. Most of theskull fragments lay at the south end of the presumed grave pit, thoughone large piece was at the opposite end. The small size of the pitargues against a full length interment, and I would suspect from thedistribution of recognizable bone fragments that the corpse had beenflexed with the head to the south. The feet were apparently doubledback against the buttocks with the knees toward the east, as if the bodyhad been laid, or perhaps had slumped over, on its right side. Bitsof red ocher were scattered about the grave; at the north end was apiece of tortoise shell; and there were two incomplete fresh-watermussel shells identified as Uniomerus tetralasmus (Say) in the pit fill.Burial No. ^.?Ten feet due south of burial No. 1, at a depth of 11inches, was an oblong area of mixed soil discolored by charcoal frag-ments. This measured 12 by 17 inches, with the long axis southwestto northeast. At the northeast end were a few small skull fragments,evidently those of an infant or very young child. At the southwestend, 8 to 10 inches from the skull pieces, and at a very slightly higherlevel were a number of toe bones since identified as those of "either aprairie wolf (Canis nuhilus), or of some large domesticated wolf-dogcross." Near the center of the mixed area was a fragmentary musselshell pendant or gorget, apparently of elliptical outline, with a par-tially broken-out perforation near the remaining end ; originally, theremay have been two perforations. Clustered over and above thispendant were some 30 or more tubular beads of bone, all badly weath-ered or etched by soil action. Eleven of these averaged 5 mm. indiameter by 10 to 16 mm. in length, were cylindrical to subtriangularin cross section ; and showed no signs of incising or other decorativemodification. Eighteen others had a diameter of about 8 mm., and alength of 11 to 27 mm. A few of these larger beads were somewhatbarrel shaped, i. e., had slightly tapered ends; one had 3 encirclinggrooves, another a single groove. There were two ringlike specimenswhose diameter exceeded their length.Burial No. 3.?The loosely flexed skeleton of another adult, probablya female, lay about 3 feet northwest of No. 2 and 7 or 8 feet southwest 470 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174 of No. 1. No grave pit could be defined, and the bones were in cleanwhite sand with little or no admixture. A few stones were notedabove the skeleton. The maximum diameter of the burial was 43inches, its depth 22 to 24 inches. The skeleton lay on its back, withthe right knee folded back under the bent left leg. The left arm wasslightly bent at the elbow, with the hand resting on or against the leftknee. The right arm was strongly flexed over the chest with thehand at the left side of the skull. The skull was at the east end of thegi'ave, turned slightly to rest on its left side against the right hand.No difficulty was encountered in determining the position of any ofthese features (pi. 75, &), because the bones were quite complete andwith few exceptions were in correct anatomical alinement. Unfortun-ately, however, they were exceedingly soft and crumbly, so that littlebut the skull, innominates, and long bones could be preserved for lab-oratory examination.There were several artifacts in direct contact with the skeleton. Onthe distal end of the right humerus lay the worked carapace of aterrapin, which fell apart at the sutures as it was taken up. I am notcertain whether this was a receptacle or part of a rattle or somethingelse altogether. Underneath, in a compact mass somewhat suggestingoriginal deposition in a bag or pouch, were 55 or 60 undecorated tub-ular bone beads, a stubby bone awl, a large stemmed projectile point,several scraperlike flints, and four spalls. A perforated mussel shelland a single tubular bone bead lay against the left wrist, and two orthree inches below the point where the left ulna touched the left, kneewas another large section of terrapin carapace. There is not theslightest reason to doubt that all these objects had been placed in thegrave at time of burial, and that they represent mortuary offerings.The beads are of two sizes, conforming to the two size groups(pi. 74, d) noted in burial No. 2. Forty-seven complete and 3 frag-mentary specimens ranged in diameter from 2.5 to 5.2 mm., and inlength from 7 to 11 mm. Four other whole beads averaged 10 to11 mm. in diameter by 22 to 26 mm. in length, and there were fragmentsof 2 or 3 additional specimens of about the same dimensions andproportions.The awl, made of some unidentifiable mammal bone, was slightlycurved, with rounded butt and a subcircular cross section. It was58 mm. long, with a maximum diameter of 7 mm. Like all bone fromthe site, it had been deeply corroded by ground moisture (pi. 74, a).Among the stone artifacts present, the projectile point and scrapersare of particular interest (pi. 74). The point, of gray chert, isshouldered, with a heavy stem expanding toward the straight base.It is 90 mm. long, 6 mm. thick, and has a maximum width across theshoulders of 24 mm. (fig. 87, a). The scrapers, seven in number, are WkdklI AN INTRODUCTION TO KANSAS ARCHEOLOGY 471 Figure 87.?Stemmed projectile point and split quartzite pebble scrapers from burial 3,Young site 14SC2. Actual size. short, wide, and thick, planoconvex in cross section, with a high dorsalridge (fig. 87 &, c) . This description, of course, would apply to count-less thousands of chipped scrapers from sites all over the Plains area,but the present specimens differ in at least two particulars : first, theyare proportionately much shorter and thicker than the common Plainstype, and secondly they seem to have been fashioned from small splitpebbles or nodules with no subsequent modification other than a slightflaking at the thick work end. They probably would stand out as adistinct group in any collection of typical Plains end scrapers, thoughI suspect they were used in the same manner as the latter. The wholespecimens show the following dimension ranges : 30 to 44 mm. long,25 to 34 mm. wide, and 9 to 18 mm. thick. Of the four spalls fromburial No. 3, one has finely retouched edges and may represent a knifeor side scraper; the others suggest rejectage or unworked knife orscraper blanks.The mussel shells from the grave, three in number, represent localfresh-water species, including Ligumia subrostrata (Say) and Lamp-silis anodontoides (Lea). 472 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174Burial No. Jf.?Twenty inches southeast of burial No. 1, at a depthof 18 inches, was a discolored area measuring 12 by 22 inches, with thelong axis running southwest to northeast. There were only a fewtraces of bone in the grave, evidently that of another infant; theorientation and arrangement of the body could not be determined.Accompanying the interment were two broken fresh-water musselshells, one possibly perforated, and both identified as JJniomerustetraldsmus (Say).Burial No. 5.?About 5 feet southwest of burial No. 3, in line withthe latter and No. 1, was the grave of an adult female. Much thedeepest of the group, the skeleton lay on clean white sand 42 inchesunderground (pi. 75, c) . The overlying grave fill included the follow-ing soil strata, from the top down: 4 inches of dark-gray hmnus;12 inches of closely packed limestone fragments intermixed with sandand bits of charcoal; 6 inches of dark sandy soil; 8 inches of fineclean white sand; 6 inches of dark sand containing some charcoal;4 to 6 inches of gray charcoal-mixed sand in which were the bones.The skeleton was closely flexed, lying on its left side with the head atthe easterly end of the grave. The arms were folded so that the handswere together, forming a sort of pillow in front of, and partly under,the face. The legs were tightly flexed, with the feet against thebuttocks and the knees toward the left within a few inches of theelbows. As in the case of burial No. 3, so here the bones were verysoft ; even with the application of a fixative, it was possible to salvageonly the skull, some limb bones, and a few other specimens.Three artifacts were taken from the grave. All lay in a tightcluster just below the left knee, against the shaft of the tibia. Onewas a chert drill (pi. 74, &), the shaft of which widened abruptly toa straight base ; it was 45 mm. long, 18 mm. in maximum width, andthe square shaft had a diameter of approximately 7 mm. A sub-elliptical or somewhat leaf-shaped knife of gray porphyry measured74 by 35 by 12 mm. The third specimen was a fragment from thesmall end of a rude scraper made, like those found with burial No. 3,by splitting a quartzitic pebble or nodule.Other -finds.?During the general excavations in the burial areaa few stray artifacts and occasional human teeth or bone fragmentswere turned up which could not be ascribed to any of the gravesnoted. Twelve inches west of burial No. 4, at 16 inches depth, wasa small barbed arrowpoint, from which a stem had apparently beenbroken. Not far from burial No. 3, but probably not associated withit, were a triangular concave-base arrowpoint measuring 40 by 26mm., and a gray quartzite end scraper. Between burials No. 2 andNo. 5 were two rude side scrapers, of chalcedony and buff chert; asubtriangular chert blade 22 by 35 mm.; and an unfinished tubular Wbdbl] an introduction TO KANSAS ARCHEOLOGY 473bone bead 7 by 16 mm. Scattered about in the sand were numerousflakes of chert, quartzite, moss agate, plagioclase basalt, and othermaterials. There were no pottery fragments, though a thick verticalrimsherd, made of fine black paste containing coarse silicious inclu-sions, and bearing partially obliterated cord roughening, was foundon the bench some 70 or 80 yards south of the burials. There is noproof, of course, that this sherd had any connection whatever withthe individuals buried on the slope of the Knob.Extended test trenches and pits elsewhere on the slope of BuzzardsKnob yielded only negative evidence as regards additional graves,and it seems probable that the site has been exhausted. Concerningthe graves graded out prior to our work, we learned no details beyondthe statement by an eyewitness that they were stone-covered and con-tained no pottery. Generalizations regarding the site and inferencesas to its significance in the prehistory of the locality must thereforebe based on a lamentably small and incomplete set of data.It seems safe to infer that burial in the flesh was practiced, sincethe bones of Nos. 3 and 5 certainly, and those of No. 1 probably, hadbeen deposited in articulation and not secondarily as dismemberedparts. None of the remains showed any evidence of exposure tocremation fires. In the case of the adults, at least, more or less com-plete flexion of the corpse was indicated. With the exception ofNo. 1, which had the head apparently at the south end, the gravesseem to have been oriented with the long axis southwest to northeastand with the skull or skull fragments most commonly at the northeastend. There can be no doubt that interment was in dug pits ; bouldersplaced in the upper part of the fill were apparently optional. Ifunworked, or doubtfully worked, mussel shells be included, mortuaryofferings were present in all observed burials, though not in strikingquantity.The probable chronological position and cultural affiliations of theYoung site burials are not readily apparent. Culturally diagnosticartifact associations are virtually absent, and this precludes allocationto known complexes of the region. It should be emphasized here,however, that neither the skeletal remains nor the accompanying arti-facts suggest a direct connection between the burials and the peoplewho lived in and about the pueblo. The stone artifacts from thegraves, for example, including the large stemmed projectile point,the split pebble scrapers, and the drill, are markedly dissimilar to theanalogous forms found at the stone ruin. The patinated and muchcorroded bone beads do not resemble as a group the characteristicforms from about the pueblo. The available crania in no way suggestpuebloid types (see Stewart, p. 679). There is thus no reason, soma-tological or cultural, to ascribe the interments to a puebloan group. 474 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174 early or late, or to a nonpuebloan group of such recent date as to havebeen in contact with white men. On the contrary, they probablyantedate the settlement of and about the pueblo by several or manycenturies, and are to be correlated with a much earlier occupancy ofthe locality by a people with Woodland affinities. Whether theyequate chronologically with such Middle Woodland complexes of theCentral Plains as the Woodruff ossuary (Keith Focus) or Kansas CityHopewellian, or alternatively precede these and betoken an EarlyWoodland or older occupancy, I am not in position to say.Our suspicion that the burials represented a horizon earlier thanthat manifested about the monument led to much surface hunting andtest pitting of all likely village locations in the vicinity. Members ofour party explored the extensive terraces on both banks of the creekbelow the dam, the tops and slopes of several mesas and other outliers,and the two larger tributary canyons?Timber and Bull?which enterthe main valley from the west and east, respectively, above and op-posite the burial site. Flint chips and bits of charcoal occurred innumerous spots in the topsoil, but nowhere was there sufficient con-centration or other corroborative evidence to justify a large-scaleexcavation. Several shallow overhangs or rock shelters were found,occasionally with charred grass, wood, and similar material mixedwith the sandy floor deposits, but as might be expected from their smallsize none showed satisfactory proof of human habitation. The arti-facts collected during this surface work consisted principally ofchipped stonework?end and side scrapers, crude knives, broken pro-jectile points, miscellaneous spalls, and occasional fragments oflarge heavy cutting or chopping tools of pale yellow chert. The fewsherds found were small and so badly weathered as to be of little com-parative value. Several were calcite tempered, and may pertain to aWoodland occupancy much more strongly represented on Salt Creek,13 miles east of the park; others suggest the sand-tempered UpperKepublican ware also found to the east and north. Most of the finds,however, did not differ greatly from material found in the vicinity ofthe "El Cuartelejo" monument; there was nothing obviously anddistinctively of very ancient type nor, conversely, did we find anythingclearly attributable to known tribes of the historic period. Theubiquitous occurrence of worked flints, tools, and chips, even atopdifficult buttes a half mile and more from water, does not suggest asettled occupancy. Rather it would appear that small parties ofhunters camped briefly and carried on certain of their activities onevery terrace, butte, and mesa, and in every canyon, in the vicinity. Ofadditional puebloan structures rumored to exist in and above the parkwe could learn nothing concrete. What may lie beneath the waters ofthe lake, particularly along the submerged bank of the former creek Wbdbl] an I]SrrRODUC'nON to KANSAS ARCHEOLOGY 475just above the dam, there is now no way of ascertaining, but it is quiteprobable that several terraces well suited to aboriginal habitation andpossibly utilized by Woodland and other prehistoric groups have beenflooded and thereby made inaccessible for investigation.SURFACE FINDS NEAR SHALLOW WATER, SCOTT COUNTYIn central and southern Scott County and northern Finney Countyis a sizable area of generally flat smooth land from which there is nosurface drainage. Today the district is known as the "shallow-waterbasin" because water is generally obtainable at comparatively shallowdepths by drilling. To primitive man, obviously, the ground waterwas useless ; the important characteristic of the area was the presenceof numerous shallow basins that in times of heavy precipitation weretemporarily transformed into ponds and water holes. Several shortintermittent creeks from the west terminate in this locality ; the larg-est, Whitewoman Creek, ends in a large depression south of Scott Citywhere its flood waters occasionally create a shallow lake of some size.The reports of local collectors, substantiated by a very limited surfacereconnaissance on our part, indicate that at times in the past the "shallow-water" area was perhaps not so uninviting as the casualtraveler in a dry summer today might suppose.One of the best known spots hereabouts for hunting "relics" is Pony(or Pawnee?) Mound (14SC3), a natural eminence 6i/^ miles southand 41/^ miles east of the town of Shallow Water, and about a milenorth of the Scott-Finney County line. Immediately west of themound are several dry depressions or playas. During the dust stormsof the "dry thirties" as much as 12 inches of soil is said to have beenstripped from the mound surface, thereby exposing fire-blackenedhearth (?) areas and concentrating among them great quantities ofanimal bones, burned stones, chipped flints, spalls, and flakes, lessernumbers of potsherds, occasional large blades, and a few glass beads.At the south end of the mound is a depression some 15 or 18 yards indiameter, locally reputed to have been used as a "fort" by soldiers. Asmall test hole dug in the center revealed only a thin black line of silt,such as might have been deposited by wind in standing water, withclean reddish sand below ; the origin and purpose of the basin, if man-made, are problematical. Other tests in the hearth (?) areas showedonly a gradual thinning out of mixed and discolored earth, with nosigns of disturbance by man below 12 inches depth or in the surround-ing soil. Half a mile east of Pony Mound more burned spots and otherremains were noted, but here also there was a notable absence ofsubsurface evidence. Other similar localities are known east of Shal-low Water. 476 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174Our small artifact collection from Pony Mound is of interest pri-marily because of the variety of types suggested. Among the sherds,all of which are small and much weathered, there are half a dozenthick "hole-tempered" pieces with cord-roughened exteriors?un-doubtedly a Woodland ware from which angular particles of calciteaplastic have been leached. More plentiful are thinner harder sherds,also cord roughened but with rounded gravel or sand inclusions, thatare probably of Upper Republican origin. Reminiscent of potteryfrom the Blue River district of northeastern Kansas is a collaredcrisscross incised rimsherd with slate-gray core and thin buff-orangeinner and outer surfaces, the latter cord roughened. Another smallthin recurved rim, finely tempered with sand and containing micaparticles, closely parallels sherds from the Dismal River middens atthe Scott County pueblo ruin (14SC1). Projectile points includestem fragments of heavy corner-notched pieces, but small triangularnotched and unnotched specimens are more common. Broken knives,scrapers, drills, and other materials are of less diagnostic value. Theglass beads include small blue and white forms, as well as larger redones with white core. No extremely old forms are represented in ourmaterial, though there are reports of Folsom or Folsom-like bladeshaving been picked up in the district.Inadequate as our sample is, it clearly attests the presence of a suc-cession of peoples in the locality?none with established residenceand all, no doubt, here to hunt game attracted to the transient waterholes. If the several sherd types noted abov^e are correctly identified,it may be inferred that Indians residing usually farther to the eastmade occasional or periodic excursions to this westerly locality ; muchof the stonework may have been left by nonpottery groups who sincetime immemorial roamed the western Plains to gain the major partof their livelihood from the bison. The shallow-water basin was, inshort, a hunting area, and gives little or no promise of revealing, insitu, stratified camp or village sequences such as may be expected inthe better watered creek valleys to the north and east.ABORIGINAL CHERT QUARRIESCOWLEY COUNTY QUARRIESSo far as I have been able to learn, the only quarries in Kansasthat can be certainly attributed to aboriginal chert gatherers, are inthe Flint Hills upland. The largest and best developed are in south-ern Cowley County, near Maple City (14C05). Here, and southwardinto Kay County, Okla., are extensive outcrops of a thick-beddedPermian limestone interbedded with nodular cliert. The nodulesrange in thickness up to 6 or 8 inches. Freshly broken faces have a Wedel] an introduction TO KANSAS ARCHEOLOGY 477bluish-gray color, but on weathering this becomes tan, reddish, orrusty brown. Some of the nodules show a characteristic banded struc-ture that on a fractured surface gives a very pleasing effect compara-ble to the grain in a fine piece of cabinet wood. The chert, moreover,though tough and serviceable, is tractable and easily worked. Thatit was highly esteemed by the Indians is attested by the finding ofartifacts made of this material in Rice County, more than 120 milesnorthwest of Maple City.A brief examination of the Maple City quarries was made in com-pany of Bert Moore, on August 8, 1940. Immediately southwest ofthe town, along the edge of a grassy hill overlooking a small intermit-tent headwater tributary of Little Beaver Creek, were numerous shal-low partially filled-in depressions. They were irregular in outline,from 6 to 10 feet wide, 10 to 60 feet long, and from a few inches tonearly 2 feet in depth. Scattered about the depressions were lime-stone boulders and quantities of whole and broken chert nodules.Most of the latter were banded, of dun, gray, or brown color, andmany contained an abundance of Fusidina fossils. We found no re-jects or quarrj^ blanks, but were told that many had been gatheredup in past years. Six or seven nodule fragments were collected fromabout the depressions, and a nearby resident, W. H. Uteley, presentedus with two additional specimens picked up on the site. One of theseis a large shaped elliptical blank 16.5 by 10 by 3 cm. in size.Our chert samples from Maple City, together with nine representa-tive chipped artifacts excavated from Indian mounds in Cowley andRice Counties, were submitted to L, G. Henbest, United States Geo-logical Survey, for examination. Mr. Henbest reports that the twoUteley samples contain "a few specimens of ParaschivageriTia Jcansa-sensis (Beede and Knicker). This is characteristic of the Florenceflint but may not be restricted to that horizon." Some of the nodulefragments picked up by myself "contain Paraschwagerina aff.kansasensis? (Beede and Knicker), and one or two other species ofLower Permian fusulinids that, as exposed, are difficult to identify,whether Schwagerina^ Paraschwagerina^ or PseudaschwagerinaP Asfor the artifacts, specimens from both counties ? contain an assemblage that resembles that in 493 [my quarry samples] in severaldetails. On the basis of a superficial study I would say that they are the same.Although the range of the Paraschivagerina af?. kansasensis (Beede and Knicker)and the other Lower Permian Schwagerininae in this collection may possiblybegin in the Neva and extend to the Florence, they may safely be regarded asFlorence.It should be noted that the artifacts examined included specimens ofunhanded as well as of banded chert, this particular feature lackingthe diagnostic value which the enclosed fossils have. 478 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174I know of no systematic explorations in any of the quarry pits inCowley County or in those farther south. It may be of some interestto note the observations made by Gould (1898 a, pp. 78-79; 1899) eventhough these relate primarily to the workings in Oklahoma. In hisfirst report (1898 a, pp. 78-79) he states that ? About 3 miles south of the territory line and 8 miles south of Maple City,Kansas, are situated several hills on which are located what are known locallyas "The Timbered Mounds." They are situated some half mile east of thejunction of Myers Creek and Little Beaver.The hills in all the region consist of massive ledges of limestone, containingmuch flint and alternating strata of gray and drab shale. Near the base ofthe hill the prominent ledge on both sides of the creek is the Strong flint ofProsser, beneath which is a ledge of massive sandstone. The ledge cappingthe hills is the Fort Riley or Florence flint. It is on the last-named ledge thatthe "Mounds" are located. ... On a crescent-shaped ridge about a half a mile long and from 50 to 150feet wide the hard but brittle limestone has been quarried in great quantities,and has apparently been piled up in the form of rude edifices. . . . The stoneswhich compose these buildings seem to have been broken out of the ledge atintervals and without regularity of size. In shape they are flat, not more than6 inches thick, and usually longer than broad. There are none that a strongman cannot lift.In certain areas of perhaps half an acre the loose rocks cover the ground toa depth of three to four feet. , . .The edifices which have fallen down appear to have been either square orcircular, with a ground diameter of 5 to 15 feet. In several places can benoticed the faint outlines of structure, but usually the rocks are piled in shape-less heaps. One peculiarity is that so far as noticed all the buildings seem tohave fallen toward the center as though they had sloped inward like anEsquimaux hut. Occasionally there will be a space in the center not coveredwith rock as if the wall was not high enough to reach the center when it fell. . . .No marks of tools have been discovered on the rocks ; but in some places thereare traces of fire. . . . Besides the hill described there are at least four others,within a radius of three miles, covered with the same kind of ruins, and otherhills with traces of the same peculiarities have been found in the state nearMaple City.To S. W. Williston is credited the suggestion finally accepted byGould that ? The ruins mark the sites of ancient flint quarries. The ledge on the top of thehill contains many flint nodules, sometimes nearly as large as a man's head, andamong the loose rock these nodules are conspicuous for their absence, althoughmany flakes of flint as large as one's hand may be found. This ledge is thefarthest west of any flint-bearing ledge in the region ; and probably the plainstribes from the west obtained their arrowheads and flint implements from thislocality. This is further substantiated by the fact that the flint implementsfound in the prehistoric mounds at Arkansas City, some 20 miles northwest,contain fossil Fusulina cylindrica, which are characteristic of the flint mentioned.The edifices described above were probably temporary structures used by theworkmen while engaged in quarrying. Wedkl] an introduction TO KANSAS ARCHEOLOGY 479A second note in the year following (Gould, 1899, p. 282) recordsthe fact that ? . . . more than 100 imperfect implements, or rejects, Lave been found. Theyvary in size from 3 to 8 inches in length, and from IV2 to 4 inches in breadth, andweigh from 3 ounces to IV2 pounds. In shape they are usually oval and twiceas long as broad, ends roundish or pointed, with usually a cutting edge chippedon all sides. They are nearly always broken. . . . One or two specimens arenearly perfect, . . . . . . The first locality is south of Myers creek and east of Little Beaver, inthe Kaw reservation from 6 to 12 miles nearly south of Maple City, Kansas. Itis here the quarries were first studied, and here most of the excavations seem tohave been made; but singularly enough very few rejects have been found here.The second locality is in and around Maple City. Most of the rejects have beenfound on the farms of Mr. H. Ferguson and Mrs. George Sutton. Some interestinglocalities are found 3 to 5 miles north of Maple City.On our hurried visit to the "Timbered Mounds" no evidences ofstructures were noted; undoubtedly the remains so termed by Gouldare nothing more than piles of stone heaped up by the Indians in theirsearch for suitable chert. The workings are much more intensivethan those near Maple City?the pits are larger and deeper, up to 30feet across, and as much as 3 or 4 feet deep. In and about the pits arelarge heaps of limestone, among which we found surprisingly fewchert fragments. A road cut just west of the quarries revealed a layerof nodular chert 2 to 3 feet underground, and overlain by limestone.This, I suspect, was the stratum sought by the Indian workmen. Inwhat little chert we found among the diggings, very few pieces showedthe banding so typical of many of the Rice and Cowley County arti-facts, and almost none of our samples contained Fusvlina fossils.Gould's correlation of the "Timbered Hills" chert with Arkansas Citymound artifacts, on the ground that both contained fossil Fusulinacylindrica was probably based on accurate observation but in light ofsubsequent terminological revisions it is outdated and nomenclatorial-ly inaccurate. The species F. cylindrica^ or its closest relatives inAmerica, characterize the upper part of the Des Moines Series in theMiddle Pennsylvanian, whereas the horizon represented in the KayCounty quarries is lower Permian. The nearest possible source ofrelatives of F. cylindrica is 70 to 75 miles east of Maple City andHardy.From the quantities of debris seen at these quarries, it must be in-ferred that they were once quite extensively worked. At what periodthis activity began, by whom it was carried on, and when it finallyended, are intriguing questions for which I have as yet only partialanswers. My excavations and observations since 1930 in various sec-tions of the Nebraska-Kansas-westem Missouri region suggest thatartifacts and rejectage probably traceable to these quarries occur ingreatest abimdance in the protohistoric sites in Rice, Cowley, and484172?59 32 480 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174nearby counties along the Arkansas in central and south-central Kan-sas. Elsewhere, I have seen or collected specimens from the Fanningand Doniphan sites, in Doniphan County, and from sites near Neode-sha, Marion, Salina, and Pratt. The Salina and Pratt sites, brieflydescribed elsewhere in this paper, are pre-White contact, probably 15thcentury or earlier ; all the others have yielded White contact material.The blue-gray chert characteristically found in sites about Manhattanand elsewhere in the Kansas River drainage, especially in the prehis-toric sites, is apparently of about the same geological period as theforaminiferous chert from Maple City, but presumably is of localorigin.It may be suggested tentatively, then, that so far as semisedentaryhorticultural groups in the Central Plains are concerned, the MapleCity material was used principally during the 16th, I7th, and early18th centuries, especially by peoples in central and southern Kansas ; that it was probably used in limited quantities as early as the 15thcentury, and possibly before; and that in small quantities it wastraded or carried as far as the Fanning site. I have no informationregarding either absence or presence of the material south of certainearly historic Kay County sites in Oklahoma, or to the east or west.Possibly some of the quarrying vras done by nomadic hunting peo-ples whose traces still elude the archeologist. Perhaps, too, publica-tion of full data on more southerly sites will shed further light onthis matter. CHASE COUNTY QUARRIESMuch less conspicuous than the Maple City-Hardy quarries areother workings in central Chase County, some 90 miles almost duenorth of the previously described operations. They are situated on alofty tableland locally known as "Flattop," approximately 4 milessouth of Elmdale and 2 miles southeast of the Cottonwood River.Here there are traces of two small areas of ancient diggings, somesix or seven hundred yards apart. Both lie just within the south edgeof section 11, T20S, R7E.The westernmost of these two areas (14CS2) lies at the immediateedge of the tableland, overlooking a long westward slope in whichthe ground falls some three hundred feet to the Cottonwood valley.Here there is a thin soil cover over a limestone caprock. In this soilcover may be discerned faint evidences of old disturbances. Theseare shallow irregular grass-grown depressions not exceeding 12 inchesin depth and up to 10 or 12 feet wide. At one point, these depressionsform a sort of horseshoe opening to the west, that is, at the edge ofthe tableland where the soil covering the cap stone would have beenthinnest or absent. Inside this horseshoe a slight elevation may berecognized, and this suggests detritus cleared from the pits and piled Wedel] an introduction TO KANSAS ARCHEOLOGY 481in an area that had probably already been stripped. The entire situa-tion looks like the result of shallow digging, pushed from the edgeof the table eastward. Chert fragments, often fractured and some-times clearly shaped into elliptical quarry blanks, are scattered overthe grassy surface of the disturbed area. To the south, there appearto be other smaller workings, similarly marked by shallow irregulardepressions in the grassy terrain, and littered with fractured chert.Here and there may also be seen large limestone blocks and fragments,most of them partly buried in the sod.The eastern area (14CS3) lies on the slope of a dry swale drainingsouthward into a small tributary of Rock Creek. Here again there isa thin earth mantle overlying ledges of massive white limestone. Thediggings here seem to be linear, as though the workmen were followingsome specific ledge or vein ; and they tend to run at right angles to theslope of the land. Numerous large blocks of limestone have beenturned up, and among them, scattered all over the diggings, are num-erous fragments of chert, mostly cores and fractured blocks, butoccasionally with shaped quarry blanks. At one point, in the dirtthrown out by a large burrowing animal, many small chips from theprimary flaking of quarry blanks were found. For the most part, thechert found here, as in the western diggings, was whitish-gray ratherthan the blue-gray material so characteristic of chipped artifacts fromvillage and campsites in and adjacent to the Flint Hills.Since my only visit to these quarries was made during a vacationperiod, no excavation was possible, and so I am unable to describe thesubsurface character of the pits. They are shallower, smaller, andmuch better grassed-over than are those in the Maple City-Hardy lo-cale, and give an impression of being considerably older. That theyactually represent aboriginal quarries cannot be proved at the moment ; but no other satisfactory explanation comes to mind. Local residentsinsist that they are not a result of early explorations for limestone forbuilding or fencing purposes, nor are they in the sort of location whereroad materials would be sought. The presence of shaped quarryblanks, similar to those encountered on Indian sites in the region, andthe manner of occurrence of the chert?as nodules in limestoneledges?seem like strong arguments in favor of aboriginal quarryingoperations. There are no village or campsites within several miles,nor are there any farmsteads in the immediate vicinity.That similar workings occur elsewhere in the Flint Hills region isvery probable. The chert nodules occur at several levels throughoutthe limestone of which the Hills are mainly composed, and whereverthe ledges have been exposed by valley erosion, the chert was madeaccessible to native man with the expenditure of little effort. Sincemuch of the Flint Hills region is primarily grazing, and since work- 482 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174ings such as those described in Chase County are relatively inconspic-uous to any but the closest scrutiny, it is entirely likely that a numberof similar features still await discovery and reporting.PETROGLYPHSThe surface geology of Kansas is not conducive, for the most part,to the high development of pictorial art on stone, or to the long preser-vation of such art as may once have existed. Over much the greaterpart of the State there are no ledges, boulders, or other smooth-surfaced outcrops of stone suitable for the delineation of designs bypecking, grinding, or rubbing; and many of the exposures that maywell have invited the aboriginal artists to try their hands, consist ofsoft stone with little resistance to weathering by wind-borne dust andsand, by rain, and by frost. It is possible, of course, that a determinedand sustained effort to locate rock carvings would reveal many moreoccurrences than are recorded here, though I doubt that many groupsof any size or complexity have escaped detection. The occurrencesbriefly noted in the following pages are necessarily incomplete, sincethey represent sample data gathered more or less incidentally in courseof village site explorations and are not the fruits of an intensive searchdirected primarily at petroglyphs.With two observed exceptions, and one or two others briefly recordedin the literature, petroglyphs appear to be pretty much restricted to asmall area in the central part of the State north of the Arkansas River.Here their occurrence is correlated with the presence of scattered out-crops of Dakota sandstone, a friable brown to reddish-brown materialwhose vertically exposed faces offered an easily worked surface. Out-side the Dakota sandstone belt, in the soft shaly and calcareous forma-tions to the west and in the limestones to the east, petroglyphs areapparently very scarce and inconspicuous.So far as my information goes, all Kansas petroglyphs are carvedinto the rock surface. Presumably the designs were first lightly out-lined by scratching and the desired lines were then deepened andwidened by grinding or rubbing with a thin-edged stone tool. In therelatively soft sandstone of the region, even the most involved figurescould be made in a relatively short time by this process. I recall nopetroglyph that could not have been made thus. We saw and heardof no paintings, nor would I expect that designs in color had theyever been made would have lasted more than a few years in any of thelocations we examined.It may be stated at the outset that there is no way of datingaccurately the petroglyphs recorded below, or of determining by whattribes or for what purposes they were made. Representations of thehorse, or of momited riders, occur in a few places, and these obviously Wbdhl] an rNTRODUCrrON TO KANSAS ARCHEOLOGY 483date from a time after the introduction of that animal, i. e., within thelast two or three centuries. For the great majority of the figures,however, there is no such fixed maximum date line, and such criteriaas the degree of weathering, relative crudeness, or superposition ofone design over an older are manifestly unreliable. It is a safe guessthat not all the figures were made at one time or by the same tribe orcultural group. Unhappily, in no case do the rock designs resemblethose on known pottery types or other native remains, and there isnot the slightest bit of evidence linking any of the carvings with any ofthe archeological horizons of the area.As to the motives that prompted the carvings we can only speculate.Life forms, both animal and human, unreal creatures of various kinds,and a variety of simple geometric and rectilinear designs occur. Someof the latter may have been purposeless scratchings. Some of theunrealistic creatures may have had ceremonial or religious signifi-cance, and for this or some other reason were inscribed on rockwalls near springs or watercourses. Horses and horsemen, and per-haps the linked human figures, may have been memorials or remindersof some otherwise unrecorded event. Occasional simple figures occurin such obscure and solitary places that one wonders if they mighthave been made by lonely youths seeking a vision, or perhaps bymembers of a hunting or raiding party temporarily immobilized asby proximity of an enemy or other menace, real or imaginary.In the following pages I have listed and briefly described the petro-glyph sites for which I have direct, or reliable indirect, evidence.With exception of the first, to which access was denied by the owner,I have visited all of the localities personally and obtained sketches,photographs, or both, of the drawings.Site No. 1. Smohy Hill River, Ellsworth County {IJfEWl ) .?Thisis probably the outstanding petroglyph site in Kansas, as it is also oneof the very few that has found its way into the literature. Under thedesignation "Inscription Rock," it was briefly noted, and part of thecarvings illustrated, by Miller (1869, p. 383 and pis. VII and VIII).Mallery (1893, p. 80 and fig. 44) reproduced tracings from Miller'sdrawings but on a greatly reduced scale. Though I was not able toinvestigate the spot myself, G. L. Whiteford of Salina has generouslysupplied me with 18 photographs showing outstanding portions ofthe gallery. From these photographs I have traced a number of thepetroglyphs, not, however, to scale or in their correct position relativeto each other (fig. 88).The locality is on the left bank of Smoky Hill River approximately15 miles southeast of the city of Ellsworth. I have been informed byWilliam O. Leuty, United States Engineers Department (letters ofMarch 10 and 19, 1942, and accompanying map) that village sites 484 BUREAU OF AMERICAN ETHNOLOGY [BULL. 174 Figure 88.?Petroglyphs at Inscription Rock, 14EW1, on Smoky Hill River, IS milessoutheast of Ellsworth. (Not to scale or in correct relative position.) Wedbl] an introduction TO KANSAS ARCHEOLOGY 485and rock cairns occur at a number of points along this portion of theSmoky Hill, but their age and cultural affiliations, as also their rela-tionship, if any, to the petroglyphs, remain unknown (see also Smith,1949 a, p. 293).The "Inscription Kock" or "Indian Hill" petroglyphs include awide range of figures, in all stages of preservation and unquestion-ably of varied authorship. There is, of course, the usual profusion ofrecent dates, names, and initials left by picnic parties, sightseers, andothers. Five-pointed stars, an American flag, certain tentlike andboatlike designs, a human figure w^earing a wide-brimmed flat-crowned hat, an occasional horse, perhaps a Maltese cross in a circle,and a few other carvings are probably or possibly non-Indian, thoughthey may date from well back in the last century. The great majorityof the petroglyphs, however, are undoubtedly Indian, even thoughmany are so deeply marked, so fresh looking, or otherwise so out ofthe ordmary that one suspects modem faking or retouching.Human and animal representations occur in profusion. Theformer vary greatly in size and style; there are triangular, rectan-gular, and hourglass-shaped bodies, most of which give no detailsregarding dress or ornamentation. Heads are commonly outlined,with facial features shown, and have a pair of curved bisonlike hornsor other appendages, perhaps representing a dance headdress ; or theymay be nothing more than a small drilled pit surrounded by shortrays. Some of the figures have plumelike designs pendent at theelbow from partly outspread arms. One or two seem to be carryinga lance or ceremonial staff, but there are few representations of thebow and arrow or of other recognizable weapons. Animals includechiefly quadrupeds, apparently, but identification is not easy. Thereis an Indian mounted on horseback, with headdress trailing behind,and one or two figures suggest grazing cattle or bison. What may bean elk or deer with upraised head is also shown. Birds, serpents,water animals, insects, and imaginary monsters seem to be almostabsent from the panels shown in my photographs (pi. 76, 77, a).Geometric figures, though probably plentiful, cannot be easily dis-tinguished in my views. There are several ladderlike designs, somesingle, others double and flanked with zigzag lines; squares contain-ing an X; concentric triangles; circles or small drilled pits withradiating lines; multicelled oblongs with a pit in each cell; "deer-track" symbols, consisting of two elliptical hollows separated by anarrow vertical ridge; vertically bisected circles; bird tracks; andmiscellaneous other patterns. No doubt a painstaking examinationof the bluff face under varying conditions of light would reveal manyadditional figures, both geometrical and otherwise. 486 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174The very crowded nature of much of the petroglyph surface and thefrequent superimposing of one figure over another is suflScient evidencethat different periods of time are represented. In places, much of thecarved surface has been dimmed by weathering, but whether this iscorrelated with greater relative age or is due to a more exposed locationI cannot say. It is to be hoped that a more extended examination ofthe site can be undertaken before its scientific worth has been destroyedby time and man.Site No. 2. Elm Creek^ Ellsworth County (UEWH ) . ?Two weath-ered petroglyphs were seen on the west side of Elm Creek 214 milesdue south of Carneiro. They are on an east-facing sandstone ledgeperhaps 30 feet above the stream, which carries a small but steady flowof clear water, and is fringed with burr oak, ash, walnut, elderberries,gooseberries, chokecherries, etc. On the flats across the creek thereare said to be circular depressions possibly representing pit-house ruins.The valley would offer unexcelled opportunities for camping in con-cealment by hunting or raiding parties, and may well have had apermanent village. The petroglyphs are anthropomorphic (fig. 89,c, d) ; each has the arms upraised, and one wears a horn at the side ofthe head. They are 8 or 10 feet apart. Nearby may be found a fewrectilinear scratches, but the origin is uncertain.Site No. 3. Cave Hollow farm., Ellsworth County.?This is a locallypopular picnic spot about 41/^ miles north by east of Carneiro, in Sec-tion 29, T14S, R6W. A steep-sided sandstone promontory on the northside of a small creek valley is pierced near its terminus by a curvingtunnel about 15 yards long that opens at each end onto the creekvalley. At the base of the promontory, perhaps 25 feet below thetunnel, the creek forms a small pond, augmented by a large seep justbelow the level of the tunnel and between its openings. The openingsface south and southwest. On the ceiling and about the entrances area number of petroglyphs. Recent vandalism and exfoliation of thestone has obscured or destroyed many others.A few of the less extensively damaged carvings from the tunnelceiling are shown in figure 89. They include a representation of ahorse carrying an unidentified object on its back, several anthropo-morphic figures, tridents, or bird tracks, and vestigial zigzag lines inopposed pairs.On the south-facing wall between the tunnel openings and abovethe seep are traces of other more elaborate carvings. They are badlyweathered and exceedingly difficult to reconstruct accurately. Theprincipal figure appears to be a monster several feet long with broadoval head from which two parallel lines run zigzag along its back.Within the zigzag lines are triangular elements ; below, are closely setslanting lines. No feet or legs are discernible in the photographs I WkdkM an ENTRODUCTION to KANSAS ARCHEOLOGY 487 /'iif^K-'^n^ ^ Figure 89.?Petroglyphs from localities 2 (c, d), 3 (Cave Hollow Farm), 4 (a, b),Ellsworth County. 488 BUKEAU OF AMERICAN ETHNOLOGY [Bdll. 174have, and the creature cannot be entirely disentangled from severalother smaller and possibly later zoomorphic figures. Nearby is a fig-ure about 3 feet high consisting of vertical lines crossed by shorthorizontal lines in ladder fashion. Bird tracks and vertically bisectedcircles may also be seen.Site No. J).. Ash Creek^ Ellsworth Coy/nty.?Eight miles south ofEllsworth and 7 miles north of the Rice County line State Highway14 crosses a small dry canyon draining eastward into Ash Creek.Ledges of soft yellow-brown sandstone outcrop here and therethroughout the canyon, but only a few were explored. One, about200 yards west of the highway (NEi^, Section 33, T16S, R8W) andwith a southerly exposure, has several much weathered petroglyphs ofsome antiquity, as well as others that have been recently deepened orotherwise retouched. One figure (fig. 89, a) is apparently that of anIndian standing upright on a saddled horse ; the arms are spread, andfrom one hangs a plume or feathered object. Another individual(fig. 89, h) wears a tall feathered headdress, has a bow in one hand andan arrow or stafflike object in the other, and a circle at each side of thehead. The figures are 8 to 10 inches (20-25 cm.) high.Site No. 5. S'priggs Rocks, Rice County {URCl).?This interest-ing group is in a small north-facing spring cove in the SWi4, NW^,S13, T19S, R7W, 2 miles west of Little River and about 200 yardssouth of an artificial lake on a short unnamed tributary of the LittleArkansas River. The cove consists of a semicircular sandstone ledgeperhaps 20 yards across by 20 feet high, from beneath which issuesa moderate seep of water. Several large blocks have fallen from theledge in recent years, and I was informed by G. L. Whiteford, whofirst showed me the spot, that many of the best petroglyphs are onwhat is now the underside of one of these boulders. The most inter-esting carvings now are on the east wall, many of them at heightssomewhat greater than a standing person can reach. A very tenaciousfilm of lichens, and the inevitable profusion of recent visitors' names,dates, and "doodlings" obscure much of the native art. The groupdeserves more extended study than we were able to give it (pis. 78, a,79).Anthropomorphic figures from 7 to 68 cm. high, most or all withrectangular bodies, are rather common (fig. 90). The bodies aregenerally bisected by a vertical midline, on each side of which areclosely spaced horizontal or slanting lines somewhat reminiscent ofthe elaborate tubular breast ornaments of the historic Plains tribes.The limbs and necks are thin single-line affairs ; heads are either smallopen circles or shallow circular pits, with or without short rays ortreelike (feathered?) headdresses. A row of narrow-bodied humanswith outspread arms, some of them carrying stafflike objects, maj WBDBLJ AN INTRODUCTION TO KANSAS ARCHEOLOGY 489 n ?rI Figure 90.?Petroglyphs at Spriggs Rocks, 14RC1, near Little Ri 490 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174 represent dancers or celebrants (fig. 90, c) ; and one of two largefigures has a circle superimposed on the upper part of each leg thatmight represent a tortoise-shell rattle (fig. 90, a). One obscurelimbless and headless torso has a triangle pendent from the shoulderline, reminiscent of the biblike front of a buckskin shirt or dress.None of my photographs shows mounted figures, horses, or otherquadrupeds.Geometric designs include a small long-rayed disk with hollowcenter (cf. Site No. 1, supra) ; a biconvex or lens-shaped element withvertical lines above and below ; numerous small rectangles containingtwo short vertical marks, either at the center or near the lower end ; two short horizontal rows of contiguous circles, one row about 15 cm.above the other, with vertical lines connecting each circle in one rowwith the corresponding one in the other ; and trident, bird-track, andmiscellaneous other designs. There are indications of older anthro-pomorphic and other figures that have been obscured and mutilatedby those now clearly marked.Site No. 6. Peverley jann^ Rice County {IJ^RCIO).?Three and ahalf miles northwest of the preceding and about half a mile south ofthe "council-circle" on the Hayes site (p. 321) is a north-facing tree-sheltered sandstone bluff with a spring at its base. Drainage is eastand north by way of a normally dry branch to the Little Ai'kansas.Various simple petroglyphs, most of them apparently geometric, arescattered sparingly over the bluff face. Anthropomorphic figureslike those at Spriggs Rocks and elsewhere seem to be absent. Mostof the readily decipherable petroglyphs we saw are shown in figure91 (pi. 77, h). ^V ? CD;x^*;^ A-Figure 91.?Petroglyphs from locality No. 6, Peverley farm, 14RC10.Site No. 7. Near Gait, Rice County {URCll).?About half amile east of Site 6, in the NW4 Sec. 35, T18S, K7W, is another smallsandstone ledge on a dry creek. It faces west, is about 15 or 20 feethigh, and has several seeps at its base. The spot has a few trees andcould have served as a watering place for small parties of Indians. Wbdhl] AN INTRODUCTION TO KANSAS ARCHEOLOGY 491 O (D ?Figure 92.?Petroglyphs from locality No. 7, near Gait, 14RC11.Above the main seep are a few inconspicuous petroglyphs (fig. 92).The principal one now visible has a central cavity about 10 cm. acrossby 3 or 4 cm. deep, encircled by a line about 3 cm. distant. Betweenthe hole and the encircling line is a rounded ridge, with short radiallines running up out of the cavity. There are traces of a secondincomplete encircling line. Just below and to one side are faintevidences of a circular or oval design, vertically bisected by twolines, with each part of the circle filled in with parallel horizontallines. There are a number of simple circles, each with a vertical linerunning partly or entirely across it. The direction of exposure is anunfavorable one, and it is quite possible that a number of other figureshave been obliterated by weathering (pi. 77, c).Site No. 8. Near Minneajiolis^ Ottawa County {UfOTJ/).?Thisrather unusual group is on a northeast-facing sandstone bluff about6 miles southeast of Minneapolis, in the SE14 Sec. 33, E3W, TllS,and about 3I/2 miles southeast of the village site designated Min-neapolis 1 (Wedel, 1935). A swale fronts the bluff, suggesting anabandoned channel of the Solomon Eiver, now some hundreds ofyards distant. The bluff is screened by trees, and at its base is asandy accimiulation whose upper portions cover parts of some of thecarvings. I know of no village site in the immediate vicinity, thoughit must be admitted that no very thorough search for one has yetbeen made.The most striking figures on the cliff, apparently repiesentingmythological creatures or monsters, cover a more or less continuouspanel perhaps 25 feet long by 6 or 8 feet high. They are quadrupeds ; 492 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174three have antlered or plumed heads, and all have a row of triangularplates or slanting lines running down the back. Body areas arefilled in with closely spaced parallel lines, triangles, and crosshatch-ing; one has a row of circles on the back. The only remotely com-parable figure I recall from the general area is the large one dimlyvisible at Site 3.There are also a number of small simple human figures, some ofuncertain authenticity; tipi and pennantlike designs, etc.Site 9. Near Arkansas City^ Gmdley County {IJi-GOJi).?On theHall farm immediately west of the Arkansas City Country Clubvillage site, on low ground between the bluffs and the Walnut River,is another unusual group of petroglyphs. They are on the flat uppersurfaces of limestone boulders and ledges near a good spring. Hereare to be found scores of shallow hemispherical pits from 25 to 55mm. in diameter and about half as deep. Some are the heads ofsimple "stick" figures, i. e., single-line human representations, withuplifted arms and spread feet; others are connected with each otherby broad shallow lines, or have a straight line running out 15 to 45cm. to end in a Y or a three-pronged bird track. One pit is encircledby a line opening toward the east through two parallel lines. Themajority of the pits, however, have no appendages, and do not appearto be parts of larger figures or configurations. The entire group isshown in figure 93, copied from a diagram furnished me by LouisEssex of Arkansas City. I know of no similar group in the State,and can offer no suggestion regarding their age or possible relation-ship to the nearby Country Club mound group.Site 10. Near Liberty, Montgomery County (l^MTl).?Thisgroup is located about a mile or slightly more east of Liberty, on alofty hilltop outcrop overlooking Big Hill Creek and the Verdigrisvalley to the west. There are two or possibly three war-bonnetedhorseback riders, a number of small bored holes that tend to occur inpairs, paired elliptical hollows suggesting deer tracks, etc. My verybrief inspection of this site was made late in the afternoon when thelighting was not altogether favorable. There has been considerableweathering but the designs were carved unusually deep and so retaintheir outlines fairly well. The mounted figures, of course, are fairlyrecent, but some of the others may be considerably older (pi. 77, d).By way of augmenting our rather meager findings and on the possi-bility that some future investigator may be stimulated to more ex-tended field studies, it may be well to enumerate some of the scatteredprevious records of petroglyphs in Kansas. Unfortunately, little morethan enumeration is possible, owing to their extreme brevity. As with8 of the foregoing 10 localities, most of these earlier records alsopertain to the north-central part of the State. They contain little in- Wedbl] AN INTRODUCTION TO KANSAS ARCHEOLOGY 493 3prt'n??r Figure 93. -Pit-and-line markings on limestone boulders near spring on Hall farm, 14C04Cowley County.formation beyond the general location of the petroglyphs; and somesurvive apparently only as titles of unpublished papers. To this lattergroup belongs a notice of "The Pictured Rocks of Pipe Creek," a smallnortherly tributary of the Solomon, in Ottawa County (Mason, 1882,p. 4) ; and another of "Indian Sculpture in the Upper MedicineValley," a southerly aflfluent of the South Solomon in Eooks County(Gould, 1899, p. 10) . There is passing mention by Mead (1906, p. 16)of Indian pictographs in the hills up Paradise Creek, a tributary of theSaline which traverses Rooks, Osborne, and Russell Counties. Brower(1899, pp. 77-78) alludes to "an old and weathered inscription upon arock, said to be of Spanish origin," at a spring near a hilltop on Gyp-sum Creek, above Roxbury, McPherson County ; and to a Spanish flagreportedly cut on stone on Big Creek, in Ellis County. Neither of theselocalities was seen by Brower. Finally, Stockton (1883, p. 685) reportsthat a cave in the Verdigris drainage near Toronto, Woodson County,was said to have carvings on the walls. 494 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174In the absence of a comprehensive and detailed analysis of Plainspetroglyphs, it would obviously be premature to attempt to indicate inhow far the groups recorded above are representative of the regiongenerally. Some of the simpler designs, of course, are very widespread,including the rayed disk, ladderlike figures, three-toed bird tracks ortridents, quartered squares, zigzag lines, etc. Indian figures on horse-back, heads with horn headdress and/or upstanding hair or plumeeffects, tipis, bows and arrows, and other designs with late Plainsattributes also have a wide distribution in the region. Cuplike pitswith attached linear carvings somewhat like those at Site 9 have beenreported from Loving County, Texas (Jackson, 1938, p. 137, pis. XCIIand XCIII). Rectangular-bodied humans with square or pointedshoulders and outspread arms are not unique to Kansas, but I havebeen unable to find record of the more elaborate variants, such as thoseat Spriggs Rocks, in reports dealing with adjoining areas. The fan-ciful monsters of Site 8, near Minneapolis, likewise appear to have nocounterpart among recorded Plains petroglyph groups (cf. Mallery,1893; Renaud, 1936; Hill and Cooper, 1938).OTHER SITES AND LOCALITIESIn addition to the village sites so far discussed, at most of which ourobservations included excavation, a number of others merit briefnotice here. These have all been examined by the writer, but noexcavations were undertaken. Since the sites are in localities where noextended work was done by the United States National Museum expe-ditions, they help to fill in some of the geographic lacunae in our ownfindings. For such information as is available concerning the occur-rences and the nature of the associated artifacts, I am indebted tovarious private collectors who permitted examination of theirmaterials. SITES NEAR BAZAAR, CHASE COUNTY (14CS1)The principal sites to be considered here are situated a little morethan a mile south of Bazaar, in the heart of the Flint Hills upland.They are on the east (right) bank of the South Fork CottonwoodRiver, which courses thence north by slightly east to join the Cotton-wood about 3 miles below Cottonwood Falls. The South Fork Cotton-wood is, or formerly was, a perennial hardwood-fringed stream flowingin a narrow flat-floored valley about a mile wide with bordering bluffs200 feet or more high. The alluvial valley floor, across which thestream winds in a deeply incised channel, is mostly cleared and undercultivation ; the bordering bluffs remain largely in grass and serve forgrazing of livestock. Rock Creek joins the South Fork Cottonwoodfrom the west just below the sites. WuDBL] AN INTRODUCTION TO KANSAS ARCHEOLOGY 495The exact nature of the archeological zone here (14CS1) is some-what uncertain. The valley floor is actually the broad firet terrace,some 500 yards or more wide and extending about 1,200 yards northto south between the river chamiel and the foot of the bluffs. Hereand there may be seen curving swales that look like segments offormer river channels. The owners of the property, Frank andGeorge Roniger, have a large collection, mainly of chipped-stone ob-jects, gathered over the years from the terrace ; but I have the impres-sion that the materials are, or were, of rather uneven and spottyoccuiTence. At one point, on the terrace edge overlooking an oldchannel and some 200 yards east of the present stream, a relativelyheavy surface yield prompted the owners to put a bull-dozer intoplay some years ago. I was informed that a good many specimenswere uncovered at the time, that these were chiefly artifacts ofchipped stone and bone refuse, and that ash deposits or fireplacescame to light; there seems to have been very little or no pottery, how-ever. At other places on the terrace, plowing turns up small concen-trations of limestone blocks and fragments, occasionally fire-red-dened, along with chipped flints and rejectage; and it appears thatthese features are somewhat more plentiful near the foot of the bluffsalong the east edge of the terrace than they are farther out on thevalley floor. On a short visit to the locality, in August 1953, I foundpotsherds at only one point?a terrace tip near the present riverchannel, where bits of grass-impressed wattling clay, small projectilepoints, bits of charcoal, and shell fragments were also noted.The Roniger collection includes a very large and varied assortmentof chert blades, scrapers, drill points, knives, projectile points, andmiscellaneous cutting and chopping tools. Complete analysis anddescription would require far more time than I have been able to givethe material. A good many of the pieces suggest unfinished, orroughly shaped, quarry blanks rather than finished implements.Projectile points vary from small triangular notched and unnotchedspecimens to large comer-notched and stemmed forms, suggestingmore than one cultural complex. Two polished %-grooved axes,numerous large corner-notched points, and several chert "cones" arereminiscent of Kansas City Hopewellian materials (Wedel, 1943) ; and among the surprisingly scanty pottery present, there are severalheavy grit-tempered sherds with cord-roughened surfaces, punchedbosses, dentate stamp and cord-wrapped stick impressions, zoneddecoration, and other traits (pi. 80, a-e) that suggest minority waresat the Renner site. I saw no cross-hatched or rocker-roughened bodysherds.Another small group of potsherds ranged in color from brown toorange-brown, had all-over exterior cord roughening on body and rim,484172?69 33 496 BUREAU OF AMERICAN ETHNOLOGY [Bdlu 174direct unthickened and slightly collared rims that were apparentlyvertical or outflaring, and plain lips, all apparently on large full-bodied vessels (pi. 80, f-j). These sherds, I was told, were picked upin association with small projectile points, mainly in the same areawhere I found sunilar fragments along with wattling clay, smallpoints, and other detritus suggesting house site remains.Much of the heavy stonework is reminiscent of that reported byWinchell, Brower, and others at sites south of Kansas Kiver, whereit occurs usually with little or no pottery. I am inclined to regardthis as workshop material, the stone originating somewhere nearbyin the ledges of gray, blue-gray, and other chert with which theFlint Hills abound. The Eoniger brothers sliowed me two probablequarry localities a few miles distant (see p. 480), and informed methat they believe others existed in the region. The cultural afiilia-tions of the people who worked these quarries and utilized the pre-sumed workshop areas remains imcertain ; but a Hopewellian or otherMiddle Woodland period affiliation for some of the remains may besuggested. It is not clearly demonstrable, of course, that the groovedstone axes and other Woodland or Hopewellian materials actuallywere contemporaneous with the workshop materials ; they may belongto a group who camped or lived in the same locality, but without anyother connection with the flint work.On the basis of the very limited ceramic materials I have seen inthe Roniger collection, I believe that at least two occupations bypottery-making peoples can be inferred. One of these appears to beassignable to a Middle Woodland horizon, possibly or probably withHopewellian influence or connections. The other, and presumablylater, may correlate with a square pit-house people better known fromlimited excavations in village sites in the Smoky Hill-Kansas Riverdrainage to the north. Unless and until systematic excavations arepossible in the Chase County locality, it is impossible to go furtherin this matter of suggesting the affiliations of the complexes appar-ently represented. It is interesting to note further, however, thatthe ceramic materials which we may presume on the basis of findingselsewhere to be latest here appear to occur along or near the banksof the present stream, whereas the earlier Woodland materials arefound mainly at a distance from the present channel and in proximityto old channels whose remnants cross the terrace surface here andthere. The possibility of such correlations between temporally dis-tinct archeological materials and the changing topography of thestream valley should, of course, be checked more carefully. Fromthe standpoint of the cultural problems involved, one wonderswhether the cord-roughened wares and small projectile pointsassociated with wattling clay represent a square-house agricultural Wbdbl] an introduction TO KANSAS ARCHEOLOGY 497people of attenuated or incipient Upper Republican type, or areperhaps a late Woodland manifestation transitional to Upper Re-publican.Other sites and remains were also reported to me in this generallocality. Chipped-stone artifacts and burned stones are said to occuron virtually every suitable flood-free terrace along the South ForkCottonwood from a point below Bazaar upstream to Matfield Greenor beyond. Pottery is nowhere abundant, to judge from the infor-mation given me. The Roniger brothers stated that buried hearths,some of them at considerable depth, have been noted from time totime at various places along the banks of the stream ; but they wereunable to show me any diagnostic or other artifacts from such occur-rences. A stone-walled or stone-covered burial caim was openedmany years ago on the bluff top east of the Roniger terrace ; but therewere no artifacts and the cultural affiliations and time horizon of thefeature thus remain unknown (but see Eyman, MS.) . That an interesting and important group of sites formerly existedin this locality is clear; and I am convinced that at least two pre-historic horizons are manifested. There is some evidence of a thirdin two small L-shaped red stone pipes in the Roniger collection,which are very likely to be correlated with the protohistoric LittleRiver Focus remains farther west. All of this strongly suggests fre-quent and repeated use of the South Fork Cottonwood valley, wherearable land, good water, timber, and an abundance of gi-azing landfor bison, elk, deer, and other game animals were formerly foundin close proximity to one another. To what extent the several groupsactually settled and dwelt here, perhaps growing com and caiTyingon their domestic activities for prolonged periods, I cannot say. Itis possible that some of the variety in remains noted is a result ofthe former attractiveness of the locality for himting groups in questof plentiful and easily taken game. The picture suggests, however,that a sustained and thoroughgoing effort should be made to disen-tangle, either at Bazaar or else in some of the other sites nearby, thevarious aboriginal cultural complexes for which we now have indica-tions in the region.SITES NEAR MARION, MARION COUNTYThis locality lies near the headwaters of the Cottonwood River.The town of Marion, formerly Marion Center, is situated about 3miles east of the junction of the South Cottonwood with the Cotton-wood, and at the confluence of the latter with a small stream known asBrook Luta. The river valley here is a mile or somewhat more inwidth, with bluffs perhaps 100 to 150 feet high. The streams areperennial, with bordering belts of hardwood timber beyond which 498 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174 are the rolling grasslands characterizing the western slopes of theFlint Hills upland.For such information as is available regarding the archeology of thelocality, we are dependent almost entirely on observations and col-lections made more than 70 years ago. These were made by localcollectors, principally one M. O. Billings, of Marion, and by EdwinCurtiss, who opened several earth mounds and rock cairns south ofMarion in 1879, under the sponsorship of F. W. Putnam of the Pea-body Museum. The exact location of the various features reported bythese men remains obscure, but they seem to have lain within 4 or 5miles of Marion, chiefly to the south of the city and apparently on thewest side of Cottonwood Kiver.In 1881, reporting on Curtiss' fieldwork and certain additions to thecollections in the preceding year, Putnam wrote (1880, pp. 718, 738) : Besides making an examination of a large burial mound in Marion Co., Kansas,Mr. Curtiss removed several cairns nearby. These seem to be the monuments oCa later people than those who buried in the mound. These piles of stones, hefound, were erected over bodies which had been placed on the bare rock. Theskeletons under all the cairns that Mr. Curtiss removed were so much decayedthat not a single cranium could be obtained. A number of objects of variouskinds were found, which had been placed with the bodies. Among the most in-teresting of these, perhaps are the several minute pipes made of catUnite.Under one of the piles of stones an arrowpoint of obsidian was found associatedwith a glass bead. This single glass bead was the only object of European makediscovered, but it Is conclusive evidence of the comparatively recent period whenat least one of this group of cairns was erected.Among the additions to the museum for 1879 were reported thefollowing : 19517-19715. Fragments of clay slate and a large grooved hammerstone froma shell heap near Marion Centre, Kansas ; stone pipes, drills, scrapers, knives,hammers, and implements for grinding and polishing, all of stone, from the sur-face on Mr. Leachman's farm and other places near Marion Centre, Kansas ; shellornaments, flint knives, clay and bones, rubbing stones, grooved hammers,and other implements of stone ; a glass bead, bone awl, and other worked bones,obsidian flakes, unio shells, flint drills, arrowheads and hoe of stone, red chalkand pipes of catlinite from caims on the west side of Cottonwood River, 3 milessouth of Marion Centre, Kansas ; bone whistle and awl, celt and other implementsof the same material, implements of flint and horn, arrow straighteners of stone,and grooved hammerstones from a mound on Mrs. Gibson's farm near MarionCentre, Kansas; ....According to Billings (1883, pp. 211-212), the remains aroundMarion at the confluence of Cottonwood Eiver and Muddy Creek(Brook Luta?), consisted of mounds "on high ground around thejunction of streams adjacent, in irregular groups averaging 8 moundsto the group. Inside of 3 miles each way from Marion there are fiveof these groups." The mounds, according to Billings, were from 10to 60 feet in diameter and from 1 to 114 feet high ; they contained no Wbdbl] an introduction TO KANSAS ARCHEOLOGY 499burials. There were also "Smaller mounds [which] seem to be theremains of adobe huts . . . [and] are circular while the large ones haveon the southeast side a spur about % the size of the mound." All thesemounds were said to contain trash, pottery, shells, etc. Billings alsomentions fire beds (caches ? ) , one of which was 6 feet deep, 5 feet acrossthe bottom, and 3 feet across the top. These he said, occurred at sev-eral places. Nowhere does Billings describe any of the artifacts fromthese sites ; but he indicated that the local antiquities could be dividedinto three categories, namely, Moundbuilders, Cremators, and ModernIndians.Despite the apparent former abundance of aboriginal materials inthe Marion locality, as judged from the observations of Billings andCurtiss, comparatively few artifacts from here seem to have foundtheir way to museums where they were preserved. Probably the bestsingle collection is that made by Curtiss, and now preserved in Pea-body Museum of Harvard University under catalog numbers 19517-19613. Most of this, judging from the accompanying catalog entries,came from cairns, not described in detail, 3 to 4 miles south of Marion,apparently on the west side of Cottonwood River; from two earthmounds (middens?) and adjacent plowed fields on "Mrs. Gibson'sfarm on Cottonwood River near Marion Centre" ; from a shell heapon the Bowers' farm ; and from the surface of the Leachman farm.^^The only field notes left by Curtiss, so far as I am aware, relate tothe two mounds opened on Mrs. Gibson's farm. The first of these was45 by 30 feet, with a height of 4 feet. It was said to contain charcoal,ashes, shells, bones of animals, turtles, and birds, and miscellaneousworked and unworked material of bone, stone, shell, etc. Curtiss sayshe "occasionally found pieces of pottery, fragments only, and that wasmade like our southern pottery [with] shells, clay, and sand." Muchmaterial was found on plowed fields nearby. Mound No. 2 was closeby; it was smaller, about 2 feet high, and contained bones, charcoal,ashes, shells, and other refuse. No burials were encountered in eithermound ; and from their general nature, as Curtiss described them,there can be little doubt that they were primarily refuse accumula-tions from a village occupation nearby. ^5 Neither the very sketchy notes left by Curtiss nor the published reports by Putnamlocate exactly any of these sites ; and the consistent omission of first names or initialswhen property owners are mentioned makes it difficult to identify today the farms named.In answer to my inquiry, Miss Lillian Pierce, register of deeds of Marion County, writes(letter of May 24, 1954) that Mrs. BarzUla Gibson owned the Si^NW% and part of theE^SWVi of Section 8 T20S R4E, just south of Marion; that Joseph Bowers held titleto the E%NW% of Section 20, Francis Bowers to the W%NWi4 of Section 20, boththese tracts lying west of Cottonwood River ; and that John L. Leachman held land insection 18, about a mile west of Cottonwood River on a small unnamed tributary (SpringCreek?) which enters the river a short distance to the southeast. All these lands aresouth of Marion, within 3 miles of the south edge of the city, and on the west side of theCottonwood. Exact relocation of the archeologlcal features, If still feasible, will have tobe done on the groand. 500 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174Among the collections deposited in Peabody Museum is an interest-ing variety of stone, bone, shell, and other objects. Strangely enough,there is but one potsherd?a straight vertical rim, shell tempered, withdiagonal incisions crossing the lip. There are several grooved maulsof varying sizes and materials, including at least one well-made cylin-drical specimen of Sioux quartzite with flanges bordering the mediangroove. Five L-shaped pipes of catlinite or of a fine-grained redsandstone have tall cylindrical to slightly bulbous bowls from 24 to 64nmi. high, with very short stems; the smallest has a single incised lineencircling the bowl near the top, another has two Imes around the endof the stem, and a third has about a dozen notches cut lightly into thelip of the bowl. Longitudinally grooved shaft smoothers of Dakotasandstone include three matched pairs 13.5-15 cm. long and two addi-tional fragments, all said to have come from "Cairn III, Pile rocks 6'long." There are numerous chipped flints, including small triangularunnotched projectile points, expanded base and straight (pipe) drills,planoconvex end scrapers, side scrapers, and a variety of notched andbeveled knives and other cutting or chopping tools, some of whichwere fashioned from the banded foraminiferous chert from the "Tim-bered Mound" quarries described by Gould. Bonework includes a "rib-edge" awl ; a bison rib shaft-straightener with single perforation ; a polished bone tube, 12 by 52 mm., with dressed ends ; a single-stopbird bone whistle, incomplete; and several other odds and ends.There are also miworked freshwater shells, as well as a subcircularshell gorget, 80 by 90 mm., with four central perforations. One ob-sidian flake and two glass beads, one of them a star bead, were alsonoted.Partially supplementing the above collection is a small lot of speci-mens sent by Mr. Billings to the United States National Museum in1880, and another received in 1892 from J. W. Lambuth. There areno details concerning provenience of these two lots, except that theLambuth specimens are said to have been taken from "Ash pits orfire-vaults in Marion County, Kansas." The Billings collection in-cludes several well-made triangular unnotched points, end scrapers,drill points, a flake knife with finely retouched edges, an L-shapedpipe of red stone, and five sherds. The latter include shell and bone-tempered pieces, one of them a plain rim. The Lambuth materialincludes an incomplete red stone L-sliaped pipe, several small frag-ments of worked bone, and three sherds. One of these is a shell-tempered plainware piece; another is finely sand tempered, withdiagonal incisions on the flattened lip. The third is grit-tempered,and bears a band of dentate stamp impressions bordered by wide deepgrooves beyond which are undecorated areas. Wbdbl] an introduction TO KANSAS ARCHEOLOGY 501As to cultural affiliation of the great bulk of these materials, theevidence is open to little doubt. The Curtiss specimens are virtuallyidentical with the protohistoric Little River Focus materials collectedby the United States National Museum in 1940 in Rice County, some75 miles to the west. With exception of the bird bone whistle, everyartifact type shown in the Peabody ISIuseum collections from MarionCounty can be duplicated by regularly recurring items from theLittle River Focus sites. So far as they go, moreover, the very smallcollections in the LTnited States National Museum from Lambuth andBillings also conform. They suggest strongly that the Marion localitywas occupied at one time by peoples closely akin to those who left thelarge sites in the vicinity of Geneseo, Lyons, Lindsborg, and else-where in Rice and McPherson Counties (Wedel, 1942). Whetherthe cairns opened by Curtiss were actually burial sites, I have no wayof knowing ; they are not described in detail nor are we told how theartifacts occurred in them. The earth mounds Curtiss excavatedin the same locality were in all probability refuse heaps. In gen-eral, the Curtiss, Billings, and Lambuth collections suggest occupa-tional, that is, village site, deposits much more strongly than they doburial areas ; and whatever the actual situation with respect to formermounds in and around Marion, I strongly suspect that there are, oronce were, fairly extensive and prolific village site deposits, doubtlessincluding refuse heaps, storage pits, and perhaps habitation imits.It would be interesting to know whether the cairns identified byCurtiss as burial structures were also made by the people who leftthe occupational debris, but neither Curtiss' own records nor thebrief observations of others help on this score.My own observations in and around Marion in 1940 tend to confirmthe above suggestions as to the general nature of the aboriginal oc-cupancy of the locality. I was unable to relocate any of the sitescertainly dug by Curtiss or to find anyone who knew exactly wherehis digging was done. However, about a mile southeast of Marion,where a short flowing spring branch enters Cottonwood River fromthe east, on a pasture blufi' overlooking the river bottoms towardMarion, we found three low mounds. These were approximately40 feet in diameter by 2 or 3 feet high, and about 30 to 60 yardsapart. Each had apparently been dug into many years ago, andstill had a small sodded depression in the center. A fourth mound,said to have been opened by Billings and others, lay about 200 yardsto the west. The only surface finds made by my party included twoshell-tempered and two grit-tempered sherds and a few nondiagnosticflints. None of this material conflicts with the suggested Little RiverFocus affiliation for the Marion materials ; and the mounds look super-ficially like the middens of the Little River-Cow Creek region in 502 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174Rice County. To the best of my knowledge, pottery and other ma-terials indicating the former existence of Little River Focus villageor burial sites have not been found or reported east of the Marionlocality, though occasional L-shaped stone pipes and other items ap-pear occasionally on sites farther down the Cottonwood drainage.While the great bulk of the materials I have seen from the Marionlocality thus indicate a protohistoric occupancy closely related to thatrepresented by Little River Focus sites in Rice County, there are sug-gestions of earlier penetrations into the area from the east or north-east. Both the Billings and Lambuth collections in the NationalMuseum include sherds of dentate stamped ware with the stampingevidently arranged in zones bordered by grooves (fig. 94). These are Figure 94.?Dentate-stamped and zone-decorated potsherd from vicinity of Marion(USNM 149294).without much doubt of Hopewellian type, closely related to materialsfoimd at village sites near Kansas City (Wedel, 1943). Among theremains we saw near Marion in 1940, therefore, it is interesting tonote the possible occurrence of chambered, or so-called vault, mounds.Some 5 miles south and 1 mile east of Marion, on a high bluff lookingnorthward across an unnamed creek and the Cottonwood valley, wefound three small badly disturbed mounds. In two, no stones werevisible, and our impression was that they resembled the small earthmounds of the Manhattan locality, 70 miles to the north. The third,approximately 18 feet across, contained a large number of limestoneblocks. On clearing some of these away, we found clear evidence ofa dry masonry wall on the west and north sides, with a well-madecorner in the northwest. Most of the structure, which appears to havemeasured about 5 by 7 feet, had been destroyed, and we could findno evidence of an entrance passage or vestibule, such as character-izes many of the chambered mounds in the Kansas City area. If thiswas indeed a chambered burial mound, its occurrence a hundred milesor more southwest of any other recorded structure of the sort and therecovery of Hopewellian sherds in the nearby area, add support to thethesis of a somewliat more intensive penetration by Hopewellian peo- Wmdel] an introduction TO KANSAS ARCHEOLOGY 503pies into the trans-Missouri plains of Kansas than we have hithertosuspected (see also Smith, 1949 a, p. 297). I believe that thorough re-connaissance of the upper Cottonwood-Neosho drainage area wouldprobably tlu-ow further light on this interesting possibility.SITE 14PT1, PRATT COUNTYThis interesting and very promising site is situated on the north,or left, bank of Ninnescah Kiver approximately 1 mile southwest ofPratt. It was brought to the attention of the United States NationalMuseum late in 1953 when Eugene Wing, of Pratt, sent in a series ofbone implements for identification and comment. Most of these wererecognized at once as so-called "digging stick heads"?a type of arti-fact not heretofore reported from any sites in Kansas, to my knowl-edge. To my request for additional information, Mr. Wing re-sponded with lengthy and informative letters and, ultimately, withsamples of pottery, stone, bone, and other artifacts collected by him-self from the surface of the site. According to Mr. Wing, the loca-tion has been known to collectors for many years, and a considerableamount of material has doubtless been carried away by them ; his ownactivities here have been carried on for some 12 to 15 years. So faras is known, there has been no excavation on the site.According to Mr. Wing (letter of October 30, 1953), "About 100feet north of the [Ninnescah] river, the terrain forms a bowl about200 yards long north and south, and 150 yards east and west. In thisbowl lies the campsite." In another letter, dated November 12, 1953,he added the following observations: "When this campsite wasfarmed this last spring, it was plowed deeper than ever before andeach campfire showed plainly on the surface of the ground. Therewere very many and I noticed that they were not arranged in a man-ner suggesting one camp layout. From this appearance, it seemedto me . . . that this location may have been occupied at several dif-ferent times." This observation is directly relevant to certainproblems arising from my analysis of the artifacts sent me by Mr.Wing.Artifacts from this site examined at the present writing numberapproximately 150 specimens. Included are about 100 potsherds, 29chipped-stone objects, 8 items of ground stone, 17 of bone, 1 of shell,and 1 of glass. To what extent they are truly representative of thetotal complex, or complexes, represented at the site, I am obviouslyin no position to say at the moment. They seem, however, to be ofsufficient importance and promise to merit description in some detail.With respect to the pottery, the sherds at hand suggest a rathersurprising variety of wares as regards the fabric, surface treatment,and other characteristics?more variety, at any rate, than is ex- 504 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174hibited in most of the sites with which my own Kansas researches andexcavations were primarily concerned. Since the sample is rathersmall, I have not attempted to set up and name what may turn out tobe distinct types ; rather, I have chosen to group the sherds into whatI am tentatively calling wares, each described as fully as the handfulof specimens permits. If future excavation, and the use of a largerand more definitive sample, confirms the distinctions I think I see,it may become possible to establish types that have true historic orgeographic significance.Ware A, most abundant in our sample, includes 60 sherds, all ofwhich have cord-roughened exterior surfaces (pi. 81, e? o o Figure 100.?Miscellaneous sherds from Wyandotte County.excavations by the United States National Museum at the Rennersite, a Hopewellian settlement, and in related chambered burialmounds nearby, in Platte County, Mo., have been reported at lengthin another place (Wedel, 1943). It was pointed out there (ibid.,pp. 98-103, and fig. 1, sites 1-4) that related materials occurred atseveral locations in northeastern Kansas. The best represented ofthese, and probably the most important of which we now have anyrecord, is the Trowbridge site near Bethel ; others are known on Plum484172?5? -36 544 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174Creek north of Leavenworth, in Klamm Park and at 9th Street andChelsea Trafficway, Kansas City, and on a small creek near Edwards-ville. I have reexamined none of these localities since our 1937investigations; and for the most part, they are now either destroyedor else have yielded too little material to permit satisfactory definitionof the complex to which they belong. Through the courtesy of H. M.Trowbridge, however, I have been privileged to examine several timesan extensive collection he has gathered over a period of years fromthe site near Bethel.The Trowbridge village site 14WY1, now largely destroyed byconstruction work, occupied some 3 or 4 acres on a sheltered small-creek terrace. Cultural materials occurred in a dark, refuse-laden soilstratum 12 to 18 inches thick and overlain by up to 2 feet of slopewash. Trash-filled (cache) pits like those at the Renner site werepresent, but there were no traces of post-molds, hearths, or houseoutlines. No beans, maize, or squash remains Avere noted duringthe digging; but the first two were found at the Renner site andwere, in all probability, known also to the inhabitants of the Trow-bridge site.The pottery remains, which occurred in abundance, have not yetreceived the careful analysis they merit. It is clear, however, thatgrit-tempered wares with plain, rocker-roughened, cord-roughened,and rare rouletted surfaces are predominant. There are no completevessels; but a large jar, vertically elongate, with slightly constrictedneck and wide mouth, appears to have been a common shape (seeWedel, 1943, fig. 4, j). There is evidence of smaller quadrilobatevessels also. Other forms can probably be determined or recon-structed from the extant sherd materials. Dentate and complicated-stamp impressions occur frequently on body sherds; and these, likethe rocker roughening, were sometimes arranged in alternate plainand stamped zones separated by incised lines or broad grooves.Cross-hatched rims seem rather less common than at Renner; othertypes include punched bosses, alone or with cord-wrapped stick ordentate stamp impressions above on the rim exterior. Generallyspeaking, rim form and decorative treatment parallel those at Renner.In addition to vessel fragments, pottery remains include crude clayfigurines such as bird effigies, a small nipplelike object, and a humanbust (?) ; and there is also a fragment unmistakably from the bitend of a clay platform pipe (fig. 102, h). A selected series of sherdsand other pottery items in the Trowbridge collection from this siteis shown in figures 101 and 102.There is also a varied and considerable series of nonpottery mate-rials (see fig. 103). Chipped stone includes numerous large stemmedor corner-notched projectile points, and lesser numbers of scrap- Wedel] an introduction TO KANSAS AftCHEOLOGT 545 "*'*?i-*.t42.^t6,Lt?-C^**-'~-'-k-.ClS?'l!?'i?'i^y,, ^wmmw^^Bimm l-icuaii iOi.?Dciiiaie-stamped.zone-decorated, andother sherds, Trowbridge site, 14WY1. ers, disks, drills, flake and other knives, and one flake of workedobsidian. The three-quarter grooved ax, of which several werefound at Eenner, has not been reported from the Trowbridge site;but there is a polished hematite celt, along with stone pendants,gorget fragments, sandstone abraders, and worn and grooved pumice 546 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174 Figure 102.?Miscellaneous sherds, clay pipe stem fragment, and crude effigies. Trowbridgesite, 14WY1. * Wkdbl] an introduction TO KANSAS ARCHEOLOGY 547lumps. Present also are conical socketed projectile points (fig. 103, d) ,flakers, cylindrical rubbing tools, and a small dentate roulette (Wedeland Trowbridge, 1940) for marking pottery (fig. 103, e) , all of antler;and beaming tools of split deer metapodials (fig. 103, a) and innomi-nates, mammal bone awls, and ulna punches. Two fragmentary, butworked, turtle carapaces suggest bowls or dippers.No burials have been found at the site. Across the Missouri, thereis evidence that small mounds containing square dry masonry cham-bers are assignable to the local Hopewellian complex, and perhapsearth mounds as well. There are a few chambered mounds on theKansas side of the Missouri (Wedel, 1943, p. 159), but these, likeHopewell village sites and sherd areas, have not been reported northof the Nebraska-Kansas boundary.The materials from the Trowbridge site parallel, in general, theassemblage fomid at the Renner site, 5 or 6 miles distant ; and thereis no doubt as to the close relationship between the two, or of theirbasic relationship to Hopewellian culture. They are clear evidencethat Hopewellian conmiunities of some size and permanence wereestablished in the trans-Missouri region. How far westward theHopewellian occupation extended is not yet clear. Evidence of ithas been recognized up the Kansas River valley in the vicinity ofWamego; in burial mounds and habitation areas near Manhattanat the mouth of the Blue; and at the Younkin mound (Schultz andSpaulding, 1948) in Geary Coimty on the lower Republican. Therelationships of the Kansas River materials to those around Kan-sas City still await clarification ; but the Yoimkin mound, and prob-ably some of the lesser tumuli briefly described in the present paperfrom the Manliattan locality, suggest important differences (see alsoEyman, MS.).Hopewellian materials have been found in what may be a smalloccupational area in Ellsworth County (Smith, 1949 a, p. 297) ; and, inthe form of a rocker-roughened sherd, in the Woodland level at thePottorff site on Salt Creek in Lane County, yet farther west. Almostnothing is known of the distribution south of Kansas River ; but thereare dentate-stamped sherds with zoned decoration from the Cotton-wood valley near Marion and again near Bazaar. There is also someevidence of former stone-chambered burial mounds in the Marionlocality. I have unconfirmed verbal reports of Hopewellian potsherdsfrom south-central Kansas in the Arkansas-lower Walnut drainage,and have seen dentate-stamped sherds from the vicinity of Independ-ence. It seems likely that additional such materials will be recordedin time from the Neosho and Verdigris drainages, where connectionsbetween the Kansas City district and the Delaware County, Okla-homa, Hopewellian locality (Bell and Baerreis, 1951, p. 27) may beexpected. 548 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174 m urn til \ 'Jl & |fV*v if.'- '"i M Figure I03.-Miscellancous artifacts of bone, stone, and antler including roulette forpottery marking (e), from Trowbridge site, 14WY1. Wedel] an introduction TO ICANSAS ARCHEOLOGY 549No Hoj)ewellian village or burial sites have been reported fromNebraska; but the presence of a few rocker-roughened body andcross-hatched rim sherds in an otherwise dissimilar Woodland potterycomplex near Peru (Hill and Kivett, 1941, p. 198) should be noted.Along the Missouri River and on tributary creeks in the southeasternpart of Nebraska, there are other remains assignable to the Woodlandperiod. First of these to be recognized by a name, and one of whichsome traces might be expected in northeastern Kansas, is the SternsCreek culture (Strong, 1935, pp. 175-198; Champe, 1946). This isbest-known from the Walker Gilmore site in Cass County, the onlylocation from which any appreciable amount of excavated material isavailable. Here numerous hearths, ash lenses, and other remainsoccur at depths up to 27 feet beneath the ground surface, and for somehundreds of yards along the course of the creek. Small post moldsand masses of grass suggest the remains of pole and thatch dwellings,of an Eastern Woodland type. There are abundant remains of thebush summer squash and the bottle gourd, but none of maize or beans.Bone refuse includes many remains of deer, small mammals, andbirds, few of bison. Worked bone consists of numerous eyed needles,awls, a few tubes or beads, a longitudinally perforated deer-toe bone,antler flaking tools, and picks. Worked stone is not plentiful; itincludes roughly fashioned knives or picks, hammerstones, a polishedcelt or ax fragment, and crudely chipped notched and unnotchedprojectile points. Pottery is grit tempered and usually smooth sur-faced, but with some sherds that suggest straw roughening. Small tomedium-sized vessels are indicated, apparently with more or lessconical base, vertically elongate shape, slightly to strongly constrictedneck, and simple rims. There is little or no decoration beyond scallop-ing or incising of the rim exterior and occasional single-cord impres-sions on the rim or neck. Clay pipes, presumably of straight form,are indicated.The geographic range of Sterns Creek culture is very incompletelyknown, but recognizable remains of this sort do not appear to haveextended very far westward from the immediate valley of the Mis-souri River. Sherds regarded as assignable to this complex havebeen reported from as far upstream as Harrison Comity, Iowa(Keyes, 1949, p. 97) and downstream to Holt County, Mo. (Wedel,1940 a, p. 305) , the known north-south spread thus not much exceeding125 miles. At the moment, therefore, this appears to be the most re-stricted areally of the known Central Plams Woodland complexes, asit is also ceramically perhaps the least representative.Farther up the Missouri, and extending westward over much ofnortheastern Nebraska, is another Woodland complex of which notrace has yet been reported from Kansas. For this, the term "Loseke 550 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174Creek Focus" has been suggested (Kivett, 1952, p. 70) , the name beingtaken from a small tributary of Shell Creek in Platte Comity, Nebr.,where several village sites have been investigated. No house remainswere identified, but small storage pits and hearths were present. Pot-tery is sand- or grit-tempered and predominantly cord-roughened. Itis distinguished by frequent use of single-cord impressions on the rimexterior, these occurring as parallel horizontal lines, in triangularareas suspended from the lip or just below, as short diagonals on theouter rim edge, or in various combinations of these devices. Non-pottery traits were very poorly represented?a few chipped projectilepoints, end and side scrapers, blades, hammerstones, and choppers,and a little bonework. Of particular interest is the demonstratedpresence of kernels of a "relatively primitive type of corn."Pottery similar to the Loseke Creek ware has been noted in theWeeping Water valley in Cass County, in burial mounds on EagleCreek in Plolt County, along the Missouri River bluffs in westernIowa, and in South Dakota. In western Iowa, as the Missouri BluffsFocus, this pottery is reported to occur stratigraphically above SternsCreek and "Northern Woodland-Hopewellian" materials (Keyes,1949, p. 97). In South Dakota, as represented at the Scalp Creekand Ellis Creek sites (Hurt, 1952), it appears to be associated with amore complicated and perhaps later manifestation than that reportedfrom Platte County, Nebr. The westward limits of Loseke Creekmaterials in Nebraska have not been established.Other Woodland variants, broadly characterized by a heavy pre-ponderance of thick, coarsely tempered, cord-roughened pottery, havea yet more westerly distribution. Two of these?the Valley Focusand the Keith Focus?are represented by excavated village or campand burial sites, and so have acquired some substance in addition totheir sherd remains. Other variants are suggested by sherds andother artifact samples still too limited to permit definition andnaming.The Valley Focus is known principally from a small village site onMira Creek, in Valley County, Nebr. (Hill and Kivett, 1941, pp. 146-193). The area of occupation here was well under an acre in extent,but the occurrence of debris throughout a zone some 16 to 40 inchesthick suggested prolonged occupation rather than a camp site. Ir-regularly circular and subelliptical basins from 12 to 20 feet in diam-eter, with central hearths, are presumed to be the ruins of pole andskin, or mat, dwellings. Smaller refuse-filled pits were probably usedfor storage. Subsistence was by hunting and gathering; the bonesof bison, deer, smaller mammals, and birds were fairly plentiful, butthere was no evidence of domestic plants or of horticultural tools ordevices. A single semiflexed burial in the upper detritus-laden fill Wedel] an mTRODUCTION TO KANSAS ARCHEOLOGY 551 of one of the house basins was without associated diagnostic culturalitems. Potsherds were relatively plentiful throughout, worked boneand stone much less so.The pottery, termed "Valley Cord-roughened," is a grit-tempered,cord-roughened ware. Vessels ranged in size from one-half pint toabout 6 gallons. A truncated ovoid jar, wide-mouthed and with co-noidal base, was apparently typical. Exterior surfaces were entirelycord-roughened, the impressions running vertically, diagonally, orcrisscross. Decoration consisted of punchmarks or bosses on theexterior below the rim, cord-wrapped rod impressions, or a combinationof the latter with punclimarks or bosses ; and there was a little incisingor trailing as well.The scanty stonework consisted of stemmed and corner-notchedprojectile points of various sizes, but generally smaller than those ofKansas City Hopewellian; end scrapers; ovate to subrectangularknives, none beveled ; and sandstone sharpening blocks, but no pairedshaft abraders. Objects of bone included hollowed bison phalanges("drill rests"); split deer metapodial and splinter awls; large andsmall bone tubes ; bison scapula beamers ; three serrate tips of deer orantelope metapodial fleshers ; a bison ulna pick ; and the biperforatesection of a deer skull with horn bases, perhaps an ornament or head-dress. A simple musselshell pendant was the only piece of workedshell found.The areal extent of the Valley Focus complex has nowhere beenoutlined with certainty. In terms of ceramics, however, it undoubt-edly represents a much more widely spread Central Plains Woodlandtradition than does the Sterns Creek Focus (Hill and Kivett, 1941,p. 240). Throughout much of Nebraska, especially in the Sandhills,east to the Missouri, and southward, heavy-walled, coarsely tempered,cord-roughened sherds may be found, and with them often rims thathave exterior punchmarks or bosses. Sherds with this rim treatmenthave been reported (Hill and Kivett, 1941, pp. 194-243) from suchwidely separated localities as Cass, Douglas, Dixon, Platte, Holt, andCherry Counties. It is not yet certain, however, that other elementsof the Valley Focus complex are similarly distributed or that thesesherd occurrences always and invariably involve only Valley Cord-roughened. The type Ash Hollow Cord-Eoughened, represented inLens D at Ash Hollow Cave, Garden County (Champe, 1946, p. 56),and perhaps at the Kelso site, Hooker County, has a general resem-blance to the Valley ware (Kivett, 1952, p. 67) and may be representedby some of the surface finds.Thick cord-roughened grit-tempered sherds, including exterior rimbosses, occurred as a small minority ware at Renner, and are apparentlypresent also at the Trowbridge site. From their persistent presence in 552 BUREAU OF AMERICAN ETHNOLOGY [Bcll. 174small numbers in local collections from various lesser creek valleys innortheastern Kansas, I would suspect a wide scattering of camp orvillage sites here. Their distribution westwardly, and also south ofKansas River, has not been worked out. I have seen heavy siliceous-tempered, cord-roughened sherds from Montgomery County, near In-dependence, and from the vicinity of Larned in Pawnee County.Interpretation of this Kansas material as Valley Focus, as in the caseof many of the Nebraska surface sherd finds, is premature in theabsence of more adequate pottery samples and of other traits diagnosticof the named complex.The Keith Focus has been set up on the basis of findings at severalexcavated village and burial sites, chiefly in the Republican Riverdrainage of southern Nebraska and northern Kansas (Kivett, 1949,1952, 1953) . Indicated are small communities, each with traces of notmore than 4 to 6 structures occupying usually much less than an acreof ground ; circular to irregularly elliptical shallow basins, sometimeswith poorly marked fireplaces in the center, and thought to representthe sites of small perishable habitations constructed of poles and thatchor skins; postmolds and small refuse-filled pits scattered randomlyabout the living area ; presence of deer, small mammal, and bird bones,as well as bison ; absence of maize or other domestic plant food remains,or of implements clearly designed for horticulture ; and scarcity of pot-tery, stone, and bone artifacts as compared to those in later pottery-bearing sites of the region.Among the distinguishing characteristics of the Keith Focus is thepottery designated Harlan Cord-roughened. This is a thick-walled,calcite-tempered ware, with allover cord roughening on vessel ex-teriors. The usual shape appears to have been a large wide-mouthedjar, with more or less conoidal base, direct unthickened rim, and flat-tened undecorated lip. Fine cord or fabric imprints are often visibleon interior sherd surfaces, these running at right angles to the heavierexterior impressions. Punched bosses, cord-wrapped rod impressions,and other decorative treatment of the rims, such as occurs in ValleyFocus pottery, is absent from Keith Focus material.Work in stone includes a rather varied series of small to largestemmed and barbed projectile points, often serrate; end scrapers;small chipped celts; ovate knives, but no beveled types; irregularsandstone sharpening blocks, but no paired shaft smoothers ; peckingand hammer stones ; and cupped grinding stones. Bonework consistsof split deer metapodial and splinter awls; scapula "knives" orscrapers, but no hoes; plain, incised, and barrel-shaped bird-bonebeads. There is much more worked shell than in most other Plainssites, and this was associated for the most part with the mortuarycomplex. The shells of fresh-water mussels were fashioned into tri- Wbdel] an rXTRODUCTIOX TO KANSAS ARCHEOLOGY 553angular and crescentic pendants, often corner- or end-perforated ; diskbeads from 6 to 18 mm. in diameter ; short cylindrical beads with oneflat side, with drilling from this surface and from one end so as tointersect at right angles ; and occasionally other forms. Marine shellsare represented by short thick subcylindrical segments of Busyconcolumella, drilled lengthwise; and by spire-lopped Olivellas.The Keith Focus mortuary complex, largely through the excellentwork of Kivett (1953), is better known than that of many otherPlains cultures that have been more extensively investigated. Sec-ondary interment of disarticulated skeletal parts was customary.Sometimes these remains are found in the village or camp area, butthey are more often in terrace, bluff, or hilltop locations at a distance.In some cases, bundled or scattered bones of one or several individualswere placed in small pits 3 to 4 feet across and up to 6 feet deep, thesepits tending to occur a few feet from one another in clusters (Strong,1935, pp. 116-122; Wedel, 1935, pp. 174-179). Equally character-istic are larger basins up to 20 feet or more in diameter, within orunder which there may be smaller pits. In these larger ossuaries, asexemplified at the Woodruff burial site in Phillips County, Kans.(Kivett, 1953), have been found the scattered remains of dozens orscores of individuals?so many, indeed, in comparison with the sizeof the known settlements that it can be presumed these were the finalrepositories for the dead from several villages and camps. Charcoal,burned earth, and scorched bone fragments indicate that fire playeda role in the burial rites. Mortuary offerings include large numbers ? often running into the thousands?of shell disk beads and beadblanks, many of the latter unperforated, as well as shell pendants,chipped stone, and occasionally other items.The geographic extent of Keith Focus outside the Upper Repub-lican valley center is not certainly known. For Nebraska, Kivett(1953, p. 132) reports pottery closely similar to Harlan Cord-rough-ened from Davis Creek in Sherman County, at Amherst Reservoirin Buffalo County, from the Sandhill region, and from RichardsonCounty in the extreme southeastern corner of the State. To thesouth in Kansas, what appears to be the same or a similar calcite-tempered ware is found in the Woodland level (Occupation B) at thePottorff site and again at the Walter site, both in Lane County ; inboth places it is associated with a few non-calcite-tempered sherds thatotherwise are essentially similar. Calcite-tempered sherds were foundas surface material in Scott County ; and at Coal Oil Canyon in ex-treme western Logan County, southeast of Wallace.*^ Western Kan-sas, and perhaps parts of eastern Colorado, thus appear to lie withinthe range of Keith Focus materials. *? Among sherd samples sent to George Metcalf, museum aide, by P. H. Bowman, Wallace,Kans. 554 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174The preceding discussion will afford some understanding of thediverse nature of the Central Plains materials currently classed asWoodland, and also of their basic resemblances. Two principalpatterns seem to be indicated. The comparatively rich Hopewelliancommunities of the Kansas City district, based on a semihorticul-tural subsistence economy, possessing a varied ceramic and materialculture inventory, and with strong implications of a fairly stablecommunity life, are in marked contrast to the much simpler, lesssettled, basically hunters and gatherers suggested by the widespreadKeith, Valley, and related remains. The interrelationships, cul-tural and chronological, between the several complexes are not yetclear, though they are emerging. Kivett (1953, p. 135) has indicatedthe direction in which present evidence points by tentatively group-ing the Valley and Keith Foci into the Orleans Aspect, to which hesuggests (Kivett, 1952, p. 67) the Kelso and Ash Hollow materialsmay eventually be added also.The analysis of additional materials from a variety of sites willbe needed before the task of classification can be satisfactorily done,and the time span of the Woodland occupation of the Central Plainsestimated. At the moment, a fair case can be made for generalchronological equivalence of most of the named or otherwise recog-nized complexes. Spaulding (1949) has directed attention to theHopewellian elements found at the Younkin mound. These includedentate-stamped and rocker-marked pottery, a platform pipe, andthe "general character of the burial" (e. g., disarticulated and some-times fragmentary bones, some primary interment, evidence ofburning, mound construction). A cord-roughened vessel fragmentat Younkin (Schultz and Spaulding, 1948, pi. 29, a), on the otherhand, looks less like Hopewellian than like some of the narrow-mouthed jars from Loseke and Eagle Creek sites (Hill and Kivett,1941, pi. 28, fig. 1; Kivett, 1952, pi. 29, A). Other Younkin items,such as abundant plain and incised bird-bone beads, shell disk beads,and small finely serrate points?features once commonly found bycollectors in moimds of the Manhattan locality?are reminiscent ofKeith Focus mortuary materials. This indirect Hopewellian-KeithFocus correlation, via Younkin, is also hinted at by the finding of atypical Hopewellian rocker-marked sherd in the Woodland level(Occupation B) at the Pottorff site, along with calcite-temperedcord-roughened sherds and large stemmed or comer-notched pro-jectile points. The presence of small numbers of cord-roughenedValley ( ? ) type sherds in Kansas City Hopewellian sites is estab-lished. Spaulding (1949, p. 110) observes that the "probable occur-rence in Lens D at Ash Hollow Cave" of sherds of the "Eagle Creektype" along with Valley Focus pottery (but see Kivett, 1952, p. 67) Wbdkl] an mTRODUCTION TO KANSAS ARCHEOLOGY 555probably connotes contemporaneity; and as supporting evidence hecites the presence in Eagle Creek ware of "a noded and cord-rough-ened sherd with cord impressions above the nodes." Three-way tem-poral equivalence has also been suggested for Sterns Creek, ValleyFocus, and Hopewellian on the basis of reported sherd associationsat the Leahy site near Peru, Nebr. (Champe, 1946, p. 68) ; but thefield observers here have recorded the possibility of redeposition ofmaterials that may have originated in more than one village or campsite (Hill and Kivett, 1941, p. 199) . In the present state of our information, I do not believe that thecross finds and apparent correlations listed above can be acceptedas proof of anything more than a veiy general contemporaneity.They suggest that several culture complexes, each covering a timespan of undetermined length, probably were at times in contact withone another somewhere; but they do not indicate at what period orfor how long in their respective time spans these contacts were op-erating. The depth of the archeological zones in the Kansas CityHopewellian sites suggests some length of occupation; and it hasbeen pointed out by Kivett (1949, p. 282) that some of the westernWoodland sites also have a fairly thick, if not very rich, culturallayer. The relative abundance and wide distribution of Woodlandhabitation sites on most secondary streams of the Central Plainsalso suggest the presence of these little communities throughout aperiod of considerable length.As is to be expected, there are still divergent views regarding theprobable sequence of the Woodland complexes in the region. Spaul-ding (1949, pp. 110-111) suggests that the typologically early ValleyFocus is pre-Hopewellian, "with probable late survival in the areanot strongly influenced by the Hopewellian settlements." Some ofthe other variants, he further points out, "may fill a chronologicalgap between Hopewellian and developed Upper Republican," al-though available stratigraphic and associational data "point to aMiddle or Early Woodland date for all of the Central Plains Wood-land variants." More recently, Kivett (1953, p. 137) has suggestedpriority in time for the Keith Focus, with the Valley Focus materialpossibly representing "a Hopewellian influence on a simple Woodlandcomplex such as the Keith Focus."Between Kansas City Hopewellian and the dominant prehistoricpost-Woodland complexes of the Central Plains, such as Upper Re-publican, I see no specific similarities that can be regarded as evi-dence of contact or connection between the two horizons represented.Like the inferre-d semisedentary semihorticultural community living,such similarities as exist involve widespread elements of little use indemonstrating direct relationships. Essentially, I believe, the same 556 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174holds as between Valley Focus and the post-Woodland village cul-tures. On the other hand, Kivett (1953, p. 136) has pointed out "some basic similarities" between some Upper Kepublican and KeithFocus Woodland burial practices. These include primary and sec-ondary burials in a basinlike ossuary pit, and the occurrence as mor-tuary accompaniments of shell disk beads, conch shell ornaments,Olivella and gastropod beads, flake knives or scrapers, end scrapers,shaft wrenches of deer-leg bones, undecorated tubular bone beads, andhammerstones. I have the impression, too, that Keith Focus may in-clude more small projectile points perhaps trending toward the smalltriang-ular side-notched Upper Eepublican form than does ValleyFocus, where somewhat larger stemmed and corner-notched types re-mind one of Hopewellian forms. I am not altogether convinced bythe foregoing items; but they suggest that Keith Focus may be closerto Upper Kepublican, Valley Focus to Hopewellian. Where StemsCreek would fit into such a scheme, I am not prepared to say ; but theLoseke Focus-Eagle Creek-Missouri Blufl's materials, I think, wouldmost nearly qualify for filling the gap between Middle Woodlandand post-Woodland horizons north of the Republican, or perhaps thePlatte, River.Many of the now puzzling things about the Central Plains Wood-land materials, including focus interrelationships, will doubtless beclarified when radiocarbon dates or other acceptable devices providea sound basis for chronology. At present, there is but one publishedWoodland radiocarbon date for the Kansas-Nebraska region?that ofA. D. 611?2i0 years, for charred timbers from Woodruff ossuary(Wedel and Kivett, 1956, p. 414; but see also p. 619, this bulletin).This suggests a somewhat earlier period for Woodland here than hasbeen previously estimated, and precedes by three to five centuries theestimates for the subsequent Upper Republican occupation of theregion. If, despite the presumed time spread of several centuriesbetween Woodruff, as dated, and Upper Republican, there is a directline of relationship between Keith Focus and early Upper Republican,we may liave some measure of the persistence with which the simplewesterly Woodland cultures clmig to their Spartan way of life.For the most part, the Central Plains complexes discussed aboveare probably of the Middle Woodland period, as Spaulding (1949)lias pointed out. Evidences of earlier Woodland or immediately pre-Woodland occupations of the Kansas region are meager, and the in-vestigations of the National Museum were not directly concerned withany materials that can be certainly so regarded. That these are verylikely present, however, is suggested by various circumstances. Prob-ably the most substantial of these circumstances at the moment is theYounkin mound (Schultz and Spaulding, 1948) where long carved WkdblI AN INTRODUCTION TO KL\NSAS ARCHEOLOGY 557bone hairpins, a cut and polished human libuhi, paired clamshellpendants, and perhaps other items are regarded as indication ofArchaic affinities. Spaulding (1949, p. Ill) has suggested that thismay be evidence of an Archaic tradition that persisted in northeasternKansas into the Middle Woodland period.There have been sporadic finds of boatstones in eastern Kansas thatmay have a similar early affiliation. One in the national collections(fig. 104, h) is from Labette County, and I have seen other fine speci-mens in private collections from the vicinity of Independence, Mont-gomery County, and Elgin, Chautauqua County. Their site associa-tions, other than that they were local finds, are unknown to me. Suchobjects have been reported also from preceramic (?) burials inMcPherson and Scottsbluff Counties, Nebr. (Champe, 1949 b, p. 16 ; Kivett, 1952, p. 65), as Avell as from a site in Sherman County, Nebr.(Hill and Kivett, 1941, p. 219) , where grit- and calcite-tempered cord-roughened sherds also occurred. They are evidently early in theKansas region, but how early remains to be determined. aFigure 104.?Miscellaneous specimens from eastern Kansas, a, Effigy pipe of induratedclay, 2M miles south of Muscotah, courtesy of F. Ward; b, boatstone from mound nearOswego (USNM 232588).LATE WOODLAND-MISSISSIPPI COMPLEXESThe Central Plains equivalent of the Late Woodland-Mississippitransition in the Eastern United States is represented by several pre-historic semisedentary pottery-making cultures in the Nebraska-Kansas region. Sites are numerous and widely distributed, with indi-cations that they extend as far west as Colorado and Wyoming.Village sites are characterized especially by the remains of substantialearth lodges, along with quantities of refuse and artifacts in stone,bone, shell, pottery, and other materials. They evidently indicate amuch more stable pattern of settlement than is inferrable from the 558 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174 earlier western Woodland sites. There is direct evidence that subsist-ence was based in part on a small-scale, hoe-linked, maize-bean-squash-simflower horticulture and in part on hunting, fishing, and gathering.The general time period has been thought to include roughly the 13thto 15th centuries, but these dates are subject to revision. The bestknown of these manifestations at present include the Nebraska Aspectalong the Missouri River bluffs zone and the Upper Republican Aspectin the loess plains to the west. A third is herein suggested for theSmoky Hill-Blue-Kansas River drainage in eastern Kansas.The Upper Republican Aspect was first studied and is still bestknown in the valley of the Republican River in southern Nebraska(Strong, 1935 ; Wedel, 1934, 1935) . Here it is represented by numer-ous small village sites strung along the tributaries of the Republicanand, to a lesser extent, along the main stream. A second importantarea of occurrence is in the Loup River drainage in central Nebraska,where several variants are represented (see, e. g., Champe, 1936).Closely related materials are suggested in the Solomon, Saline, andSmoky Hill drainages of northern Kansas ; and the upper componentat the Pottorff site in Lane County, described herein, is clearly UpperRepublican. The southward extension of the complex has not yetbeen worked out.Unfortunately, most of the investigations reported to date representmainly limited digging at widely scattered sites?essentially an inten-sive survey?and so they do not provide a complete picture. Muchmore comprehensive in scope were the still unreported operations ofthe Nebraska State Historical Society and Works Progress Adminis-tration in 1939 at Site 25HW6 on Davis Creek, in Howard County,Nebr., where 21 house sites were opened; and the 1948 investigations ofthe River Basin Surveys (Kivett, 1949) and of the Nebraska HistoricalSociety (Wedel, 1953 b, pp. 12-18, 41-42) at Medicine Creek reservoirarea. Full publication of these major projects will increase greatlyour understanding of the Upper Republican Aspect and some of itsprincipal variants.Upper Republican villages were characteristically small, unfortified,and often so unformalized tliat their exact size and limits are difficultto define. There appears to have been a marked preference for resi-dence on small streams, perhaps because suitable building timbermay have been more abundant here than on the main rivers, as seemsto have been the case in the 19th centui-y (Bryan, 1857, p. 475). OnMedicine Creek, the settlements consisted of single-house units scat-tered over the terraces and higher hills at intervals of a few yards toseveral hundred feet, or of clusters of two to four units similarly sep-arated from other small groups or single units. A similar loose clus-tering of habitations is indicated for the Davis Creek area. One or Weubl] .\N introduction TO K.\NSAS ARCHEOLOGY 559more small subfloor cache pits, seldom exceeding 4 to 4^/2 feet indepth, commonly occur in the lodge sites or just outside near the door.About the entrances may be fomid small midden areas; sometimes asingle such refuse area evidently served sev^eral nearby houses.The house sites, averaging 20 to 32 feet across, are generally squareto rectangular in floor plan, rarely circular; they have rounded cor-ners, four main roof supports arranged in a rectangle around an un-lined central fireplace, and an entryway from 8 to 12 feet long openingtoward the east, south, or west, or in an intermediate direction. Nearthe outer edge of the floor area are small postholes spaced at intervalsof 1 to o feet, marking the secondary roof-wall supports. Floorsshow no burning or other attempt at hardening, as in some houses ofthe historic Village Indians. Sometimes the house floors were exca-vated from 6 to 20 inches so that a shallow pit house resulted. Inmany instances, however, the structure was evidently erected essen-tially on the contemporaiy groimd surface from which little morethan the sod had first been stripped ; and the pitlike appearance of theexcavated ruins is attributable to the deposition of a variable tliick-ness of wind-blown soil subsequent to abandonment of the site (Wedel,1934, 1941 a; Kivett, 1950). Such terms as "pit house" and "semisub-terranean" thus are often incorrect when applied to Upper Republi-can houses, and c>annot be used without some qualification.Artifacts are usually much more plentiful than in the earlier Wood-land sites. The potteiy is sand or gravel tempered, fairly hard, andpredominantly gray in color. Vessel exteriors are nearly alwaysroughened with a cord-wrapped paddle; the impressions are oftenclear and well marked, usually vertically oriented, and may be partlysmoothed over. Full-bodied jars are the common form; they havea rounded base, flattish upperbody, and constricted neck. Rims arecharacteristically collared or wedge shaped, or else present a simpleunthickened vertical or outcurving profile. Collared rims usuallybear incised decoration, \^Olich may consist of 3 to 8 horizontal parallellines, of groups of diagonals slanted in opposing directions or alter-nating with blank areas, of crossliatching, of paired lines running zig-zag around the vessel, or of other motifs. The lower edge of thecollar is not uncommonly notched or shaped into nodes or scallops.Handles, lugs, or other appendages are very rare or absent. Vesselforms other than jars seem to have been uncommon, but some bowlsand miniatures are indicated.Nonceramic materials include a variety of stone and bone artifacts.In chipped stone, there are small triangular notched and unnotchedprojectile points; numerous end and side scrapers; thin ovoid or el-lipsoidal blades ; diamond-shaped beveled knives ; straight-shafted andT-shaped drills; chipped celts; and some heavy chert or quartzite484172?59 37 560 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174chopping tools. Ground stone includes longitudinally grooved sand-stone shaft smoothers; irregularly shaped sandstone sharpeningblocks; grinding slabs with elliptical milling surface; manos; stonepipe bowls, either equal arm or with slightly projecting prow, andsometimes with incised zoomorphic decoration ; a few carved humaneffigy heads; pecking and hammerstones ; small pendants; and rarepolished celts. Work in bone consisted of numerous bison scapulahoes; split mammal leg bone and splinter awls; eyed needles; fish-hooks ; edge-slotted knife handles ; shaft wrenches of bone or antler ; split metapodial gouges or fleshers, miserrated; pierced bison phal-anges; ulna punches; dressed cylindrical antler sections; bison spinebeamers; deer mandibles with polished diastema; oblong bone "counters" ; plain tubular beads ; incised bracelets or "bowguards" ofthinly scraped antler or bone ; and possibly deer metapodial beamers.The scanty shellwork is virtually limited to a few disk beads, taperedpendants, and musselshells with serrate edges.The burial complex is not fully known, but apparently includesboth ossuaries (secondary burial) and individual primary interments.The former (Strong, 1935, pp. 103-114) are large basins in whichdisarticulated bones and occasional articulated burials were depositedalong with pottery, bone and shell ornaments, and other it^ms. Theseburials include far fewer finished disk beads than occur in the "shellbead" ossuaries attributed to the earlier Woodland peoples (Kivett,1953, p. 134).The Nebraska Aspect is represented by numerous sites along theMissouri River in eastern Nebraska, western Iowa, and northeasternKansas (Strong, 1935, pp. 250-267; Bell and Gilmore, 1936; Hill andCooper, 1938; Cooper, 1940). It has not yet been reported fromnorthwestern Missouri, although I have seen sherds apparently ofNebraska Aspect type in the Kansas City locality. Its presence inDoniphan County, Kans,, has long been known, and the pottery andother materials recovered in 1937 by the National Museum party intwo house sites near Doniphan falls readily into the pattern as estab-lished m Nebraska. There are variations in the complex from localityto locality, but these have not yet been worked out in the necessary-detail.Nebraska Aspect sites occur commonly on the narrow ridges andbluffs along the Missouri, overlookmg the flood plain; and to lesserextent on the bluffs of the Elkhom and lower Platte. They are foundusually as straggling lines of house pits, rarely in groups, and appar-ently nowhere as compact settlements. Isolated house units alsooccur. The lodge sites are marked by pits ranging in size around45 feet in diameter ; Stems reports one of 60 feet ; and we were shownone in Doniphan County that measured well over 75 feet. The Wkdel] an introduction TO KANSAS ARCHEOLOGY 561houses themselves are square or rectangular, with rounded corners,four main roof supports about a central fireplace, and have a longentryway. They were excavated to depths of 21^ to 4I/2 feet from thecontemporary ground surface; and, unlike the Upper Republicanstructures, they seem everywhere to represent purposeful semisubter-ranean constructions, i. e., true pit houses. Cache pits occur withinthe houses, often in some numbers.The burial complex is virtually unknown. On the basis of incom-plete information, Strong suggests that scattered bones and partiallyarticulated skeletons in natural hillocks or mounds associated withNebraska Aspect potsherds may belong to the culture. If morespecific data have since been gathered, I am unaware of theirshowings.Pottery is tempered with sand or crushed granite, rarely with shell ; and the proportion of aplastic varies widely from sparse to abundant.In color, there is a higher incidence of buffs and reddish browns thanin Upper Republican. The basic shape is a globular jar with con-stricted neck and recurved rim. Simple unthickened rims predomi-nate, but there is a small proportion of thickened collared rims withincised decoration like that on Upper Republican collared rims.Strap and loop handles occur much more frequently than in UpperRepublican. Vessel exteriors were either left smooth or else werecord roughened; and the latter were frequently wiped over so thatthe cord impressions have been partially or largely obliterated. Thecord impressions, when present, are rarely as clear-cut as in much ofthe Upper Republican ware. Incised body decoration usually occurson shell-tempered pieces, often with polished surfaces, that are e^d-dently due to alien influences originating in Middle Mississippi(Strong, 1935, p. 255; Wedel, 1943, p. 213). Also of pottery are anumber of effigy heads somewhat reminiscent of those on MiddleMississippi pottery bowls, and many bent tubular pipes which com-monly have incised or modeled decoration.Nonpottery remains are fairly plentiful and varied. Chippedstone includes numbers of fairly heavy triangular projectile points,as well as small notched and unnotched forms ; end and side scrapers ; flake knives; ovoid and diamond-shaped knives, some of the latterbeveled; celts; and T-shaped drills. Ground stone consists of lon-gitudinally grooved shaft smoothers; variously shaped sharpeningstones ; polished celts ; discoidal hammerstones ; anvil stones ; possiblystone trowels or modeling implements for pottery making; and oc-casional stone pipes that are sometimes carved in part to suggestanimal and other forms. Bonework includes bison scapula hoes;socketed and unsocketed cylindrical antler sections; splinter andmammal bone awls; shaft wrenches of bone or antler; ulna picks; 562 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174eyed needles, rare ; conical antler knapping tools ; gouges or flesliers,untoothed; grooved deer jaws; fishhooks, including very large formsnotched at the bend ; toggle-head "harpoons" ; bone beads, pendants,and bracelets or bow guards; deer jaws broken and polished at thediastema; knives or scrapers made of scapula fragments; and oc-casional digging tools made of bison frontal with attached horncore. Work in shell is rather more common than in Upper Repub-lican ; it consists of tanged spoons, perforated musselshell hoes, pend-ants with incisions suggesting birds or fishes, triangular gorgets,tapered pendants, disk and short tubular beads, and simple pendantsfashioned by a deep cut across the umbo.From almost the very beginning of organized archeological re-search in the Central Plains in the 1930's, it has been apparent thatno sharp line separates Upper Republican and Nebraska Aspectdistributions. As several workers have noted (e. g., Cooper, 1940,pi. 12; Gunnerson, 1952), there is a small proportion of incised col-lared rims of Upper Republican type in Nebraska Aspect sites alongthe Missouri; and conversely, Upper Republican materials fromcentral Nebraska sometimes include rim effigies, vessel handles, andother features that are much more characteristic of eastern Nebraskawares. Northward, there are sites that can be fairly regarded as alate prehistoric merging of the two (Cooper, 1936). Much of thearea in Nebraska and northern Kansas that lies between "classic"Upper Republican and Nebraska Aspect localities is occupied byculturally intermediate sites whose bias roughly reflects their re-moteness from one, and proximity to the other, of the two namedcomplexes (Strong, 1935, p. 168; Hill and Cooper, 1937). Thisunhappy circumstance, though not unexpectable, complicates theproblem of pigeon-holing the many local variants in taxonomic terms.The transitional character of many of these remains is, I suppose,one reason for a recent move to jettison the two existing aspect namesand to replace them with a single designation?Aksarben (see Ste-phenson, 1954, p. 19). The basis for this proposed revision has no-where been presented, so far as I am aware; and since I am notpersuaded that there are any real gains in making such a change, Icontinue to regard the Upper Republican and Nebraska Aspects asvalid categories and their designations as above reproach.The very limited information available for northern Kansas sug-gests that, as in southern Nebraska, Upper Republican materials arewesterly, Nebraska Aspect easterly. The former appear to be mainlywest of the Smoky Hills and north of the Arkansas River; theNebraska Aspect is essentially restricted to the Missouri Valley innortheastern Kansas. Between these two areas, and fanning outto the south, there is evidence of another widespread small-village Wei.elI an introduction TO KANSAS ARCHEOLOGY 563complex that shares its basic traits with the two named aspects andalso shows some noteworthy divergences. Pending more extendedand intensive investigations, I can do little more than suggest at thistime the nature of this complex. I believe it is represented by theWhiteford and Grilling sites described elsewhere in this paper, andperhaps also by the Minneapolis site (Wedel, 1935, pp. 210-237) onthe lower Solomon. It seems probable that when the village sitesnoted many years ago by Brower and others in the lower SmokyHill-Kansas River drainage have been properly investigated, theywill be found to show consistent similarities and strong relation-ships to those hei-e summarized. The "Upper Republican variants"on the Blue River, with cord-roughened sherds and bits of burnthouse daub, perhaps also belong in this grouping. Provisionally,1 suggest that the Grilling and nearby related earth-lodge villagesites be classed as the Manhattan Focus, the Whiteford and Minne-apolis sites as the Saline Focus; and that both be assigned to theSmoky Hill Aspect.Village sites of the Smoky Hill Aspect are generally small, varyingfrom one or two house units on a small creek terrace to a dozen ormore widely scattered units strmig along a creek or one of the largerstreams. There is no indication of defensive works. House sites aresquare to rectangular, with rounded corners, four mam roof supportsset aroimd a central fire basin, and an extended entryway. Cachesoccur inside the houses and perhaps also outside. In the Minneapolis,Griffing, and Whiteford sites, the only ones in which house ruins haveas yet been excavated, the lodges were visible prior to excavation asslight elevations ; and it is evident that the structures were originallybuilt on the sod-stripped ground surface, not let down into it as werethe Nebraska Aspect dwellings.Pottery and other materials are moderately plentiful. The pot-tery is usually tempered with sand or gravel, but crushed sherds andbone are sometimes present. A few bowls are indicated, but largejars were the prevalent form; their height tends to exceed the diam-eter, but some with flattish upperbodies in the Upper Republicanstyle are also present. Rims are mostly simple, imtliickened, and ver-tical or outcurved ; occasional examples show some thickening ajid acrude attempt at a collared effect. Very rarely, the collared rimsbear incising; the lower edge may be scalloped or noded; but morecommonly, the allover cord roughening is carried from the body overthe rim to the lip. On the lower Blue and Kansas Rivers, many ofthe sherds have a characteristic orange-brown color and show a strong-tendency for the surfaces to peel or flake away. Small strap handles,usually arising from the vessel lip and connecting this with the neck,are present. Essentially, the pottery remains are what I have desig- 564 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174nated Riley Cord-roughened on the basis of our admittedly smallsample from the Griffing site.Less is known of the nonceramic materials, but they seem to parallelgenerally those of the Upper Republican and Nebraska aspects.Among chipped artifacts, projectile points are roughly flaked and in-clude triangular notched and unnotched forms ; end scrapers are pres-ent; and knives include thin lanceolate and diamond-shaped forms,the latter commonly beveled, as well as occasional long thin well-chipped curved forms with rounded ends. Corner-tang knives arenot imcommon in the general area, but their association with the cul-ture under consideration has not been established. Ground stone in-cludes longitudinally grooved shaft smoothei*s; polished and chippedcelts, as well as some which are partly polished and partly chipped ; pipes with projecting stem or prow; mullers; and many large flat orslightly hollowed grinding slabs. Work in bone consists of bisonscapula hoes ; awls of split mammal leg bone ; deer metapodial beam-ers ; ulna picks or punches ; plain tubes and tubular beads ; and shaftstraighteners. I suspect that some of the material classed by Broweras "Harahey," including curved bone fishhooks and shell beads, mayalso represent this complex; but the lack of specific provenience forhis specimens, the impossibility of rechecking many of his sites, andthe uncontrolled character of his work discourage miqualified inclu-sion of his data here.The Smoky Hill burial complex is uncertain ; but if the Whitefordsite is correctly included here, primary flexed interments accompaniedby occasional pottery, stone, and shell objects may be added to thetentative trait inventory. The massed burials at this site suggestthat it served a number of nearby communities, and so may or maynot be entirely representative of other manifestations of the SmokyHill complex. That the "Harahey" mounds of Brower belong to thiscomplex remains to be demonstrated, but I doubt that they do.The exact relationships of the Smoky Hill materials to those of theUpper Republican and Nebraska Aspects remain, of course, to bedemonstrated when larger bodies of controlled data are in hand. Thatstrong relationships exist is clear; and I would expect that the SmokyHill sites nearest the Upper Republican and Nebraska Aspect local-ities wUl exhibit stronger similarities than those more remotelysituated. That the Smoky Hill sites can be satisfactorily assigned toeither of these aspects I nevertheless doubt. To me, the ceramic re-mains of the Smoky Hill complex sites, and the meager but provoca-tive hints we have of them yet farther south, suggest a more general-ized tradition than either Upper Republican or Nebraska Aspectwares. It is tempting to see in this the "sort of highest commonceramic factor" out of which developed the "classic" Upper Repub- Wedel] an introduction to KANSAS ARCHEOLOGY 565lican pottery to the west and northwest and the Nebraska Aspect pot-tery to the east?the fonner perhaps by stimuhis from Late Wood-land cultures in the upper Mississippi-Great Lakes area, the latterthrough contacts with Middle Mississippi influences coming up theMissouri. From here, it is only a short, if at the moment hazardous,step to propose that the Smoky Hill cultural complex generally is notfar from the sort of generalized "ancestral culture" which Spaul-ding (1956, p. 79) aptly suggests may have arisen in the CentralPlains in post-Hopewellian times and fathered, so to speak, the UpperRepublican and Nebraska Aspects as we now know them.The geographical range of the Smoky Hill Aspect also awaits fur-ther study. Cord-roughened sherds with the characteristic orange-brown color, friable surface, and simple or slightly thickened cord-roughened rims occur widely in eastern Kansas. This sort of potteryis scattered along the Smoky Hill-Kansas River valley from Salinaeastward, and occurs on many of the lesser tributaries as well. I haveseen it, or closely similar material, from the vicinity of Ottawa, fromsites in the upper Cottonwood drainage, in Sedgwick County, fromsouth-central Kansas, around Humboldt in Allen County, and invarious localities in southeastern Kansas. According to my fieldnotes, it occurs in many of these localities in direct association withlumps of burned grass-impressed clay that almost certainly are fromearth-lodge remains. A sustained attack on these obscure, but seem-ingly abundant, remains is needed to permit clearer definition of theirdistribution south from the Smoky Hill-Kansas River locality and toclarify their nature and variations.The suggested southward distribution, as down the Neosho drain-age into Oklahoma, should certainly be worked out. As has beennoted elsewhere in this paper, locally exotic pottery commonly asso-ciated with the Spiro Focus was found in direct association withSmoky Hill burials at the Whiteford site. This, aside from itsobvious connotations of chronological equivalence, suggests the pos-sibility of more fundamental relationships that should not be en-tirely overlooked because of the wide and obvious differences other-wise between the two manifestations. The house types attributed byOrr to the Early and Middle Spiro components, for example, are notfar, in respect to floor plan, from the typical Central Plains earth-lodge, which appears first in the general time period and culturalstage with which we are here concerned; and the Spiro houses aresaid (Orr, 1946, p. 230) to have been "a few inches to a foot underthe surface." As I have pointed out above, many of the CentralPlains earth lodges were not pit houses but were surface or near-surface structures, apparently like those of the lower Arkansas-RedRiver area. In the Central Plains, moreover, the square or rectan- 566 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174gular earth lodge is associated with hoe tillage, direct evidence offishing, a basically eastern pottery complex, and a way of life gen-erally much more eastern than western in flavor. Wliatever our finalviews regarding its source, the case against a Southeastern derivationdraws no support from the assertion that the Plains earth lodgecannot have come "from the Southeast where the pit house does notoccur" (Jennings, Eeed, et al., 1956, p. 87).Returning to the main thread of our discussion it is immediately ap-parent that the three cultural complexes just summarized?the UpperRepublican Aspect, the Nebraska Aspect, and the Smoky Hill As-pect?share a good many traits and trait complexes. These includeparticularly the following : (1) A subsistence economy divided about equally between maize horticultureand hunting.(2) Semipermanent villages, small, unfortified, and arranged in no apparentorder.(3) Villages located sometimes along the larger streams, and as often on thelesser creeks, adjacent to water, wood, and arable land.(4) Houses built on the ground surface or semisubterraneau.(5) Square to rectangular earth lodges, with rounded corners, four primaryI'oof supports arranged around a central fireplace, and extended entryway.(6) Bell-shaped and cylindrical cache pits, in and between the houses.(7) Burial by primary single interment, and also in communal ossuaries.(8) Grit-tempered pottery, surfaces cord roughened or smoothed; simplevertical to flared or collared rims, latter with incised or cord-impresseddecoration.(9) Chipped projectile points, small, uunotched, side-notched, or side- andbase-notched.(10) Diamond-shaped and ellipsoidal knives.( 11 ) Chipped end scrapers.(12) Pipes of pottery (bent tubular) and of stone (equal arm or with slightlyprojecting stem) . (13) Scapula hoes.(14) Longitudinally grooved sandstone shaft smoothers.(15) Bone awls of split deer cannon bone.( 16) Arrowshaf t wrenches of bone and antlor.(17) Bone fishhooks.(18) Smooth-bladed fleshing tools of mammal leg bone.(19) Grinding implements of stone.(20) Scanty and simple work in shell.As is well known, some of these characteristics have a wide distribu-tion, spatially and temporally, throughout the Great Plains. Theyinclude, with some modification, the traits listed by Lehmer (1954 a, p.139) as diagnostic of the Plains Village pattern. To these have beenadded others that he regards as distinguishing one of three broad cul-tural traditions within this pattern?his Central Plains tradition(Lehmer, 1954 a, p. 143) , or the Central Plains phase of Hurt ( 1953, p. Weoel] .\N introduction TO KANSAS ARCHEOLOGY 56754). ""^ I have omitted one item?figurines?included by Lehmer andHurt as a Central Plains diagnostic ; and have added two or three that,while not peculiar to this phase or tradition, are more or less regularlyrecurring and have, within the Central Plains geographical area, alimited and significant distribution in time or in space. In otherwords, I have sought to characterize broadly the Central Plains phaserather than to emphasize its distinctiveness with respect to other pro-posed categories, contemporary or otherwise.The relationships of the Upper Republican and Nebraska Aspectsto more or less contemporary prehistoric cultures to the north alongthe Missouri River have been discussed by others (see, for example,Lehmer, 1954 a, 1954 b; Spaulding, 1956) and need not detain us here.Southward, however, the relationships of the Central Plains phase areless w^ell known, though it is clear that there are comiections with sev-eral late prehistoric manifestations in Oklahoma and the Texas Pan-handle. These connections perhaps merit more extended considera-tion than can be given them here, but in any case wdll probably not beclearly understandable until systematic investigations have been un-dertaken at various localities in southern Kansas and northern Okla-homa. Such investigations should be pointed toward clarifying therelationships of the late prehistoric Kansas-Nebraska materials tosuch southerly Plains Village complexes as the Panhandle Aspect ofthe upper Canadian, the Custer Focus of west-central Oklahoma, andthe Washita River Focus farther to the east.The Panhandle Aspect (Krieger, 1946, p. 74), including the Ante-lope Creek (Krieger, 1946, pp. 41-71) and Optima (Watson, 1950)Foci, shares a great many traits with western manifestations of theCentral Plains phase; and its resemblances to Upper Republican inparticular have been repeatedly pointed out. There are noteworthydifferences, of course, that have not escaped notice. The Panhandlehouses diverge in many particulars from the Central Plains earthlodge, though sharing the four-post roof supports, vestibule entry-way, and other features. The artifact inventories are generally verysimilar; but such Panhandle items as bone digging stick heads andtransversely scored bones are conspicuously absent from Upper Re-publican and, conversely, fishhooks and arrowshaft wrenches con-sistently occur in the latter but are absent from Panhandle sites.The Panhandle pottery complex lacks the specialized collared andincised rims of Upper Republican, thus apparently conforming tothe more widely spread and generalized ceramic tradition representedby the Smoky Hill sites of Kansas. I have the impression that the *'' For an earlier, somewhat diflferent, and now untenable use of the term Central Plainsphase, correctly Including however the Upper Republican Aspect, see Wedel, 1935, p. 251,and Strong, 1935, p. 2. 568 BUREAU OF AMERICAN ETHNOLOGY [Bcll. 174chipped flint industry of the Panhandle is rather more varied andextensive than that of Upper Republican; and this, along with thespecialized digging stick heads and scored ribs, suggests a slightlylater time level than that of Upper Republican. Krieger's (1947,p. 143) estimate of circa A. D. 1300-1450 is probably substantiallycorrect (see also Sulim, Krieger, and Jelks, 1954, p. 392) . The Custer and Washita River Foci likewise present artifact as-semblages reminiscent of Central Plains materials. They divergemost widely perhaps in their pottery, which includes various aplas-tics, smooth and cord-roughened surfaces, and deep jars with someflat disk bases. In both foci, however, there are square earth lodgeswith four center posts and a central fireplace, as well as numbers ofcache pits. Bone digging stick heads and "notched or grooved bonessuggesting the musical rasp" are reported for both, the bone arrow-shaft wrench only from Washita River. Socketed scapula hoes haveno parallel in the Central Plains phase, but do occur in southernsites of the later Great Bend Aspect in Kansas. The house complexseems to be close to Central Plains; but otherwise the artifact in-ventory again suggests a somewhat later time level, probablyapproaching the Lower Walnut Focus of southern Kansas.Typological considerations, buttressed by stratigraphic and otherevidence, have long ago established the fact that the Upper Republicanand Nebraska Aspects followed in time such specifically Woodlandmanifestations as Kansas City Hopewellian and the several foci ? Keith, Valley, etc.?of the Orleans Aspect to the west. The generaltime level during which these cultures flourished can be approxi-mated; but unfortunately it cannot yet be said that our knowledgeincludes a well-controlled chronological sequence.Elsewhere, I have estimated (Wedel, 1947 a, p. 150 and fig. 51) thatUpper Republican, as it is known from the western part of its range,probably falls between circa A. D. 1200 and 1500, with later variantssuch as St. Helena in the east. The Nebraska Aspect was placedat circa 1300, with a questionable terminal date of about 1500 and asuggested break that I now question between it and the later Oneota.Substantially the same dating (A. D. 1300-1500) was suggested aboutthe same time by Champe (1946, fig. 17), whose estimate derivedsome support from dendrochronological determinations on charcoalfrom Ash Hollow Cave. More recently, a radiocarbon date of A. D.1176 ?150 years has been obtained (Crane, 1956, Sample M-113) fromcharcoal reported to come from the Woods site, "an Upper Republi-can site [in Clay County, Kans.] in the middle period of the de-velopment of this culture." The cultural associations here have notbeen published, and the nature of the "middle period" UpperRepublican materials has apparently nowhere been set forth. Wedel] an introduction TO KANSAS ARCHEOLOGY 569Two additional radiocarbon age determinations by Crane of theUniversity of Michigan Memorial Phoenix Project KadiocarbonLaboratory and included here by courtesy of E. L. Stephenson, chair-man of the Missouri Basin Chronology Program, pertain to the UpperKepublican Aspect. Charcoal (M-835) from a house post at theCoufal site (25HW6) on Davis Creek, in Howard County, Nebr.,is dated at A. D. 1138?200 years, which "should date the transitionbetween Upper Kepublican and, Nebraska Culture." Another charcoalsample (M-844) from a house post at site 25FT70, on Medicine Creek,Frontier County, Nebr., yielded a date of A. D. 1458?200 years.Both dates fall within the expectable time span; but if additionaldeterminations confirm the greater age of the Coufal site over site25FT70, and by more than three centuries, some reorientation ofcurrent thinking on Central Plains prehistory will be in order.Certain other Kansas data throw some light on this matter ofchronology, if obliquely. It will be recalled that the Wliitefordburial pit, here included in the Smoky Hill Aspect, yielded an effigyclay pipe fragment and a large sherd of Crockett Curvilinear Incisedpottery. The latter resembles pottery assigned by Orr (1946, fig.32, e) to the Middle Spiro component, which also includes squarehouses with four center posts. The Spiro situation, unfortunately,is a highly complicated one and its interpretation has given rise towide diversity of opinion. The problem of cross dating is not simpli-fied by recently published radiocarbon dates from Spiro (Crane, 1956,Samples M-14 and M-54) . Five of these average out to circa 330 B. C.?200 years, and the sixth is A. D. 1316 ?250 years. Only the last ofthese is within a period I would consider likely for contact betweenthe Smoky Hill Aspect and the lower Arkansas area, as indicated atthe Whiteford site.Farther west, the Upper Republican manifestation at Pottorff, likethe better known and more extensively worked sites in Nebraska, hasyielded no Southwestern potsherds that might be useful for crossdating. This is the southernmost site now known to me in whichsuch typical Upper Republican materials as collared pottery, bonefishhooks, and four-post earth lodges have been found. These itemsmay turn up, of course, in the Arkansas River drainage yet farthersouth if diligent search is made. It may be significant, however, thatthe only comparable materials yet reported from southwestern Kan-sas, as at the Pratt site, are divergent. They include preponderanceof cord-roughened pottery, along with other items?for example, bonedigging stick heads, transversely scored ribs, and "rib-edge" awls ? that suggest the Panhandle-Custer-Washita River materials morestrongly than they do Upper Republican. Moreover, this complex,without the preponderance of cord-roughened pottery, is curiously 570 BUREAU OF AMERICAN ETHNOLOGY [Bdli.. 174like the Lower Walnut Focus to which it may be in part ancestral.With it is associated Southwestern pottery of the 1425-1550 period-earlier, in general, than the puebloan trade pieces in Great Bend As-pect sites, and confirming the pre-Wliite time level. If there is trueUpper Eepublican material in this section of the State, it presumablyantedates the Pratt complex.With the general lengthening of our time estimates in light ofdendrochronology and radiocarbon dating, it now seems very prob-able to me that the Upper Republican Aspect, or perhaps a moregeneralized antecedent not far from the Smoky Hill Aspect in nature,may have been in existence as early as the 11th century of our era.Almost certainly, the more conservative estimates with which thisdiscussion began are too late to acconunodate the beginning stagesof the Central Plains phase.The western part of the Upper Republican territory was appar-ently largely abandoned by earth-lodge-dwelling people prior to 1500,perhaps because of horticultural difficulties brought on by widespreaddrought (Wedel, 1940 a, p. 329; 1941 a; 1953 a) or possibly for otherreasons. I find unacceptable Lehmer's view (Lehmer, 1954 a, p. 149)that this abandomnent was caused by the drouglit of 1539-64, andthe further inference (Lehmer, 1954 b, p. 147) that the settlements ofHarahey reported to Coronado "may well have been Upper Republi-can." I know of no evidence, and Lehmer offers none, of Upper Re-publican villages occupied thus into White contact times. On thecontrary, the Nebraska villages contemporary with the Quiviran set-tlements identified in Kansas with Coronado's visit are specificallyLower Loup in type, not Upper Republican.The late date adopted by Lehmer seems open to argument on an-other score also. Strong (1940, p. 382), and I following him (Wedel,1941 a, p. 26, n. 12), suggested long ago that since the transitionalstages between Upper Republican and Lower Loup-Pawnee wereapparently missing in the Platte-Loup drainage, perhaps the connect-ing threads were to be sought outside the Nebraska area. Lehmerhas elaborated this thought lucidly in his Central Plains-MiddleMissouri-Coalescent sequence; but in bringing the Central Plainspeople out of Nebraska after the 1539-64 drought, he has allowedmuch too short a period for their conversion to a Coalescent traditionand subsequent return to the Platte-Loup locality by Lower Louptimes.In my opinion, a 15th-century retraction of Upper Republican ter-ritory in Nebraska and northern Kansas seems much more likely. Forthe putative early 16th-century Upper Republican manifestation atAsh Hollow Cave, Champe (1946, p. 48) has offered an explanationthat seems to me more realistic than the application of that dating Wedel] an introduction TO KANSAS ARCHEOLOGY 571to a general occupation of the region. If any single drought triggeredthe sort of exodus that has been suggested, I believe that of 1439-54is a more likely prospect?or else the cumulative effects of this and theshorter following one of 1459-68 (Wedel, 1941 a, p. 25). In any case,what probably followed the indicated movement out of the westernplains, and the ultimate outcome of the Upper Eepublican andNebraska Aspect complexes when their carriers shifted toward theeast and north, may be surmised from Spaulding's recent monographon the Arzberger site ( Spaulding, 1956 ) . THE GREAT BEND ASPECTThe Great Bend inspect, as defined in this paper, is based very largelyon two groups of partially excavated sites in central and southern Kan-sas. Most of our current information comes from several village sites(Tobias, Thompson, Major, Hayes) on the headwaters of LittleArkansas Kiver and from another (Malone) on Cow Creek, these loca-tions all lying in Rice County. There are at least two large andimportant sites (Paint Creek, Swenson) in the valley of Smoky HillRiver some 20 miles to the east, these actually being more preciselydescribed as on Paint and Sharps creeks respectively, in McPhersonCounty. I have unconfirmed reports of one or two additional sites inthe Smoky Hill valley in the general vicinity of Lindsborg. In thispaper I have classed the Rice County sites as the Little River Focus,to which those in McPherson County should probably also be allocated.A second and slightly deviant cluster of sites on Walnut River, nearits union with the Arkansas at Arkansas City, constitutes the LowerWalnut Focus. Limited surface collections I have seen suggest relatedsites on Grouse Creek to the east ; farther up the Walnut near Augusta ; and probably elsewhere in this section of the Arkansas River basin.The analytical work needed to allocate these scattered sites to one ofthe existing foci, or perhaps to another as yet undefined, remains tobe done.Pottery and other materials from localities outside those specifiedsuggest a somewhat wider distribution for the complex. It is certainlyindicated on the headwaters of Cottonwood River near Marion, wherethere are, or formerly were, village sites of some size and richness ofcontent (p. 497). Westward, there is at least one site on PawneeCreek near Larned, and apparently another, or several, on WalnutCreek near Great Bend. Smith (1949 a, p. 293) has reported a campsite of the culture on the Smoky Hill in Ellsworth County. The CowCreek and Little xlrkansas River drainages in Rice County are lit-tered with sites, in addition to the ones that have had some systematicinvestigation. Small sites without visible refuse mounds, but withtypical pottery and flint work, occur on Sand Creek and other small 572 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174streams near Newton, in Harvey County ; and I strongly suspect thatthis situation will be found to hold for many other lesser waterwaysof the Great Bend lowland, especially where the soils are not toosandy or the terrain subject to recurrent flooding.So far as my present information goes, there appear to be no re-ported village sites of the Great Bend Aspect north of the Smoky HillEiver, west of Larned on the Arkansas Kiver, east of Marion, Butler,and Cowley Counties, or south of the Arkansas. Much of the countrywithin the great bend impresses me as poorly suited to occupancy bysemihorticultural Village Indians, though they undoubtedly huntedthere; but it would seem strange indeed if the lower reaches of suchstreams as the Ninnescah, the Chikaskia, Slate Creek, and Bluff Creek,were entirely overlooked by the Great Bend peoples.The known geographical distribution of the Great Bend Aspect inKansas (fig. 105), as outlined above, may also be expressed in termsof physiographic subdivisions. The sites composing the complex liebetween the Flint Hills upland on the east and the High Plains onthe west, extending but a short distance northward into the DissectedHigh Plains of north-central Kansas. They are essentially restrictedto the Great Bend and McPherson Lowlands as these are delimitedby Schoewe (1949, fig. 22). Their eventual discovery also in theWellington Lowland can probably be predicted. Southward exten-sion of the complex into Oklahoma, where it might be expected tooccur, has not been established.The northern sites that I have designated the Little Kiver Focusare found on streamside terraces or, where they occur near the head-waters of a stream, on sloping hillsides and prairie ridges. None ofthe sites I know is directly on the larger streams, such as the SmokyHill and the Arkansas. In most cases, however, they are in proximityto fairly sure water supplies. The Tobias and Thompson sites, forexample, are situated about as far up the Little Arkansas as live wateris to be found. They are separated by a streamlet one can almoststep over in the summer months ; but from the rocky ledges borderingit at this point there issued, even in the summer of 1954 after a succes-sion of very dry years, a steady seep of cool potable water. Belowthis spot for many miles the Little Arkansas consisted mainly of dis-continuous stagnant pools unfit for domestic consumption. Thesoutherly sites of the Lower Walnut Focus are situated on terracesand on bluffs immediately overlooking the Walnut River. In theMarion vicinity, the sites appear to have been either on flood-freeterraces overlooking the Cottonwood River bottoms or else on bluffswith the river immediately at their base.The principal known sites of the Great Bend Aspect, where culti-vation has not blurred or effaced the surface traces, consist of low Wedkl] AN INTRODUCTION TO KANSAS ARCHEOLOGY 573 Figure 105.?Map showing location of certain Central and Southern Plains archeologicalsites and complexes.spreading refuse mounds scattered in random fashion over areas of3 or 4 up to 40 or more acres. Among these middens, and apparentlyunder many of them, there are usually abundant cache or storage pits.On some sites, these are present by the hundreds ; and they may runto considerable size. Characteristically, they have undercut sides anda bell shape, with constricted neck and a flat or slightly concave floor.Examples have been found with floor diameter of 7 or 8 feet anddepth of 8 to 10 feet. Shallow depressions up to 20 feet in diameterand not over 6 to 12 inches deep suggest house ruins ; but this identi-fication lacks verification by excavation, and we have no actual indica-tion of the house type. Other than the numerous middens and thesmall elevations associated with the "council-circle" complexes, thereappear to be no mounds on the Little Eiver sites. Larger moundsthat may be planned constructions occur on at least one Lower WalnutFocus site, but their true nature remains obscure. I know of noevidence of fortifications, stockades, or ditches at any of the GreatBend sites, nor of any associated burials or burial grounds. 574 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174A notable feature of several of the Little River Focus sites is thepresence of what, for want of a better name, have come to be knownlocally as "comicil-circles." Their nature and purpose remain moreor less obscure, despite extensive t-ests made by our party in one suchcomplex at the Tobias site. They may be described as consisting ofa wide shallow ditch some 30 to 60 yards in diameter, with a moundedcenter. In some cases, as at the Hayes site (14RC3), the ditch ap-pears to have been continuous, with a present depth of 12 to 18inches and width of 12 to 15 feet, and with a well-marked raisedcenter consisting in part of ash and refuse. In others, as at Tobiasand the Paint Creek site, the ditch is discontinuous, consisting of aroughly circular arrangement of five or six shallow elongate depres-sions of A-aiying length and without visible central elevation. Thesecomplexes have been noted on at least three sites in Rice County, allwithin a mile of one another, and at the Paint Creek site (Wedel,1935, p. 240) ; and I once noted what appears to have been a some-what similar arrangement of depressions at the Swenson site onSharps Creek. No known village site has more than one of thesefeatures, and not all sites have them. They have not been reportedat any of the Lower Walnut Focus sites in the Cowley County area ; but it is possible that long-continued cultivation on the sites we sawhas blurred their traces. Our findings in the mound 17 complex atTobias, including discovery of postholes, firepits, caches, and burntwattle-impressed clay in two basins adjacent to parts of a discon-tinuous ditch system, would indicate that some sort of structures oncestood here. I am indebted to Dr. J. R. Swanton, formerly of the Bu-reau of American Ethnology, for the suggestion that these remainsmay represent the sites of temples or ritual centers analogous to thecommunity centers in Caddo villages. One wonders if the largerearthen mounds at the Arkansas City Country Club site were perhapstemple platforms or otherwise served as comparable ritual centers fortheir local community.In comparison with most earlier culture complexes of the Kansasregion, the, Great Bend Aspect has a relatively rich and varied arti-fact inventory. Broken pottery is plentiful at most of the majorsites, especially subsurface; and through excavation, we have a gooddeal of information on the principal characteristics of the ceramiccomplex here represented. For the most part, the pottery is of medi-ocre quality and suggests the product of a decadent industry or,alternatively, perhaps a heavy emphasis on strictly utility wares. Inthe Little River Focus sites of Rice and McPherson Counties, 95 to98 percent of the ware is grit or sand tempered, the inclusions rang-ing from fine to medium in size and usually being present in generousquantities. At Cowley County sites of the Lower Walnut Focus, shell Wbdel] an introduction TO KANSAS ARCHEOLOGY 575tempering is decidedly predominant, constituting about 99 percent ofour sample. Vessel surfaces, usually brown to gray or black in color,were imevenly smoothed, especially on the interior ; and there is littleor no evidence of polishing or burnishing. From 60 to 89 percent ofbody sherds in Little River sites have plain surfaces (type GeneseoPlain) ; and 10 to 40 percent show parallel ridges produced by stamp-ing with grooved or thong-wrapped paddle (type Geneseo SimpleStamped). On the shell-tempered Lower Walnut pottery, virtuallyall sherds are plain surfaced (type Cowley Plain). Incised, trailed,or other neck or body decoration is very scarce ; but decorative effectsof a sort were sometimes achieved on Little River pottery by the ad-dition of fillets, plain or diagonally incised, below the rim, or by rowsor groups of small applique nodes. A small proportion of LittleRiver sherds have cord-roughened exteriors (type Little River CordRoughened (Smith, 1949 a, p. 295) ) ; and another small group includespieces with a red-painted surface (type Geneseo Red Filmed).The usual vessel form is a medium to large jar, from 20 to 40 cm.tall and slightly less in maximum body diameter, with simple verticalor slightly outcurved rim, constricted neck, rounding shoulder, sub-conical or globular underbody, and rounded or flat base. No otherforms have been recognized in Little River pottery of the graywareseries (types Geneseo Plain and Geneseo Simple Stamped) ; but globu-lar small-mouthed vessels are suggested in Geneseo Red Filmed ware.The Lower Walnut complex apparently includes some deep round-bottomed bowls; and probably some miniatures belong to both foci.Flat disk bases are much more plentiful in Lower Walnut than inLittle River sites ; at the latter, there is some evidence that they maycorrelate with the rare shell-tempered pottery. Vertically placed loopor strap handles, two per vessel and usually attached by tenoning, com-monly connect the neck and upperbody. Here again focus differencesseem to be indicated, with the Lower Walnut pottery including a sig-nificantly higher proportion of handles of all kinds. In both foci,the handles sometimes have a small nipplelike or laterally compressedprotuberance in the upper, lower, or both angles of attachment ; andsmall perforate or imperforate lugs sometimes are found. Vessel rimsare usually unthickened and unspecialized ; those from the Little RiverFocus tend toward higher straight forms ; and in both foci thickeningat the lip occurs only where the lip has been decorated. About 40 to60 percent of Little River rim sherds and roughly 30 percent of LowerWalnut rim sherds bear small incised diagonal or punctate units orfinger-impressed scalloping on the lip.The presence of exotic sherds on some Great Bend Aspect sitesshould be noted here. Most abundant, perhaps because they are morereadily distinguished from local and other Plains wares, are Soutli-484172?59^?38 576 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174 western sherds. It will be recalled that late Rio Grande glaze pieceswere found by our party at the Thompson site (Little River Focus)and also at two sites of the Lower Walnut Focus. Smith (1949 a, p.295) reports surface finds of Late Glaze E or F pottery at the Hayessite (14RC3), and of Late Glaze C to Early D sherds from SpriggsRocks (14RC1), the latter group unfortunately without associatedlocal types. From the Tobias site there is a sherd of ChupaderoBlack on White; and an incised tubular clay pipe fragment rem-iniscent of some of the late pipes at Pecos was recovered at the Thomp-son site. Contacts in a different direction are indicated by severalred-filmed sherds from the Thompson site which suggest effigy potteryfrom the Arkansas-St. Francis region; and downriver relationshipsare also hinted at by engraved sherds and by others with fingernailindentations from the Larcom-Haggard and Arkansas City CountryClub sites, both assigned to the Lower Walnut Focus. The sherdwith trailed body decoration and cloistered rim found by Udden atthe Paint Creek site (Udden, 1900, fig. 10) points in still another di-rection. Paste and tempering pretty certainly indicate local manu-facture, but the decoration and rim treatment are typically LowerLoup (Wedel, 1935, p. 246) in all respects.Continuing our inventory of Great Bend material culture, we notenext the abundance and variety of stone and bone artifacts. Chippedstone is especially plentiful, ranking next to potsherds in quantity.Planoconvex end scrapers are very common ; most are small to mediumin size, and practically all have been carefully shaped and finished.Almost equally plentiful are side scrapers in a variety of forms andsizes ; single-edge retouching is the rule but there are many exceptions.Bifacially chipped knives include unbeveled elliptical, ovate, oroblong-rounded forms; and also many specimens with two or morealternately beveled edges. Slender two-edged knives with a pair ofshallow notches opposite each other above a rounded, pointed, or oc-casionally flattened, base are the most common form; and there isevidence also of lanceolate and lozenge-shaped pieces. Projectilepoints are generally small, basically triangular, and always well made ; unnotched forms are decidedly predominant over notched. There area few cruder leaf-shaped, subtriangular, and stemmed points; andfrom the Tobias site came a series of large triangular notched speci-mens duplicating in every respect but size the minority type. Drillsare also plentiful and include several forms. At most sites, expanded-base drills in which a fairly slender shaft widens abruptly to a largebasal flange, are predominant. Plain-shafted medium to heavy drills,occasionally with one end worked to a definite stem, were especiallyabundant at Tobias, but occur elsewhere also. Large coarsely chippedovate to subelliptic choppers are found at most sites, but not in great Wkdel] an introduction TO KANSAS ARCHEOLOGY 577 numbers. They sometimes have broad shallow notches that suggestplaces of attachment for a handle. Excepting the choppers, the chip-ping on all these implements was usually done with care and skill, andthe industry compares very favorably with that in any ceramic horizonof the Plains. Many of the artifacts were fashioned from bandedfossiliferous Florence flint, a colorful and attractive stone obtained inthe southern Flint Hills. Chert, chalcedony, agate, and other stonewas also employed, and there is a little obsidian from New Mexico orelsewhere to the west.Ground and pecked stonework also occurs in considerable amount.Grooved mauls are present in a variety of sizes and shapes, rangingfrom beautifully fashioned cylindrical specimens with flanged grooveto slightly altered cobbles whose principal modification is a shallow,sometimes discontinuous, groove for hafting. They include most ofthe large and small forms that have been variously described as stakedrivers, pemmican pounders, and club heads. Milling slabs or metatesup to 26 inches in length are of two general kinds; some have anelliptical depression on their upper surface resultmg from a rotarygrinding motion whereas others are flat-surfaced and surely indicatea to-and-fro action. One- and two-hand manos are present, as arediscoidal or bun-shaped grinding or rubbing stones. Particularlyplentiful are fragments of shaped longitudinally grooved sandstoneshaft smoothers of the type generally used in pairs. Rectangular orvariously shaped blocks of sandstone with irregular, incomplete, orcrisscrossed grooves doubtless served for sharpening awls or othersmall pointed implements. There are occasional deeply groovedblocks of limestone which I take to be shaft polishers ; others, also offine-grained stone and with numerous grooves irregularly disposedalong their edges, may be sinew-stones. Pipes are almost always ofstone and occur in several varieties. Highly typical is an L-shapedpipe with tall cylindrical or slightly bulbous bowl and a short stem.Much less common are a low bowled form with projecting stem anda simple tubular form. Some of these pipes are almost certainly ofcatlinite ; more are of a very fine-grained red sandstone that is easilymistaken for the Minnesota stone ; and I have seen at least one tubularspecimen of steatite or a closely related material. Perforate and im-perforate sandstone disks up to about 3 inches in diameter and ofuncertain use came from the Tobias site, as well as various concretionsand pseudomorphs. The sandstones and limestone used by the GreatBend peoples were probably available at no great distance; but theSioux quartzite and other tough crystalline stone used for mauls wasevidently carried in from the glaciated region of northeastern Kansas.Catlinite and turquoise were clearly imported ; and so, probably, weretwo greenstone celts found at the Elliott site in Cowley County. A 578 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174cupstone from this site may also be a trade piece. Hematite andocher occur everywhere as pigments.Work in bone includes numerous bison scapula digging tools at allsites, but with noteworthy focus variations. In the Little River Focus,the great majority have had the distal extremity or head of thescapula removed. In the Lower Walnut specimens, much or most ofthe head remains and there is either a socket bored from the glenoidfossa into the neck of the bone or else a deep groove runs down thedorsal surface of the scapula. This suggests, of course, differentmethods of hafting, the Lower Walnut specimens evidently havingbeen mounted in a fashion similar to the socketed digging stick headsand scapula hoes of certain Oklahoma-Texas culture complexes. Awlsare of several types, the most common being those fashioned from theedge of the neural spine of the bison and having a triangular or cir-cular cross section. Flat awls split from large mammal ribs are nextcommonest in Little River sites, but we found none of these in theLower Walnut area. Awls of split mammal legbone occur every-where, but in relatively small numbers. Large pointed daggerlikeobjects of split rib occurred in Little River sites, but not in LowerWalnut. Bison rib arrowshaft wrenches were found by us only inthe Little River area; but two made of deer tibia came from theElliott site in the Lower Walnut district. Transversely scored mam-mal ribs ("musical rasps" or "tallies") occurred several times in eachfocus, as did wedge-shaped paint applicators and other objects ofcancellous bone, and tubular beads. Hide dressers made of bisonepiphyses are present. The metapodial flesher with serrate or smoothblade was not found by us at any of the sites ; I picked up the serratetip of one at a putatively Great Bend site near Larned, but presentevidence suggests that the type was rare at or absent from most sites.Probably to be included in the Great Bend inventory, though wefound them exclusively at the Tobias site, are a number of well madeand variously shaped four-sided, tapered cylindrical, and cylindricalstemmed objects, all presumably representing projectile points, andseveral smaller bipointed implements. Short blunt awls or polishers,with triangular cross section and evidently fashioned from rib edges,recur at nearly all sites. Worked bison hyoids came from two LittleRiver sites, gaming chips (?) and incised slips of bone only fromTobias.Worked antler was found only at the Tobias site. It includesespecially the following: carefully shaped partially hollowed scoop-like tines, a conical socketed projectile point, a dressed and socketedbasal section (handle?), a small transversely scored and perforatedrodlike object, and a number of narrow strips with bored and/ornotched ends. Eyed pieces that may be needle fragments were found Wedel] an introduction TO KIANSAS ARCHEOLOGY 579 at both Tobias and Thompson. I venture the suggestion that ifand when properly controlled and systematic excavations on a reallyadequate scale are undertaken at other sites, especially in and aroundsome of the "council-circle" complexes, many of the types that nowappear more or less unique to Tobias may prove to be present at othersites as well.Objects of shell are somewhat more plentiful at Great Bend sitesthan in other Central Plains complexes, but again most of our speci-mens come from the "council-circle" complex at Tobias and onlyfurther work will show how truly representatives the sample is. In-cluded are disk beads, whole and cut shell pendants, spoons or scrap-ers, a notched shell, and two tapered imperforate objects. Withpossible exception of the disk beads {OliveUa^ , all worked shellappears to be of fresh-water species.Articles of wood and other perishable substances were, of course,scarce in all our excavations. In fact, the little evidence we haveof work in these materials we owe to the conflagration that destroyedwhatever structure once stood in basin 2, mound 17, at the Tobiassite. Of especial interest is the coiled basket or tray, constructed ona single-rod foundation and presumably used as a container forshelled com. There was also a small piece of twisted two-ply grassor shredded comhusk cordage. A few small rods suggestingarrowshafts and several small slips of cut wood complete the list.Finally, we must note briefly the non-Indian objects that, as gifts,trade goods, or otherwise, originated with white men. Iron objects,either as scrap or as identifiable items, came from three Little RiverFocus sites. "What was evidently a double-pointed awl and an axblade came from Tobias, and several bits of chain mail from Thomp-son. Rolled tubular copper or brass beads were also recovered atTobias, as were the remains of a necklace composed in part of about150 blue glass beads. There are no White contact materials in ourcollections from Cowley County (Lower Walnut Focus) . As any close student of Plains archeology will realize by this time,the Great Bend Aspect traits listed in the foregoing pages include agood many that occur widely throughout the Central Plains, partic-ularly among the semisedentary cultures on the late prehistoric andprotohistoric time level. Relatively few features in its inventorycan be shown to be limited to Great Bend. The "council-circle" com-plex is certainly one of these and perhaps the most noteworthy ; thepotter}^ complex is another. High-bowled L-shaped stone pipes, inthe Central Plains at least, can be regarded as distinctive of GreatBend; and so, apparentl}^, can hollowed scooplike antler tines andperforated sandstone disks. I am inclined to think that stemmedand four-sided bone projectile points are much more plentiful in 580 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174Great Bend than in neighboring cultures to the east, north, and west ; and the chipped stone industry likewise seems more prominent andadvanced. The Great Bend peoples, as judged by our finding ofwell-developed ears of com, of numerous large storage pits, of scap-ula hoes, and of the metate-mano complex, were certainly more ad-vanced horticulturists than were those of the Dismal River Aspect,discussed in more detail in the next section; and they probablyranked in this respect close to their Pawnee kindred of the Lower LoupAspect in east-central Nebraska.The pottery of the Great Bend Aspect, considered as a whole, isreadily set apart from any other in the Kansas region ; and it is alsofairly distinct from other ceramic complexes of the Central Plains.It should be pointed out, however, that a good many plainware bodysherds from Little River Focus sites are not easily distinguishablefrom some of the Dismal River sherds we collected at 14SC1 in ScottCounty. As a complex, the Scott County variant of Dismal Riverpottery differs from Great Bend pottery in the general absence ofsimple stamping, of vessel handles or lugs, of lip decoration, and offlat vessel bases. The generally darker color, inclusion of mica flakes,and smaller sherd size of Dismal River pottery are less trustworthycriteria for distinguishing between the two complexes. As regardsvessel shapes, it should be noted that the one restorable vessel from14SC1 (pi. 68, h) fits about as well into the Little River series as it doesinto the one extant group of Dismal River vessels?that reported byHill and Metcalf (1942, pis. 5 and 6) from the Lovitt site.The resemblances between Dismal River and Great Bend extendto other categories of their culture besides pottery, representing inpart actual intercourse between tlie two groups and in part joint par-ticipation in widespread Plains practices and customs. Shared non-pottery traits include small triangular notched and unnotched pro-jectile points, end and side scrapers, a similar variety of chippedknives and drills, paired sandstone shaft smoothers, grooved mauls,grinding stones, scapula hoes, stemmed bone projectile points, ribwrenches, a similar variety of bone awls, cancellous bone paint appli-cators, epiphyseal hide grainers, tubular bone beads, and rare tubularstone pipes. Also in both are very limited amounts of Wliite andSouthwestern trade or other contact materials.The Great Bend Aspect also shares a number of traits with theLower Loup Focus (protohistoric Pawnee) in east-central Nebraska.Here detailed comparisons are impossible except on very limitedscale, since only a small fraction of the Lower Loup materials avail-able in Nebraska have been published in the desired detail (Dunlevy,1936), and there are otherwise only generalized and certainly incom-plete trait lists extant (Strong, 1935, p. 68; Wedel, 1938 c, pp. 6-9). Wedel] an introduction TO KANSAS ARCHEOLOGY 581With respect to potteiy, the Lower Loup wares have been describedas grit tempered; and examination of several small site samples inthe United States National Museum suggests that there is perhapsless tempering than in most Little River Focus pottery, but withmore frequent occurrence of large sand gi'ains or crushed siliciousparticles. Shell tempering, which predominates at Lower Walnutsites, is very much in the minority at Lower Loup sites. Lower Loupvessel forms show some similarities to those of Great Bend ; but theypresent much more variety and the large jars tend strongly towarda more globular form as contrasted to the often vertically elongateshapes in Great Bend. Simple stamping is far more characteristicof Lower Loup, as are handles in a wide variety of shapes and sizes.Much of the Lower Loup utility ware is relatively simple and plain,and sherds from such pottery would probably parallel closely muchof the Great Bend ware. In addition, however. Lower Loup includesmany rims that are elaborated into thickened or collared form withincised or punctate decoration on the rim exterior, frequent lipdecoration, ornately done handles, and incised or trailed vessel bodies.These features, other than lip decoration, are absent from Great Bendpottery. The attractively done cloistered rims from such LowerLoup sites as Burkett, Gray, and Barcal are likewise generally absentfrom Great Bend. That there are some general relationships be-tween the pottery wares of the two areas is evident. The Great Bendpottei-s, however, seldom progi*essed beyond the strictly utilitarian,whereas those of the Lower Loup went on to a more varied and artisticseries of creations.In the matter of nonpottery traits. Great Bend and Lower Loupapparently share the following: abundance of small triangular un-notched projectile points ; numerous end and side scrapers ; a varietyof unbeveled and beveled knives; several kinds of chipped drills;grinding stones (the metate-mano complex rare or absent in LowerLoup) ; paired sandstone shaft smoothers; grooved mauls and club-heads ; sandstone hones or sharpening blocks ; stone elbow pipes withlow bowl and projecting prow-like stem; abundant scapula hoes, notsocketed in Lower Loup; stemmed bone and socketed antler tipprojectile points; a similar variety of split-rib, "rib-edge," and splitmammal legbone awls; short blunt awls or polishers with subtrian-gular or round cross section ; bison rib wrenches ; transversely scoredbison ribs ("musical rasps") ; wedge-shaped cancellous bone paint ap-plicators ; epiphyseal hide grainers ; bison ulna picks ; narrow slips ofantler or bone, bored and/or notched at one end ; tubular bone beads ; perhaps serrate-bladed metapodial fleshers; catlinite; and limitedwork in shell. Non-native traits shared include small amounts ofWhite contact goods. Conspicuous differences that may be noted here 582 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174include absence at Great Bend sites of tlie remains of the circularearth lodge, of any evidence of fortification of the communities, andof burials. Precise determination of the actual nature and extent ofsimilarities and dissimilarities between the two cultures must awaitfuller publication of data on the Lower Loup complex and itsvariations.Southward, the Great Bend Aspect shows a number of resemblancesto certain late prehistoric complexes in central Oklahoma and north-ern Texas. In Oklahoma (Bell and Baerreis, 1951, pp. 75-83), thesimilarities appear to be closest with the Washita River Focus inGarvin and Grady Counties (Schmitt, 1950; Schmitt and Toldan,1953), less so with the more westerly Custer Focus in Custer County(fig. 105). Sites of the Washita River Focus are situated on stream-side terraces or upland promontories ; like those of Custer Focus, theyare marked by accumulations of cultural detritus, remains of squarehouses, and cache pits. In both complexes, the subsistence economywas based on hunting and on maize-bean horticulture, as in the GreatBend Aspect. There are socketed bison scapula hoes, shaft wrenchesof mammal leg bone, transversely scored bones ("musical rasps"),tubular bone beads, leaf- and diamond-shaped beveled knives, longi-tudinally grooved sandstone shaft smoothers, ground stone celts, split-rib awls, metates and manos, end, side, and flake scrapers, and use ofKay County fusidinid flint for chipped objects?all reminiscent ofthe Great Bend Aspect, and particularly of the Lower Walnut Focus.Plain pottery predominates in both of the Oklahoma foci, but withsome cord-roughened sherds also present ; shell tempering is presentin Washita River Focus, often with limestone, sandstone, and bone.Flat disklike bases occur in both foci, along with occasional handles,lugs, and applique features; and these items, along with the CusterFocus vessel forms described (Bell and Baerreis, 1951, p. 83) as "deepconical jars which may have a small roiuided or flattened base," arealso reminiscent of Great Bend traits.There are also some notable differences between the Oklahoma andKansas materials. Present in the former but missing in the Great Bendinventory are the following: square houses with four center posts,central hearth, extended entryway, and earth or wattle covering ; hoesof bison frontal bone with horn core handle ; socketed bison metapodialdigging stick heads ; rib and deer metapodial beamers ; bone fish hooks ; deer antler headdresses ; small figurines, elbow pipes, and perforateddisks of clay. Conversely, the Oklahoma complexes apparently lackthe following Great Bend items : grooved mauls and hammers ; two-edged side-notched chipped knives; high-bowled L-shaped stone pipes;sandstone disks ; "rib-edge" awls ; wedge-shaped cancellous bone paintapplicators; stemmed and four-sided bone projectile points; conical Wedel] an introduction TO KANSAS ARCHEOLOGY 583 antler tip projectile points; and hollowed scooplike antler tine imple-ments. Noteworthy, too, is the fact that no European contact materialshave been reported from either the Washita River or the Custer Focussites ; and this, together with some of the trait differences noted above,suggests that the Oklahoma complexes mider consideration probablydate from a somewhat earlier time period than do the sites on whichthe Great Bend Aspect has been set up.Farther to the south, we must consider the Henrietta Focus of north-central Texas, which may also be somewhat earlier in time. Krieger(1946, pp. 148-154) has already discussed this matter with insight andin some detail ; but his observations were made before completion of myanalysis of the Great Bend materials and some further comment iscalled for here. As Krieger notes, the house type for neither culture isknown. Both, however, followed a subsistence economy divided be-tween maize horticulture and hunting, with some fishing indicated inHenrietta. Both, likewise, had a rather extensive and varied chippedstone industry, with many, but by no means all, artifact types in com-mon. The shell-tempered Nocona Plain potteryware of Henrietta,which includes some flat-based jars, is certamly remindful of the Cow-ley Plain pottery in the Lower Walnut Focus ; and the socketed bisonscapula hoes are another point of similarity. To the list of sharedtraits given by Krieger (1946, p. 150) can now be added several more,including the following : back and forth grinding on shaped flat slabs(metates) ; two-hand manos; sandstone hones; hematite blocks withscraped and polished faces ; thick greenstone celts with rounded polls ; and Chupadero Black on White trade sherds. Beyond this, the respec-tive trait inventories diverge, with Great Bend exhibiting generallymuch less similarity to Henrietta than it does to Lower Loup, DismalRiver, and perhaps other protohistoric Central and Northern Plainsmanifestations.A good many of the material culture items inventoi'ied for the GreatBend aspect occur also in late prehistoric and early historic phases ofthe sequence in the northeastern Pueblo area of New Mexico. Thus, atPecos, from which an exceptionally complete and admirably describedseries of artifacts of this period has been published (Kidder, 1932),such items as chipped end scrapers, side scrapers, beveled knives, I'ibwrenches or shaft straighteners, cancellous bone paint applicators, andhumerus-cap (epiphyseal) hide gi*ainers, are considered indicative ofrelationships with Indians to the east, that is, in the Plains. For themost part, where stratigraphic provenience could be established, thesetypes were found at Pecos in greatest frequency, and in some casesexclusively, in late Glaze III to Glaze V context (ca. A. D. 1500-1700) . Well-chipped end scrapers, side scrapers, and beveled knives, as wehave noted, are all very common in Great Bend sites. Here, however. 584 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174the two-edged beveled knives often have well-finished round or pointedbases and lateral notches, which are not evidenced on the illustratedPecos specimens. The four-edged knives from Pecos and the two "exceptional specimens" (Kidder, 1932, p. 34 and fig. 16) would notbe out of place in the series of chipped artifacts in our Great Bendseries ; and the plain shafted heavy drills with edges "ground so smoothin service that they have an almost water-worn appearance," the ex-panded base drills, and the occasional scraper-drill combination, arealso found in Great Bend. There is, in fact, a rather striking similaritybetween the chipped stone illustrated and described by Kidder fromPecos and our collections from the Great Bend Aspect sites of centralKansas. Other shared items include : the metate-mano complex ; sand-stone shaft smoothers or rasps ; three categories of awls, fashioned re-spectively from flat sections of split rib, from rib or neural spine edgesplinters, and from dressed mammal leg bone; four-sided projectilepoints, and others with stems ; blunt awls or polishers, subtriangularin cross section ; tubular bone beads ; and transversely scored imple-ments or "sounding rasps'' of bison rib. Further work in Great Bendsites may show, though ours did not, that the serrate-bladed metapodialflesher should be added to the list.Most of the shared traits enumerated in the foregoing paragraphare found also in other protohistoric Plains complexes, as we havealready noted, and are not peculiar to the Great Bend aspect. Theiroccurrence at Pecos and elsewhere along the eastern periphery of thePueblo area thus indicates contact between Pueblo and Plains cul-tures, but not necessarily between the Great Bend peoples and thoseof the Southwest. Yet there were certainly some direct contacts here,since peculiarly Southwestern elements occur in Great Bend sites.These include Rio Grande glaze paint pottery, turquoise, fine-grainedarrowshaft straighteners or polishers, probably obsidian, and verylikely occasional incised tubular pottery pipes.The general chronological position of the Great Bend Aspect has,I think, been acceptably determined. In considerable measure, thisdetermination rests on the recurrence in and on Great Bend Aspectsites of datable Southwestern potsherds. As scattered surface finds,and occasionally as a result of random digging, such materials havebeen gathered for many years at various locations on Cow Creek andLittle Arkansas River in Rice County, and on Grouse Creek in CowleyCounty. Among the Rice County finds (Wedel, 1942, p. 6) are RioGrande glaze paint sherds of Mera's Group C, which was "in stylefor a short time during the last half of the 15th century" (Mera toWedel, letter of August 13, 1940). Others, including a number ofsherds recovered by my 1940 United States National Museum partyfrom pit 2 and mound 2, Thompson site, from pit 14, from mound 17, _ J Wedel] an introduction TO KANSAS ARCHEOLOGY 585and on the surface at Larcom-Haggard, and from pit A, mound 1, atthe Arkansas City Comitry Club site, have been identified by Meraas of his group E and by Kidder (letter of May 6, 1941, to Wedel)as glaze IV ; both agree that the sherds should be assigned to the pe-riod 1525-1650, and Kidder "would prefer to place them prior to1550." At the Tobias site, we recovered one sherd of ChupaderoBlack on White, a long-lived type (13th to 17th century) describedby Mera as of little use for close dating. It will be recalled thatSmith (1949 a, p. 295) surface-collected at the Hayes site (14RC3) a"Pueblo sherd associated with Great Bend pottery. It is of RioGrande glazed ware. Late Glaze E or F, dating from the 17th cen-tury" ; and that he made another collection of Pueblo sherds, but with-out local pottery associations, at Spriggs Rocks ( 14RC1 ) , these beingidentified as "Late Glaze C to Early D, characteristic of the GalisteoBasin and dating from the pre-Spanish portion of Pueblo IV."Sherds from Grouse Creek, presumably surface-collected, were iden-tified by Mera (letter of May 8, 1946, to Wedel) as Pojoaque Poly-chrome, an early 18th century ware.These exotic pieces, then, range through a time span running fromabout the last half of the 15th century to the early 18th century.Those we excavated ourselves and whose exact provenience is knownare all from the more limited period, 1525-1650. In all cases, thepuebloan pieces occurred in or on sites, or else in sherd areas, cer-tainly or very probably attributable to the Great Bend Aspect. Thisrecurrence at widely separated Great Bend Aspect sites of Pueblosherds centering around 1525-1650 is strong evidence that these com-munities were flourishing during the 16th and I7th centuries?thatis, during the period when the Spanish were exploring the upper RioGrande valley and working outward into the Plains toward the north-east. The scant traces of White contact we found in two sites inRice County, including bits of chain mail, are what might be expectedin communities existing at the very beginning of such an explorationperiod. In other words, it seems in the liighest degree probable tome that some of the Little River Focus sites were inliabited when thefirst mid-16th century Spanish explorers reached the region; and Iconsider it quite likely, further, that such sites as Malone, Tobias,Thompson, and Paint Creek were among the Quiviran villages actu-ally visited by Coronado and his men in that memorable summer of1541.There is, of course, another facet to this problem of chronology,namely, the matter of time perspective within the aspect itself. Thewide range in time?some two and a half centuries?represented bythe puebloan sherds suggests that there may be some sites that pre-date the arrival of the Spanish, and others that followed the initial 586 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174 contact. I realize fully that some of the Southwestern sherds mayhave come into central Kansas long after their manufacture on theRio Grande and thus do not necessarily date the sites in or on whichthey were found. Still, there are numerous sites in the Great BendAspect and probably not all were simultaneously inhabited. Moreextended field investigation and, of course, meticulous laboratoryanalysis of materials, will be required before we can be sure whichwere earl}^ and which late.It is not even clear at the moment which focus?Little River orLower Walnut?has priority in time over the other. Smith ( 1949 a, p.295) has suggested that the Major and Thompson Creek sites maydate from pre-Coronado times since no European contact materialswere found at either in course of his investigations. On this basis, thesites on which I have established the Lower Walnut Focus must alsobe considered pre-White, despite our finding there of puebloan sherdsdatable to the 1525-1650 period. Our work in Cowley County, how-ever, was much more limited in scope than that in Rice County, and Iam by no means convinced that we have adequate grounds for judgingtime relationships as between individual sites. Some 60 years elapsedbetween Coronado's entrada into Quivira and that of Onate ; and afterthe latter, there were other long interludes between visits to the GreatBend area by Europeans. Thus, trade or other White materials musthave been extremely scarce in central Kansas for some time after Coro-nado ; and at sites inhabited mainly or entirely during one of the in-tervals between European visits, there may be little or nothing toindicate that such contacts had taken place.With reference to our immediate problem?time perspective withinthe Great Bend Aspect?I can otfer only tentative suggestions whichare highly impressionistic. I am inclined to believe that the LittleRiver and Lower Walnut Foci were in part contemporaneous, as theexotic Rio Grande glaze paint sherds suggest; and that both wereprobably flourishing during the Coronado-Onate period. I suspect,however, that Lower Walnut may have persisted later than LittleRiver, and that it was perhaps the source of the shell-tempered sherdswe found at the Scott County Dismal River site, 14SC1, for which alate 17th- or 18th-century date is otherwise indicated.'*? In part,Lower Walnut may represent in its later stages a southward drift bythe people represented in Rice and McPherson Counties by the LittleRiver Focus sites. Whether any of the Little River Focus sites ac- " Several thickly shell-tempered sherds, including parts of a flat disk vessel base, wererecovered in 1930 from house 1, Gray site (see Dunlevy, 1936, fig. 6), 2 miles north ofSchuyler, Nebr., by a University of Nebraska Archeological Survey party under my fieldsupervision. These were recognized at the time as unusual In the local (Lower Loup)complex. I now believe those pieces originated somewhere in the Great Bend region ofKansas, perhaps as far south as the Lower Walnut Focus sites. Wkdel] an introduction TO KANSAS ARCHEOLOGY 587tually go back to an earlier time than any of the Lower Walnut sites,I am not prepared to say. Neither can I suggest when, where, or bywhat mechanism the Little River Focus peoples, assuming they diddrift southward in or soon after the I7th century, changed from theGreneseo sand-tempered to the Cowley shell-tempered potterywares.Probably the answer to this and other intriguing problems lies in someof the many yet undug sites scattered through the Great Bend lowlandof central and southern Kansas.Acceptance of a general mid-16th century dating for the GreatBend Aspect sites we investigated brings us to my final point here ? the identity of the people who were responsible for it. On this pointI have already set forth my views elsewhere (Wedel, 1942), and Ihave found nothing since that impels me to alter that position.Briefly, I hold that what we know of the distribution of the sites andof their basic cultural uniformity indicates that they were the habitatof a widespread and numerous semisedentary people practicing amaize-bean-squash horticulture, along with much hunting, somegathering, and perhaps a little fishing. There is evidence in the chammail fragments, iron, brass, and glass beads of limited White contact.Turquoise, obsidian, glaze-paint sherds, and probably other items in-dicate relationships with Pueblo groups of the mid-16th century onthe upper Rio Grande. Viewed in light of the etlinohistorical datapresented elsewhere in this paper (pp. 60-62), the archeological andgeographic evidence at hand sustains my conviction that the Quiviraof the 16th- and 17th-century Spanish documents and the sites of theGreat Bend Aspect were the habitat of one and the same people. Thesewere in all likelihood the Wichita. It is gratifying to note that Bolton(1949, pp. 291-295), who painstakingly retraced the route of Coro-nado, has also identified as Quivira the sites classed herein as the LittleRiver Focus of the Great Bend Aspect. In this identification, it is onlyfair to note, both of us were anticipated more than 20 years ago by a "country editor" in Lyons, Kans. (Jones, 1929) . The archeological remains we liave discussed in this section com-prise one of the most distinctive and noteworthy manifestations ofnative Indian culture in Kansas. Before we move on to the reviewof other complexes, it may be of some interest to consider briefly whatcan be inferred from the archeological evidence regarding the GreatBend way of life. The Little River Focus peoples occupied unfor-tified villages, some of considerable extent, scattered chiefly alongCow Creek, Little Arkansas River, and other small streams north andeast of the great bend of the Arkansas. The Lower Walnut peopleslived in large rambling communities, also without defensive works,on or near the immediate banks of Walnut River, and probably onsmaller streams in south-central Kansas. The number of persons in 588 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174these settlements is uncertain, but some of them must have includedseveral hundred individuals. There is no evidence that they used thesemisubterranean earth lodge, although the "coimcil-circles" of theLittle Eiver district apparently include the remains of some sort ofspecialized semisubterranean partly earth-covered structures. Thecommon type of dwelling must have been of perishable materials ina surface or very shallow subsurface construction. The grass housesnoted by Coronado in Quivira would qualify nicely here.Subsistence was primarily by horticulture, secondarily by hunting,with gathering and fishing of less importance. Corn was certainlyknown and included highly developed varieties with well-shapedgood-sized ears bearing 10, 12, and more rows of kernels; it wascultivated in the creek or valley bottoms, with the bison scapula hoeas the principal tool. We may suppose that beans, squash, sun-flowers, and perhaps other crops were also grown, though of thesethere is no direct evidence. Crop surpluses, which were probablyconsiderable, were stored in large underground pits which, whenno longer needed or suitable for storage, were used for the depositionof rubbish. Com was reduced to meal by means of mealing slab andmano; and from the length and size of some of the latter it seemsevident that they were held in both hands and used with a to andfro motion, as among the Pueblos. Hunting was also extensivelycarried on, with the bison as the principal game and the bow andarrow the chief weapon ; whether the lance was in use is not certain.Birds apparently formed a relatively small part of the bill of fare;but in the Lower Walnut sites the turkey, which must have aboundedin the wooded ravines and scrub oak cover to the south, was of someimportance. There is a little evidence of twisted cordage and coiledbasketiy in the Little River Focus. Heavy reliance on skin dressingcan be inferred from the numerous knives, scrapers, light flint drillsor perforators, and bone awls. Wedge-shaped cancellous bone ob-jects, if used as paint applicators as in historic times, suggest gailyornamented parfleches, tipi covers, and the like. The use of skinclothing may also be inferred, but there is only meager evidence ? bone beads, long slender polished tubes, a few shell pendants, tur-quoise, and red and yellow pigments?to suggest the character ofpersonal adornment.Household utensils, in addition to items already mentioned, fur-ther included pots and bowls of a rather drab potteryware, groovedmauls, choppers and hafted axes of chipped stone, shell scrapers,perhaps also rare groimd stone celts and cupstones. The soundingrasp, made of wood among the historic Wichita, may be indicated bysome of the notched or grooved bison ribs. Technology in somelines, as for example, stone chipping, was of fairly high order, that Wedel] an introduction TO KANSAS ARCHEOLOGY 589in bone perhaps less so; and sliell working was relatively unimpor-tant. For much of the chipped stone, use was made of an attractivebanded fossiliferous Florence flint from the southern Flint Hills,this stone being carried in quantities as far north as Rice County;and Sioux quartzite from the drift area of northeastern Kansas wasutilized in making mauls and other heavy duty implements.Concerning the ceremonial, sociopolitical, and other nonmaterialaspects of culture there is little evidence. Specialized structures areindicated for some Little River sites; mounds in the Lower Walnutarea may also have been erected for some ritual purpose or as templefoundations. The burial grounds which might ofl'er some clues tosocial distinctions and other matters remain undiscovered. Pipes,probably used with tobacco, may have had a ceremonial use. Thatthere was a fairly lively intercourse with peoples in remote regionsis indicated by presence of glaze-paint pottery, turquoise, and ob-sidian from the upper Rio Grande, punctate and engraved potteryfrom Arkansas or eastern Oklahoma, and catlinite from Minnesota.Not all of these contacts were necessarily direct, but they indicatethat the people of the Great Bend Aspect were not solely dependenton the resources of their immediate neighborhood.As to the physical appearance of the people, and their resemblancesand relationships to their neighbors and trading partners, we haveno evidence; and clarification on this point must await discovery ofthe burial grounds or of the ossuaries to which previously exposedbodies may finally have been consigned.THE DISMAL RIVER ASPECTThe Dismal River Aspect is represented by pottery and othermaterials from numerous sites widely scattered throughout westernNebraska, western Kansas, eastern Colorado, southeastern Wyoming,and possibly southwestern South Dakota (Gunnerson, 1959). Theeasterimiost known sites are in Harlan and Franklin Counties, Nebr.(Champe, 1949 a) . Farther west, roughly between the 100th and 102dmeridians, there appear to be clusterings of sites in certain localities ? in and around Hooker County, Nebr., on the upper Middle Loup ; inLincoln County, Nebr., around the forks of the Platte ; and in south-western Nebraska in Chase, Dundy, Hayes and Frontier Counties(Gunnerson, 1959). Probably another group could be added inwestern Kansas, since our observations revealed Dismal River sherdsat various points in Scott and Lane Counties, in addition to the villagesite at 14SC1. Westward from this belt, which may reflect localitiesof concentrated search more than it does actual distributions, thereare scattering occurrences to or beyond the Front Range of theRockies. A clustering of sites is suggested in the general area of 590 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174Denver on the upper South Platte. Sherds typical of the complexhave been found at Ash Hollow Cave (Champe, 1946) and in theupper level at Signal Butte (Strong, 1935). All these occurrenceslie west of the 99th meridian and most are also west of the 100thmeridian, so that the culture may with some justice be characterizedas a High Plains manifestation (fig. 105). It should be added thatfor most of the area of known occurrence west of the 101st meridian,there is little information beyond surface sherd samples, some ofthem distressingly small.First recognized in print by Strong (1935, pp. 212, 270), the DismalEiver culture was briefly described by him on the basis of limitedsurface collections from several sites at and near the forks of theDismal River in the Sandhills of west-central Nebraska. Some ofthe sherds from the type locality have since been classified as Wood-land; but the existence of others clearly assignable to the DismalRiver Aspect is established. At the moment, most of our detailedinformation concerning the complex comes from excavations at severalof the Nebraska sites, notably in Chase, Dundy, Frontier, Harlan,and Hooker Counties (Hill and Metcalf, 1942; Champe, 1949 a; Gun-nerson, 1959). To this pool of data can now be added the ScottCounty, Kans., site (14SC1) described in detail elsewhere in thispaper.Dismal River sites occur in a variety of locations?on terracesbordering perennial creeks, around the shores of Sandhill lakes andponds, in blowouts away from visible surface water supplies, in rockshelters and caves?and perhaps more often than not at a distancefrom the larger streams of the region. The settlements evidentlyvaried widely in size, their remains today covering areas of an acreor two to as many as 60 or 70 acres. Surface indications, aside fromscattered artifacts and other detritus, are usually of the scantiestsort?there are no refuse mounds, fortification lines or ditches, housepits, or other remains of structural features, nor have any associatedburials been reported.Excavations in Nebraska have disclosed posthole groupings that areconsidered to be house remains at two sites, 25CH1 and 25HN37.Hearths and presumed floor areas have been noted at a third, 25H021,near Mullen, but here there are no satisfactory posthole patterns inassociation. The more common grouping is a pentagonal arrange-ment, first noted at 25CH1 and subsequently confirmed at 25HN37.In this, the house unit centers around five postholes lying at a radiusof 5 to 7y2 feet from a central unlined hearth. Into these holes, pre-sumably, the main posts were set vertically to be connected at the topby stringers ; poles were leaned against these and were in turn coveredwith a grass or brush thatch. In four of the units at 25HN37, an Wbdel] ax introduction TO KAJS'SAS ARCHEOLOGY 591 additional pair of postholes east of the pentagon suggests an entry-way ; and one unit had 6 center posts. At 25CH1, in addition to thepentagonal unit, there was a 20-foot circle of postholes within whichother postholes were scattered about the hearth, the whole suggestinga somewhat different style of dwelling. The houses, as reconstructedfrom these findings, were probably more or less circular in floor plan,with a diameter varying from 15 to 25 feet. In none of the knownstructures was there any evidence from which an earthen or sod covermight be inferred ; there were no inside cache pits ; and all appear tohave been erected on or just below the ground surface. The practiceof stabilizing posts by wedging them at the base with the joint end ofa large mammal bone has been noted at several of the Dismal Riversites, including 14SC1 and 25CH1. From the excavations to date itis impossible to say how many house units may have stood at one timeor another on the sites we now know or to judge how compactly theywere arranged over the site area.Other features of the village sites include irregular basins of vary-ing size and depth, usually containing a dark fill mixed with refuse,and scattered in apparently random fashion over the area. Thesewere particularly plentiful at 25CH1, occurred also at 14SC1 and25H021, but were rare at 25HN37. They range in diameter from 1 to15 feet, most being between 4 and 7 feet, and in depth from 12 to 51inches. Their original purpose is unknown. Bell-shaped or cylin-drical cache pits, which occur in such numbers in protohistoric sitesfarther east, are exceedingly rare at Dismal River sites.At three locations, 25HN37, 25DN1, and 14SC1, have been foundpits of another sort. From 25 to 36 inches deep, and up to 60 inchesin diameter, these are characterized by a more or less constrictedmouth 20 to 48 inches in diameter, heavily burned walls and floor, anda fill usually containing quantities of ash, burnt twigs and sticks,blackened earth, and often stones, which may be fire-cracked andblackened. The fill may contain occasional artifacts and bone refuse,but their primary purpose was not storage or the deposition of trash.There is every indication that these pits were subjected to intense andprolonged heat. The most logical interpretation seems to be that theywere utilized in the roasting or baking of some vegetal or meat food,heated stones probably being used also in the process ; and thus theyhave come to be identified as "roasting pits." Perhaps "baking pits"would be a more accurate designation. They have not been found inother protohistoric complexes of the Central Plains, but somewhatsimilar features, though without known pottery associations, werenoted by Smithsonian River Basin Surveys personnel in 1948 at An-gostura Reservoir in Fall River County, S. Dak. (Wedel, 1963, p. 22and pi. 8, 5 ) ; and they have also been reported in post-Spanish levelsat Pecos (Kidder, 1958, p. 119 and fig. 33, 6) . 484172?69 30 592 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174Turning now to the Dismal River artifact complex, we may con-sider first the pottery. This is gratifyingly uniform over most of thearea, although by no means without variations ; and it is usually readilydistinguishable from other wares of the region. Generally speaking,it is a fairly thin, hard, dark gray to black ware, tempered generouslywith very fine to medium sand, and exhibiting almost everywhere apronomiced tendency to fracture into small sherds. There is no evi-dence of coiling, and the pottery was presumably shaped by lumpmodeling and by the paddle and anvil method. The few whole orrestorable vessels so far recovered, most from 25CH1, are small tomedium in size, not exceeding 25 cm. in maximum body diameter and27 cm. in height (but see p. 442). Shapes are simple, the commonesthaving a more or less globular to slightly elongate (vertically) bodywith rounded or subconoidal base, constricted neck, and vertical toslightly outflaring rim. Vessel lips are usually simple and unmodified,but may be flattened or slightly thickened horizontally, especiallywhere they have been decorated. There is no certain evidence of bowls,bottles, and other shapes ; and handles, lugs, and other appendages areeverywhere rare or absent. Much of the pottery is plain surfaced;surface roughening with a grooved or thong-wrapped paddle (i. e.,simple stamping) is often present; and very rarely, incising is found.Otherwise, decoration is limited to small diagonal incisions, punctates,chevrons, and other small units on the lip.Simple and uniform though the pottery tradition generally is, thereare local variations that are of interest. The relative proportion ofsimple stamped to plain sherds, for example, seems to show a significantchange from south to north and east in the better-known sites. Thus,at 14SC1, 95 percent of the sherds are plain and less than 1 percent bearsimple stamping; at 25CH1, the respective percentages are 69 and 30.Farther east, at 25HN37, the percentage of plainware drops to 60 andsimple stamping rises to 40 ; and still farther north, in the Sandhills,at 25H021, plain and simple stamping occur in about equal proportions(Gunnerson, 1959) . A similar change can be detected with respect todecoration of vessel lips. At 14SC1, for example, much less than 1percent of lip sherds bear ornamentation; at 25CH1 and 25HN37,decorated lips rise to approximately 15 and 7 percent; and at 25H021,decorated lips are described by Gunnerson (1959) as yet more plentifuland varied.Simple stamping is a northern pottery trait in the Plains area.It characterizes most protohistoric and historic Pawnee wares inNebraska and has an extensive distribution northward up the Mis-souri into the Arikara and Mandan territories. South of the Pawnee,it is much less common on Great Bend pottery from Central Kansas,where plainwares predominate ; and here also the decorative treatment Wbdkl] an introduction TO KANSAS ARCHEOLOGY 593 of vessel lips with punctate or incised elements is much less commonthan to the north. It may be suggested, then, that in these two par-ticulars?simple stamping and lip decoration?the Dismal Kiver pot-tery complex of western Nebraska and Kansas exhibits northerly ornortheasterly relationships rather than southerly.Another variable in Dismal River pottery is the proportion of mica-tempered or micaceous sherds. These pose an interesting problem. At14SC1, approximately 5 percent of the sherds are mica-tempered(Scott Micaceous) ; at 25CH1, a similar ware (Lovitt Mica Tempered)includes less than 1 percent of the sherds ; and at 25HN37 and 25H021,micaceous pieces were extremely scarce. With respect to the better-known Dismal River sites, therefore, micaceous sherds form a dimin-ishing proportion of the pottery from south to north and east. Smith(1949 a, p. 295) points out that Tichy has classified some of Willistonand Martin's potsherds from Scott County Pueblo as "late Rio Grandemicaceous culinary ware"; and he suggests that the mica-temperedpottery of New Mexico may represent diffusion from the Plains intothe Southwest. More recently, Stubbs has kindly examined a seriesof the sherds I collected at the same site and have described elsewherein the present paper as Scott Micaceous. Under date of November 26,1956, Stubbs reports in part that "I would not hesitate to place thesherds from the Scott County pueblo ruin in the same classification as 'Taos-Picuris' There is still a lot we do not know about this ware,but the paste and temper is quite distinctive both in prehistoric andmodern times. I would date your sherds in very late prehistoric orearly historic times."These parallel identifications by Southwestemists and the alreadynoted diminishing frequency of micaceous ware in Dismal River sitesfrom south to north, that is, with increasing distance from the puebloregion, cast serious doubt on Smith's suggestion. Instead of beingan Eastern or Plains type, the micaceous pottery seems almost cer-tainly to have been directly stimulated, if not actually imported, fromthe Southwest. Some of it may have been manufactured on thePlains by fugitives from Taos and Picuris. In line with this proba-bility. Withers has indicated to me his suspicion that much of themicaceous pottery identified in the western Plains as Dismal Rivermay in reality be Taos-Picuris ware. Since Taos micaceous wareapparently goes back as far as the 13th century in the upper RioGrande, it seems likely that some of the western Plains occurrencesof small numbers of "Dismal River" sherds on sites yielding largerquantities of cord-roughened pieces may actually reflect not a DismalRiver occupancy but instead prehistoric contacts between the upperRio Grande and the pre-Dismal River Upper Republican groups. 594 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174Clearly exotic sherds have been recovered in at least one DismalRiver site under conditions indicating association with the local com-plex, and they have been reported from others. They include vari-ous late polychrome glaze and utility wares from the upper RioGrande Pueblos (p. 465) ; and also a few shell-tempered fragmentsthat probably originated at Great Bend Aspect sites in south-centralKansas. So far as I am aware, no Lower Loup or Oneota sherdshave yet been identified in the Dismal River site samples.The tobacco pipes found to date at Dismal River sites include, aswe have seen, a preponderance of straight tubular forms made ofpottery. At 14SC1 and 25CH1, there are also specimens showingsome elaboration of fonn, as in the flattening and widening of thebit, occasional squaring of the bowl, and, especially at 14SC1, in themodeling of collars and lateral winglike or serrate appendages. Thesetraits are alien to the Plains, as is the presence of fine-line incisingthat sometimes alternates with blocks of pmictate or pricked design.There are also indications of tubular stone pipes or "cloud-blowers,"as well as of bent tubular or elbow pipes of stone ; but the latter arepresent in fragmentarj'^ condition only.Work in chipped stone is usually fairly plentiful to abundant. Avariety of raw materials was utilized, chief among them being chertsand jasper, the latter often a yellow-brown stone occurring in thinledges in the Upper Republican drainage and traded or carried widelyin Indian days. Chalcedony, quartz, quartzite, and obsidian wereless frequently used. Implements included large numbers of plano-convex end scrapers, these predominating at some sites over all otherchipped flint objects and sometimes including specimens with smallgraverlike points. Side scrapers, also planoconvex and with one ortwo working edges, are likewise common, as are bifacially chippedknives of lanceolate, oblong, quadrilateral, and indeterminate fonn,sometimes with two or more edges oppositely beveled. Projectile])oints are small and triangular, with unnotched forms usually pre-dominant, and sometimes decidedly so. There are drills of severaltypes, including some with expanded base and light point, othersthat are heavier and sti-aight or cigar shaped, and still others witha lateral tang or tang-s. At some sites, tliere are large coarse choppingor hide-scraping tools, sometimes reminiscent in general fonn of thequartzite scrapers used by tlie 19tii-centuiy Pawnee and other his-toric tribes (Wedel, 1936, p. 76). Much of the chipped stonework,including especially the smaller objects like projectile points, knives,scrapers, and drills, is less carefully vrorked than are the same cate-gories at other protohistoric Central Plains sites as, for example, inthe Great Bend Aspect of central Kansas.Objects of ground and pecked stone occur in lesser numbers andvariety than the chipped. Like those of chipped stone, they .seldom Weuei.] an introduction TO KANSAS ARCHEOLOGY 595 exhibit mucli skill or care. They include irregular but shaped millingslabs of sandstone or limestone, with flat grinding surface (metates) ; smaller grinders or anvil stones, with shallow circular depressions intheir upper surfaces; one- and two-hand manos; occasional groovedmauls, these usually rudely shaped and never showing the excellentworkmanship found on those of the Great Bend Aspect; numerousfragments of elongate boat-shaped sandstone shaft smoothers, usedin pairs; sharpening blocks or hones, also of sandstone; occasionallimestone blocks with polished grooves, suggesting the shaft polishersof the Southwest; simple tubular and elbow pipes; pecking stonesand hammer stones of quartzite or other tough stone. Bits of workedturquoise at one or two Nebraska sites should be noted, as also thepresence of grapliite. Other stone materials include marcasite andother concretionary forms; fossil shells that were evidently carriedinto the villages ; and, of course, various pigment materials, such asocher and hematite.There is also a considerable amount and range of work in bone andantler ; and much of this is competently made as well as serviceable.Bison scapula hoes were evidently present in some number at suchsites as 14SC1, 25CH1, and 25H021, though fragments far outnmn-ber complete specimens and the exact nature of the implements inthe complex is uncertain. Their presence at these sites, wherecharred com was also found, permits the inference that some horticul-ture was practiced by the natives. Awls are of several well-definedkinds, including broad flat specimens made by splitting mammal ribs ; sturdy forms round to triangular in cross section and fashioned fromthe edges of neural spines or ribs ; variously split and gi'ound deer legbone awls ; and finally, splinters that have been sharpened at one end.Punches, polishing tools, or short stubby awls triangular in cross sec-tion also occur. Fleshing tools were made from the metapodial andassociated ankle bones of the bison, and these have both smooth andserrate blades. There are a few socketed conical and stemmed projec-tile points, the latter with round or quadrilateral cross section ; andwedge-shaped cancellous bone paint applicators are present. A frag-ment of transversely scored rib implement of the sort sometimes des-ignated as a musical rasp was collected by Williston and Martin atScott Comity pueblo; and other distinctive types of rib artifactswhose purpose remains unclear came from the nearby midden andfrom Site 25H021. Also in the inventory are rib shaft wrenches,tubular bone beads, eyed needles, bird bone whistles, epiphyseal hidedressers, antler tips with notch or other device for seating or secur-ing a scraper ( ? ) blade, bison ulna picks, and occasional thinlyscraped narrow slips of bone or antler whose use is not known. Thethree-hole flageolet from Scott County pueblo has not been dupli- 596 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174 cated, so far as I am aware, at other Dismal River sites ; nor has thefragment of polished finely incised bone tube from 25CH1.Worked objects of shell are extremely uncommon. They consistprincipally of drilled and shaped oblong bits of fresh-water shell thatprobably represent pendants.Objects and materials of non-Indian origin are not plentiful at anyof the sites, but their consistent occurrence, even if sparingly, is note-worthy. Iron, brass, and/or glass beads have been found subsurfacein association with the native materials in at least six sites, includingthe following: 14SC1, 25CH1, 25DN1, 25FT9, 25HN37, and 25H021.The iron is sometimes so badly oxidized or so scrappy that its originalform is mirecognizable ; but it also includes axblades from at leasttwo sites (14SC1 and 25HN37), and awls from two or more. Brassor copper includes conical and tubular objects and small scraps.These items, besides indicating direct or indirect contact with Whites,testify to the general lateness of the Dismal River complex in theCentral Plains, and thereby add support to the dendrochronologicaldeterminations already mentioned elsewhere in this paper.The aboriginal features enumerated above for Dismal River areessentially Plains in character. Notable exceptions include the in-ferred house type with pentagonal foundation ; the roasting or bakingpits ; and, in large part, the ceramic complex, which certainly showsPlains influence in the simple stamping and lip decoration. Amongthe artifacts listed, incised tubular clay pipes, chipped driUs withlateral protuberances, end scrapers with graver points, certain un-identified objects of mammal rib, and perhaps the antler tip hafts,seem at the moment to be more or less peculiar to Dismal River.Otherwise, the items in bone, and in chipped and ground stone, gen-erally occur also in such contemporary, and perhaps somewhat earlier,Central Plains complexes as Great Bend and Lower Loup. The greatmajority of the artifacts from 14SC1, for example, could easily be lostin our collections from Great Bend Aspect sites. The Dismal Riverartifact inventory is shorter than that for Great Bend, that is, it lackssome of the types found in central and southern Kansas, and theworkmanship on Great Bend artifacts is, on the whole, of superiorquality. The Dismal River complex differs further from Great Bendand Lower Loup in the relative proportion of certain artifact cate-gories it has. It shows, for example, a greater proportion of toolsand implements suited to, and presumably intended for, the hunt andthe products of hunting; and in contrast, fewer items mdicative ofhorticultural pursuits and the storage of crop surpluses.Many of the items Dismal River shares with Great Bend and otherneighboring complexes also occur in late stages of the Pueblo sequencein tlie upper Rio Grande region. At Pecos, for example, the occur- WBDBL] AN INTRODUCTION TO KANSAS ARCHEOLOGY 597 rence of chipped end scrapers, side scrapers, beveled knives, serrate-bladed metapodial fleshing tools, rib wrenches, cancellous bone paintapplicators, and epiphyseal hide dressers is viewed as evidence ofcontacts with Plains Indians (Kidder, 1932, pp. 44, 235, 238, 241) ; andthese types at Pecos occurred either exclusively, or in greatest fre-quency, in the later Glaze levels, which would be roughly concurrentwith the Great Bend, Lower Loup, and Dismal Kiver complexes in theCentral Plains. The frequent occurrence at Pecos of flat split-rib and "rib-edge" awls is again reminiscent of Dismal River ; and the chippeddrills include both straight and expanded base types, as well as atleast one specimen with double lateral tangs which has a counterpartat 25CH1 (cf . Kidder, 1932, fig. 12a, and Hill and Metcalf, 1942, pi. 7,fig. ID) . Stemmed bone projectile points also occur sparingly at bothPecos and the Dismal River sites. That these Dismal River-Pueblorelationships were two-way is indicated, of course, by the often recur-rent presence at Dismal River sites of such Southwestern features asincised and punctated tubular pottery pipes, turquoise, obsidian, boneflageolets, Rio Grande Pueblo trade sherds, occasional fine-grainedarrowshaft polishers, finely incised bone tubes, and perhaps otheritems.Still farther afield, the Dismal River complex shows some interest-ing similarities to archeological materials from the Promontory Pointcaves in north-central Utah. Some 20 years ago, Steward (1937, p.44) noted the distinctiveness of the Promontory pottery from allknown Great Basin wares and suggested the possibility that it "wasderived from some northern Plains people." Other similarities havebeen noted by Hill and Metcalf (1942) and more recently by J. H.Gunnerson (1956), these including such items as end scrapers (includ-ing the tanged or graver-point variant), tubular steatite or schistpipes, bone punches, sandstone arrowshaft smoothers, bison meta-podial fleshers, etc. Both Steward and Gunnerson emphasize the factthat the Promontory peoples were apparently primarily bison huntersrather than seed gatherers, like the Shoshoneans of the Utah region ; and Gunnerson (1956, p. 72) suggests that the Promontory materialmay represent "an early protohistoric thrust by a buffalo-huntingAthabascan group into the Great Basin from the Plains."From the archeological evidence now available, some inferences re-garding the Dismal River way of life are possible. It seems fairlycertain, for one thing, that these people were primarily hunters andonly secondarily tillers of the soil. The artifact inventory shows ahigh proportion of stone and bone tools that were probably or cer-tainly associated with hunting, butchering, and skin-workingpractices. Among these may be included the abundant chipped-stone scrapers, knives, choppers, projectile points, and perhaps 598 BUREAU OF AMERICAN ETHNOLOGY [Bitix. 174drills, as well as the bone awls, fleshing tools, epiphyseal hidescrapers, rib shaft wrenches, paint applicators, and needles.Tools and devices from which horticulture may be inferred are,on the other hand, comparatively limited in both number andvariety?bone hoes, metates, and manos. Absence or great rarityof cache pits at most worked sites suggests either limited cropsurpluses or else storage techniques different from those ofother Plains semihorticultural complexes. The use of roots, berries,and other edible plant products is indicated; but of fishing there isno evidence, either by hooks or other devices among the artifacts orby the refuse bone. The sites themselves suggest some permanencein the presence of fixed dwellings and in the quantities of refuse ; butit is not clear whether year round occupancy was customary. Thesites at which sustained excavation has been carried on may havebeen winter stations of a normally more nomadic people, with sizableaggregates collecting seasonally over a long period of years at suchspots as 14SC1 and 25CH1. On the other hand, the location of theprincipal known sites along a north-south axis near the eastern mar-gin of the High Plains, where maize-growing was less precarious thana hundred or two hundred miles to the west, may imply that thesewere especially favored locations where some horticulture was prac-ticed year after year while other segments of the populace rangedwidely during the spring, summer, and fall months in pursuit of theherds. The quantities of small game represented at some sites, as at25CH1, suggest that there may have been periods, in winter or atother seasons, when the bison were not readily available and foragingwas necessary.Regarding the appearance of the people themselves we know nothingfrom the archeological work; nor can we reconstruct in detail thenonmaterial aspects of their community life. The scarcity of orna-ments and art objects and the apparent emphasis on useful rather thanbeautiful implements may suggest preoccupation with getting alivelihood in an often trying environment. There is no evidence thatthe horse was known, but the dog was present and may have beenused sometimes for food. Contacts with other groups is indicatedby items or practices idejitifiable with the Great Bend Aspect(Wichita) peoples of central Kansas and the Lower Loup Focus(Pawnee) of east-central Nebraska, and by others clearly originatingin the Pueblo region.As has been indicated, the Dismal River sites we know most fullyhave all been dated, either through tree rings or by cross finds ofSouthwestern materials, at approximately A. D. 1700, plus or minus25 years or so. Earlier stages of the culture, preceding 1650, remainundiscovered or unidentified. These late iTth- and early 18th- Wkddl] an LN'reODUCTION TO KANSAS ARCHEOLOGY 599 century archeological materials occui- in the region where, accordingto historical documents, the Spanish from the Kio Grande settlementscame into repeated contact with the Plains Apache around 1700 ; andthe evidence that the Dismal River culture represents the materialremains of the Plains Apache seems convincing (Champe, 1949 a;Secoy, 1951 ; and supra, p. 466) . This identification permits use ofthe Spanish documents in filling in some of the gaps in our picture ofthe Dismal River culture as derived from arclieology. As I see it,the archeological data at hand conflict in no significant way with whatthe docimaents (see especially Thomas, 1935) tell us of the way of lifeof the Plains Apache during the period in question.The material culture of the Dismal River people has been charac-terized by one student as impoverished, their way of life by anotheras dismal. Comparison of the artifacts recovered at 14SC1 and25CH1 with those from Lower Loup (Pawnee) and Great Bend(Wichita) village sites of comparable age in central Nebraska andKansas indicates that the Dismal River people had much of thebasic horticultural and hunting equipment possessed by their moresedentary corn-growing neighbors. It is perhaps true that they hadfewer of these items?that in proportion to the amomit of diggingdone there are significant quantitative differences as between the Dis-mal River people and their neighbors. The Dismal River groupsdwelt mainly in a region of notoriously low rainfall and consequenthorticultural uncertainty; and the absence of cache pits from theirsites may imply a more perfunctory maize gardening and perhapsscantier crop returas than were enjoyed by their more favorably sit-uated easterly neighbors. Horticulture, in short, was a less reward-ing and therefore a less reliable subsistence basis than was hunting.Like any people relying heavily or primarily on the chase, theydoubtless alternated between plenty and scarcity ; and if, in a seasonof scarcity, their crops failed for want of rain the winter conditionof the people may well have been a precarious one. The commentsby some of the Spaniards who had been among the Cuartelejo Apaahebefore 1720 are directly relevant (Thomas, 1935, pp. 157, 161, 173).All this suggests that the Dismal River people perhaps stood inabout the same relationship to tlieir Pawnee and Wichita neighborsas did the short-grass rancher or the marginal farmer of 50 years agoto the Corn Belt farmer of eastern Nebraska and Kansas. The DismalRiver way of life was certainly an austere one, but it appears to havebeen well adjusted to the limitations and opportunities of the environ-ment. By comparison with the struggle for survival waged by theforagers of the Great Basin, I doubt that the life of the Plains Apachewas dismal or that their material culture can be considered chroni-callj' impoverished. 600 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174THE ONEOTA ASPECTThe Oneota Aspect, unlike the Dismal River and Great Bend As-pects, is not in the main a Plains manifestation. The great majorityof known sites lie east of a line drawn along the Big Sioux River,thence down the Missouri River to Kansas City, and southward alongthe Kansas-Missouri boundary. Above the Kansas River, probablyless than half a dozen Oneota sites have been recorded west of theMissouri. Of the four actually reported in print, summarily orotherwise, three (Fanning, Leary, Ashland) are within 25 miles ofthe mainstem ; the fourth (Stanton) is not more than 50 miles distant.'*^The nature and extent of Oneota sites in southeastern South Dakota,where the Big Sioux, Vermillion, and James River valleys might beexpected to fall within the general area of distribution, remains ob-scure. In Missouri (Chapman, 1946, 1952), most of the reportedsites are situated along the right bank of the mainstem in the westernhalf of the State; but if the somewhat deviant sites identified asOsage be admitted to the Oneota category, Vernon County at the westedge of the State would be included. In eastern Kansas south ofKansas River, no tnie Oneota sites have yet been reported.Oneota sites, as contrasted to sites showing possible or probableOneota influence, thus have a scattering distribution along the east-em margin of the Central Plains. These westerly sites exliibit cer-tain interesting variations from those fartlier east. They representa clearly defined and distinctive Upper Mississippi manifestation,the nature and extent of whose influence in the development of lateprehistoric and protohistoric culture complexes in the Central Plainsof Kansas and Nebraska, and on the Middle Missouri in the Dakotas,still await clarification.East of the Missouri River, Oneota sites have a wide distributionin Iowa, southern Minnesota, Wisconsin, and Illinois. In Iowa(Mott, 1938), there is a notable concentration of village and burialsites on the Upper Iowa River in Allamakee and Winneshiek Coun-ties; along the Little Sioux in Clay and Dickinson Counties, andagain in Cherokee and Woodbury Counties; on the Big Sioux inLyon County ; on the Des ISIoines in Polk and Warren Counties ; andon the Mississippi in Muscatine, Louisa, and Des Moines Counties.The Minnesota sites (Wilford, 1955) are mainly south of Minne-sota River, centering in the Blue Earth and Root River drainages,with at least one noteworthy site near Red Wing on the Mississippi.In Wisconsin (McKem, 1945), there are at least three and perhapsfour main areas of occurrence: one along the Mississippi in Pepin,Trempeleau, La Crosse, and Vernon Counties; another on Lake Win- *? See fig. 105. The Fanning site Is reported In this paper. The Stanton site Is underanalysis by Dolores Gunnerson. For Leary, see Hill and Wedel, 1936 ; for Ashland, HIUand Cooper, 1938, p. 267. Wedkl] an introduction TO KANSAS ARCHEOLOGY 601 nebago, the nearby Fox and Wolf Kivers, and. at the southern tip ofGreen Bay ; perhaps a third on Grand River in Green Lake Comity ; and a fourth?still undescribed?around Lake Koshkonong and theupper Rock River in the southern part of the State. In Illinois, theBlue Island or Huber site in Cook County is safely classed as Oneotaand Oneota influences have been noted at the Fisher site in WillCounty, at the Crable site in Fulton County, in Jo Daviess County,and at Pere Marquette State Park in Jersey County.Despite its widespread occurrence and demonstrable importance toUpper Mississippi Valley prehistory, current information on theOneota Aspect is distressingly uneven and incomplete. Some pre-liminary reports, partial analyses, and summary statements are avail-able, and there is no lack of speculations as to its origin, age, identity,and role in midwestem prehistory. Unfortunately, however, for theseveral variants known to exist among the large and important siteseast of the Missouri, there is not a single comprehensive publishedreport detailing the results of extended systematic excavations at adefined commimity, that is, at a site where both living area and burialgrounds are represented. The nature, size, and number of dwellingunits has nowhere been established; expert identification of the ani-mal and vegetal refuse recovered at habitation areas, by which eco-logic adjustments from one locality to another might be assessed, havenot been made or remain unpublished; and in several instances, theextant materials are largely surface finds or burial accompaniments,and thus inadequately illustrate the material culture complex bywhich the people carried on their daily routine of living. All thismakes more difficult the task of outlining the material culture com-plex of the Oneota and its variants.The Oneota Aspect is represented by both village sites and burialgrounds. Areas of habitation usually vary from an acre or less upto 40 acres or more ; at the Utz site in Missouri and the Leary site inNebraska, occupational refuse covers areas of more than 100 acres;and the Hartley-Lane site on the Upper Iowa may be as large orlarger. East of the Missouri, no direct evidence as to house typehas been found and structures of perishable materials are inferred.In Nebraska and Kansas, there was some use, at least, of substantialearth lodges. Refuse-filled pits, probably used originally for storage,are characteristic; in the eastern sites these were commonly bowlshaped, whereas in the west the deep cylindrical or bell-shaped formswere characteristic. Fortifications in the form of stockade ruins,walls, or ditches are nowhere indicated. Earthen enclosures have beennoted at Oneota sites on the Upper Iowa, at the Blood Run site inLyon County, and at Toolesboro in Louisa County; but their as-signment to the Oneota culture is open to question. Other than low 602 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174middens, there are no mounds certainly attributable to the Oneota,unless those in Miimesota from which Oneota burials were takenare correctly identified with the Orr Focus (Wilford, 1955). InIowa burials were often made intrusively into mounds of earlierorigin. Interment was usually singly, extended, and in the flesh,commonly with pottery or other associated artifacts. At the largersites, considerable quantities of bone, shell, broken pottery, workedstone, and other village detritus indicate prolonged and fairly in-tensive settlement. This was clearly based in part on maize horti-culture, in part on hunting, fishing, and gathering.Artifact materials include considerable quantities of pottery, almostexclusively shell tempered and with smooth surfaces. The basicform was a globular jar, with rounded base, somewhat constrictedneck, and vertical or flaring rim. Two to four oppositely placedstrap or loop handles conmionly connected the rim and upperbody.On the upperbody and rounded shoulder, groups of parallel trailedor incised lines, in various patterns and often associated with punctateelements, formed a decorative zone. Round or elongate punctateswere used in various ways?as fringes to trailed lines, in blocks be-tween lined areas, or under handles. Vessel lips commonly carriedsmall punctate units, notching, finger impressions, or scalloping ontheir surfaces.Other artifacts included an abundance of small unnotched tri-angular chert projectile points, end scrapers, well chipped ellipticalor diamond-shaped knives, a large variety of irregularly shapedknives and scrapers, a few drills, and bipointed objects that may befish gorges. Ground stone includes longitudinally grooved sand-stone arrowshaft smoothers, celts, occasional axes, mortars, roundor oval mullers, discoidal hammerstones, disk and other forms ofpipes, and, in some localities, inscribed catlinite tablets. In bonework, there are many awls of various types, but mostly of split orunsplit mammal leg bone; tubes, either plain or incised; a few largemat-sewing needles; and scapula hoes. Antler objects include sock-eted conical projectile points, flakers, and perhaps other items. Thereis some work in shell, chiefly beads and spoons; and also, especiallyin the northeastern sites, some native copper tubes and sheet orna-ments. "White contact materials are found in some, but not all,Oneota sites.The Oneota Aspect, superficially considered, shows remarkable uni-formity throughout most of the area of its occurrence; but itsmanifestations are by no means identical in their cultural content.Analysis of the data extant, uneven though they are, reveals the ex-istence of a number of variants. These are based chiefly on consistentlocalized differences in pottery, and to some extent in other categories VVkuel] ax introduction TO KANSAS ARCHEOLOGY G03 of material culture. Some of the differences undoubtedly representunequal or inadequate sampling; others may reflect chronological,geographical, or environmentally conditioned variations. Stillotliers, however, by their regular recurrence and limited associationwith other items, probably indicate valid group distinctions. In anyevent, since the observed variations east of the Missouri are ap-parently matched by significant variations to the west, some noticeshould be taken of the focus differences now" believed to exist.Of first interest is the Orr Focus, which comprises a highly con-sistent series of materials in northeastern and extreme northwesternIowa, southeastern Minnesota, and across the Mississippi in adjacentWisconsin. Surface remains from Clay and Dickinson Counties,Iowa, in the upper Little Sioux drainage, should probably be included.The materials from the Upper Iowa River sites show substantialidentity except in respect to White trade goods?their absence orpresence, and their amount when present. To the complex here rep-resented, Keyes (1927) originally assigned the name "Oneota cul-ture," after the term by which the Winnebago are said to know theUpper Iowa River. Especial importance attaches to the concentrationof village sites and burial grounds of this locality because it can besafely ascribed on documentaiy and archeological grounds to the IowaIndians of the early 18th century (Keyes, 1927, p. 224; Griffin, 1937 a;Mott,1938).Most of the available information is derived from seven village andburial sites in Allamakee County, Iowa, and from limited excavationsat a few sites in nearby Wisconsin and Minnesota. The village sitesare usually situated on streamside terraces and range in area from afew acres up to 60 or more. Refuse-filled pits, usually bowl shaped,are fairly common; and there are considerable accumulations of cul-tural detritus over the area of former residence. Burial grounds maylie in close association with the village area, or at a little distance onelevated spots and hills. The single extended burials were sometimescovered with stones, with mounds reported only for Minnesota. Semi-seated burials, with head and shoulders raised slightly, occur inWisconsin and occasionally on the Upper Iowa. Pottery and othercultural objects are often associated with the burials.Orr Focus pottery (Keyes, 1927; Mott, 1938; Griffin, 1943, p. 287;McKem, 1945, p. 145) consists chiefly of globular jars of the basicform described above among the aspect traits, but a squatty ellipticalvariant is highly characteristic. Quite common are handles, usuallyof the broad strap variety. Many of these widen strongly towardtheir upper end; and "tails" or "crests," like those on handles fromLeary, or a prominent medial ridge, such as occurs at Fanning, arefound on Upper Iowa sites. Rims are straight, vertical or slightly 604 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174flaring, and the lip surface typically bears impressed or punctatedecorative elements. Body decoration consists primarily of groupsof parallel vertical or slanting trailed lines, combined with some useof punctates ; on the Upper Iowa, and to lesser extent in Wisconsin,punctates are frequently massed in triangular areas on the body orin the spaces below vessel handles. Broken lines or rows of shortdashes are present in some number. The chevron motif is found inIowa and Minnesota sites, but less commonly than in Wisconsin. Asa rule, the trailed lines are much narrower than are the spaces be-tween lines; and decoration occurs on about 20 to 25 percent of allbody sherds, A few sherds typical of the Lake Wiimebago Focushave been found at Orr Focus sites on the Upper Iowa.Among nonpottery traits, stonework includes the following in ad-dition to the aspect traits: round "thumbnail" scrapers, retouchedon all edges ; snubnose or end scrapers which, like the triangular pro-jectile points, tend to be shorter and wider than those from westernOneota sites ; relative abundance of catlinite, notably for disk, equal-arm, and other forms of pipe; small bone "counters," and use ofbirchbark. Inscribed catlinite tablets have been found on UpperIowa sites and at Blood Run, in Lyon County, but in no other OrrFocus components. Objects thought to be of native copper, nowhereplentiful, include rolled tubes (beads?) and small sheet metal orna-ments.White contact materials are associated with Orr Focus sites on theUpper Iowa and in Lyon County, Iowa, as well as in Minnesota. Onthe Upper Iowa, they are noticeably less abundant on upstream sitesthan on those near the mouth. They include only small simple itemssuch as helical coils of brass wire, iron fragments, and glass beads.No axes, hoes, traps, or gim parts have been identified in this tradematerial.The Blue Earth Focus (Wilford, 1955) includes remains in south-em Minnesota and in western and south-central Iowa. It apparentlycenters along Blue Earth River and the adjacent southerly bend ofthe Minnesota, with the Humphrey site in Faribault County, Minn.,as the type location. Other probable occurrences include materialsalong the Little Sioux River in Iowa which Keyes once termed theCorrectionville Focus, and several undescribed sites in Polk and War-ren Counties, central Iowa. There has been very little excavation insites of this complex, and the Iowa sites are known only from surfacecollections.Pottery of the Blue Earth Focus, represented mainly by sherds, istypically Oneota, but differs in certain particulars from that in OitFocus. There is, for example, no evidence of the elliptical jar socharacteristic of the latter. Rims are proportionately shorter ; at the Wbdhl] an introduction TO KANSAS ARCHEOLOGY 605Humphrey site and in Woodbury County, Iowa, punctate, impressed,and occasionally trailed decoration occurs as frequently on the innerrim as on the lip. Vessel ornamentation impresses one as perhapssomewhat more ornate than in Orr Focus, but with less variety ofmotifs ; and it consists largely of trailed lines and punctates. Highlycharacteristic are groups of parallel vertical lines separated by hori-zontal chevrons bordered by single rows of punctates; or the punc-tate-bordered chevrons cap the blocks of vertical lines. Similarmotifs are known from Orr Focus, but are much less common. InBlue Earth, the punctates functioned more often as a fringe to thetrailed line areas whereas in Orr they commonly constitute areafillers. There are a good many vessel handles, usually attached attheir upper end to the lip rather than to the rim exterior ; and theyusually carry more decoration than those in Orr Focus.With respect to nonpottery traits, projectile points and end scraperstend to be narrower in proportion to length than in Orr. Catliniteis scarce. Other items include occasional bone "counters," scapulahoes, antler picks, and numerous bell-shaped cache pits. There areno White contact materials from Blue Earth sites.The Lake Winnebago Focus (McKern, 1945) centers about LakeWinnebago in eastern Wisconsin. It is best known from the Mc-Cauley and Karow sites, on the west shore of the lake in WinnebagoCounty. Both are village sites and represent some intensity andduration of aboriginal occupation, with many refuse-filled pits andconsiderable accumulations of trash. The Karow site includes a burialground. Karow is described as essentially a "pure" Oneota complex,McCauley as predominantly so. Both, it is worth noting, lie wellwithin the territory occupied in early historic times by the villages ofthe Winnebago.Lake Winnebago pottery (Griffin, 1943, p. 296; McKern, 1945) isin many respects the finest of all Oneota wares. The fabric is morecompact, surfaces less pitted and better smoothed, and the potterygenerally more carefully made, than the Orr Focus or Blue Earthmaterials. The predominant and almost exclusive form is again aglobular somewhat squatty jar, which never assumes the ellipticalshape of Orr Focus vessels. Handles are rare; when present, theyare of the flattened strap variety. Rims are straight; the great ma-jority are strongly flared, with their inner surfaces sometimes assum-ing an almost horizontal position. Lip surfaces, usually flattened, aredecorated in about 25 percent of specimens, with closely spaced roundimpressions that result in a distinctly notched effect. Shoulder deco-ration is by trailed lines which are often broad shallow flutes, care-fully executed, with narrow intervening spaces. Very distinctive is arow of punctates encircling the bessel at base of the neck ; below this 606 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174row there are sometimes four trailed lines encircling the pot, withvertical lines below these running to the shoulder. Massed areas ofpunctates occur, as do punctates bordering trailed lines, dashes, andinner rim trailing. There are no chevrons or massed slanting linesin adjoining or opposed blocks.Lake Winnebago nonpottery traits include a rather limited varietyof objects in chipped and ground stone, bone, antler, shell, and nativecopper. There is less chipped stone than in Orr Focus sites. It in-cludes, not unexpectedly, small triangular unnotched projectile pointsand end scrapers, the latter less abundant and rather cruder than inOrr Focus. Fragmentary cutting tools are found, but there is noevidence of the elliptical and diamond-shaped blades. Bipointedstone objects (awls, drills, or fish gorges?) are not reported. Chippeddisks, circular to subcircular in outline, are common at McCauley;but the small circular "thumbnail" scraper type with all edges re-touched is absent. Ground stone includes paired sandstone shaftsmoothers, hammerstones, mortars, mullers, spherical biconicallydrilled pipe bowls, and worked catlinite. The disk pipes, directlyassociated in situ with Orr Focus in western Wisconsin, occur only assurface finds in the Lake Winnebago district ; but their concentrationhere suggests that assignment to the Lake Wiiuiebago Focus complexis highly likely. Bone artifacts include split bone awls of mammaland bird bone, double-pointed cylindrical and also spatulate perfora-tors, mat-sewing needles, oblong counters, and bird bone tubes, somewith incising. There is abundant shellwork, including tanged spoon-like forms, rare fish effigies, notched fragments, disks, and beads.Work in native copper consists chiefly of tubular beads and sheetpendants, to wliich may be added rare finds of conical projectilepoints ( ? ) , perforators, and imidentified pieces.Since the McCauley and Karow sites are both reported to havebeen traditionally habitation sites of the Winnebago, it is of interestto note that a clay trade pipe, glass beads, iron fragments, and gunflints were found at the former. If the association here is not acci-dental, it suggests that a combined ethnohistorical and archeologicalattack on the problem of tribal relationships might be extremelyfruitful ; but I am unable to determine that the leads thus profferedhave been developed in any appreciable measure.The Grand River Focus (McKern, 1945) is represented at presentchiefly by the Walker-Hooper site in Green Lake County, Wise,located some 40 miles southwest of the center of the Lake WinnebagoFocus. This site includes "a highly productive village site and moundgroup," with associated mound and nonmound burials, refuse-filledpits, and a detritus-laden habitational area. A "large majority" ofthe cultural remains here are identified as characteristically UpperMississippi ; but the assemblage is rather markedly deviant from the Wedel] an ENTRODUCTION TO KANSAS ARCHEOLOGY 607three Oneota Foci considered so far. Mound burials, for example, areatypical in Oneota, though said to occur in Minnesota ; and the flexedor semiflexed, and prone extended burials at Walker-Hooper have noknown counterpart elsewhere in the Oneota. Further, as McKern(1945, p. 162) notes, the "Orr and Lake Winnebago potteries aremore sunilar to each other in many respects, including decoration, thaneither is to the pottery of the intermediate Grand River focus,"Grand River pottery (Griffin, 1943, p. 293; McKern, 1945) includesas its characteristic shape a globular jar, which averages appreciablysmaller in size than Orr and Lake Winnebago vessels, and has anorifice larger in proportion to body size. A beakerlike form is alsoreported. Most jar rims are vertical or slightly flaring; about 11percent are recurved, and all tend to be shorter, i. e., lower, than inother Oneota wares. About 60 percent of lip sherds are undecorated ; where decoration occurs, it is usually on the lip or on the outer upperrim, the latter again atypical. The rare handles are of the loop varietyand without ornamentation. Opposed perforations or slots below thelip, presumably for attachment of a thong or bail, are sometimespresent. Body decoration is very rare, being represented on less than5 percent of the body sherds. It consists of trailed lines, usuallynarrow and never flutelike; and there is little use of punctates. Arelatively common motif consists of two or three parallel zigzag linesencircling the pot at the shoulder; parallel vertical or slanting linesmay fill the upperbody between neck and shoulder, or single triangu-late or circular units may repeat themselves around the upperbody.Several vessels are reported to have "a succession of large embossednodes" about the shoulder or upperbody. Presence of a few negativepainted sherds, and of others with paddle impressions reminiscent ofthe simple stamping on historic pottery from the Plains, is note-worthy. Disks cut from potsherds and centrally perforated are alsopresent.Nonpottery traits include the usual mmotched triangular chippedprojectile points; snubnose end scrapers; flake knives, but none ofelliptical or diamond shape; drills with expanded base, few or nostraight forms; mullers; celts; hammerstones ; disk pipes, mainly assurface finds; worked catlinite; sandstone shaft smoothers; plainand incised bone tubes; bone fishhooks; bone beamers; considerablework in shell, including especially tanged spoonlike objects, sometimeswith serrations, and fish effigies; and native copper in the form oftubular beads, sheet pendants, awls, etc. Notable absences includebipointed chipped-stone objects, flakers of bone or antler, and scapulahoes.There is no White contact material in the Walker-Hooper siteinventory, nor is there any indication in the published record thatLake Winnebago, Blue Earth, or Orr Focus sherds were found here.484172?59 40 608 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174This may indicate a somewhat earlier date for Grand River, but moreextended investigation is needed before this point can be cleared up.The feeling persists, too, that Grand Eiver is "off-color" as comparedto the Oneota manifestations described briefly above. The usualjar form here is deviant, as is the low rim tending to a recurving orrolled profile, the general scarcity of body and lip ornamentation, thebeaker form, loop instead of strap handles, knobbed shoulders, rimperforations or slots, and negative painting. Some of these items arecharacteristically Middle Mississippi in flavor, as for example nega-tive painting, beakers, and low rolling rim. Many of them are re-ported from the Spoon River Focus in Illinois. The quartered circle,or cross within a circle, which constitutes one type of body decorationin Grand River, is also Middle Mississippi. In this connection, it isinteresting to note that two copper disks in Middle Mississippi stylehave been reported as surface finds in Green Lake County?one from "a well-known Upper Mississippi site," the other from a Grand Riversite in Kingston (Pasco and Ritzenthaler, 1949). One of these showsthe quartered circle, with small bosses filling each quadrant.In the absence of a comparative trait list giving not alone presenceor absence, but also actual numbers, it is impossible to do more thansuggest that Grand River's possible relationships to Middle Missis-sippi perhaps merit as careful exploration as do its resemblances tothe Upper Mississippi Oneota in which its students have placed it.In northern and western Illinois, Oneota culture or influences havebeen recognized at several sites. With rare exception, these stand inmarked contrast to the comparatively rich and distinctive Orr Focus,Blue Earth, and Lake Winnebago complexes of Iowa, Minnesota, andWisconsin. The Huber site near Blue Island, where village remainsand several extended burials were found, appears to be closest to the "classic" Oneota with respect to pottery, nonpottery, and burial traits(Griffin, 1943, p. 286). Elsewhere, Oneota influence is manifested invarying degrees of admixture with other complexes at the middle levelsof the Fisher site (Griffin, J. W., MS.) and at the Crable site (Smith,1951), as apparently also on Apple River in Jo Daviess County (Ben-nett, 1945) . The general area here involved is one in which extensivetribal realinements took place in the I7th and early 18th centuries,and not improbably for some time previously. It would seem reason-able to expect mixed and/or stratified sites in such a contact area ; andthe locations at which Oneota influences appear may well be attribut-able to contacts between the Chiwere peoples from the north and north-west and Algonkian groups with Middle Mississippi or other culturein the Illinois region. A sustained archeological program oriented Wkdhl] an introduction TO KANSAS ARCHEOLOGY 609in light of ethnohistorical findings would seem particularly promis-ing throughout this region. Alternatively, depending on their finalplacement in time, some of these sites may reflect an early stage inOneota differentiation out of a Middle Mississippi-related ancestor.The exact relationships of the trans-Missouri Oneota sites to theseveral foci briefly characterized in the foregoing pages remain unset-tled. Tliis is, in fact, a complicated problem which awaits moredetailed analytical work than has yet been brought to bear on it. Likethe Iowa and Wisconsin manifestations, however, the western andsouthern sites from which most of our information derives?Leary,Fanning, Asliland, Stanton, Utz, and other sites in Missouri?clearlyreflect a basic subsistence economy divided between maize horticultureand hunting, with some gathering and probably also fishing. Theysuggest communities of varying and sometimes considerable size, withUtz and Leary doubtless inhabited by several hmidred people andcomparing favorably in intensity of residence with the largest ofthe Iowa and Wisconsin sites.As a group, the western sites appear to diverge from the named fociin a number of particulars. There is not much evidence regarding theprevalent house type, but the earth lodge was pretty certainly knownand in use at nearly all western sites. The circular form is indicatedat Fanning and Stanton, the older square form at Leary and probablyat Ashland. Cache pits are everywhere plentiful, with bell-shaped(undercut walls) and cylindrical (vertical walls) profiles prevailingover the bowllike forms of the Iowa-Wisconsin area.In its general characteristics of shell tempering, trailed and punctatedecoration, and preponderance of globular jars with constricted neck,pottery is substantially like the eastern Oneota and can be readilydistinguished from that of other eastern Plains complexes. Occasionalelliptical jars at Leary and Utz are reminiscent of Orr Focus on theUpper Iowa; but their presence elsewhere in the west has not beendemonstrated. Tapered strap handles occur at all western sites, againas on the Upper Iowa. Flattened vessel lips are rare, and sharplyflared rims virtually absent. Except at Ashland and Stanton, themajority of rims at all sites carry decoration, but in varying fashion.Ornamentation of the inner upper rim, for example, predominates inthe small sample from Ashland, is found on about half the Stantonrims, is a minority trait at Utz and Leary, and is all but absent atFanning.Among other features that appear to distinguish the western Oneotasites may be listed the following: high frequency of bison scapulahoes, bone arrowshaft wrenches, and antler projectile points ; a tend-ency toward somewhat greater size in chipped stone projectile points 610 BUREAU OF AMERICAN ETHNOLOGY [Bill. 174and end scrapers, with the latter especially present in large numbers ; more ground stone traits, including grooved mauls, inscribed catlinitetablets (cf. Iowa Orr Focus), greater frequency of catlinite, and per-haps more numerous grinding stones ; much less w^orked shell ; scarcityor absence of bone "counters" and metapodial beaming tools.Some of the items that distinguish western from eastern Oneotasite complexes, such as grooved mauls, arrowshaft wrenches, jug-shaped cache pits, and occasional (?) use of the earth lodge, arecommon in late prehistoric and protohistoric Plains cultures. Theirappearance at tlie western fringe of the Oneota area doubtless re-flects increasingly stronger orientation of Oneota material culturetoward the semisedentary Plains Indian way of life as western con-tacts became closer. A number of widespread Plains traits of thisperiod, on the other hand, do not appear on Oneota sites, or else aresporadically manifested. These include simple stamping on pottery,grit tempering, flat split-rib and "rib-edge" awls, stemmed bone pro-jectile points with square or round cross section, large straight-shaftedflint drills, cancellous bone paint applicators, transversely scoredribs, metapodial fleshers or grainers with serrate blade, and adz-shaped elkhorn scraper handles. The last two items are known fromStanton, scored ribs and paint applicators from Missouri sites.While the general affinities of the western Oneota sites are clear,it would be premature to attempt assignment of any of them to oneof the named foci or to group them in another focus. From site tosite there are differences, as well as similarities, that suggest OrrFocus relationships in some instances, Blue Earth in others, etc.Leai-y and Stanton potterywares may be considered highly typicalOneota; they probably show closer and more consistent similaritiesto the Utz materials than does Fanning pottery ; and Leaiy, Stanton,and Utz are all more suggestive of Blue Earth Focus than of anyother named Oneota subgroup. Utz, on the other hand, where thegreatest amount of digging has been done and researches are con-tinuing, shows much greater variety in all respects than do otherwestern sites, with some indications of chronological differences andof Middle Mississippi influences. An impression persists that Fan-ning is somewhat more deviant than the other western Oneota mani-festations here considered. Its pottery complex shows a high inci-dence of plain undecorated rims and body sherds, a near-absence ofpunctates as fringes on trailed zones or as area fillers, relative abun-dance of a row of small punctates or identations between paralleltrailed lines, numerous tapered strap handles with or without verticalflutings, and abundance of high straight rims. In nonpottei-y traits,our sample is lamentably small; but it suggests inferior stone- Wedbl] an introduction TO KANSAS ARCHEOLOGY 611chipping techniques and fewer stylized knives. There is a feelingthat the people who lived at Fanning, while clinging to an Oneotatradition, somehow did not care much and made little effort to achievethe standards reached by other related groups of the westernperiphery.The Oneota Aspect, as we have noted, includes sites that yieldlimited quantities of White contact material and others that do not.The complex thus bestrides the shadowy border line between theundocumented prehistoric and the post-Wliite contact periods. Thisrelativeh^ late chronological position permits plausible identificationof some of the materials with documented tribal groups. Thus, theOrr Focus sites on the Upper Iowa, where trade goods have beennoted at most sites excavated, are in all likelihood the formerhabitat of the Iowa Indians. In Wisconsin, history and traditionlink some of the Lake Winnebago Focus sites with old Winnebagosettlements; ethnohistoric data and the presence of scanty contactmaterials make it virtually certain that this identification is alsovalid. In Missouri, the highly typical Utz site, along with lesserones nearby, is surely assignable to the Missouri ; and the somev/hatdeviant assemblage in Vernon County, where Oneota influence isstrongly manifested, seems as certainly Osage. Beyond this, theproblem is not so simple. For the Blue Earth Focus, including theCorrectionville materials, Keyes once suggested an Oto authorship;and because of the similarities he noted between Leary and BlueEarth Focus, he further observed that "Leary should align with Oto."More recently, Wilford has suggested that Blue Earth is possiblyan earlier stage in the development of Orr Focus, that is, ancestralIowa. Unfortunately, Leary seems beyond reach of documentaryevidence, and so cannot now be correlated with any recorded move-ments of the Oto, Iowa, or any other tribe. Fanning has beententatively identified with the Kansa ; but the tribal identity of Stantonremains unsettled. Ashland is admirably situated to be the siteof an Oto village of the early 18th century; but absence of Whitecontact material and possible presence of a rectangular housecomplex suggest an earlier time level.While the early historic residence and wanderings of the ChiwereSiouans, and perhaps those of the Dhegiha as well, promise accept-able explanations for substantial portions of the Oneota distributions,this is certainly not the whole story. Relevant ethnohistorical datafirst appear shortly before the middle of the 17th centuiy, but theyare of little real help until the 1650-1700 period; and probably atthis time and a little later, the Orr and Lake Winnebago Foci com-plexes were presumably still functioning as the material culture of 612 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174identified tribes?the Iowa and the Winnebago. What their precisenature may have been a century or two earlier has not been deter-mined, there being no established chronological sequence in the Oneotatradition. Absence of Wliite contact material at Blue Earth sites,if confirmed by more extended investigations, suggests that this focuswas flourishing on an earlier 17th-century time level or before. IfLeary predates A. D. 1600, as suggested in figure 106, this site andUtz may be regarded as evidence that by the 16th century, or before,a well-developed Oneota complex was already established at severalpoints on the west bank of the Missouri. The presence of Oneotainfluence even earlier is suggested in the shell-tempered trailed andpimctated sherd?reminiscent of Blue Earth?figured by Cooper(1936, pi. 13, fig. 3) in an otherwise alien complex from St. Helena,along with disk pipe fragments. Elsewhere in Nebraska, an Oneotastrain has long been recognized in the pre-White contact Lynch com-plex, where it appears to be associated with late Upper Republicanand Nebraska Aspect elements (Freed, MS.) ; and more recently inthe Redbird Focus, a later complex with protohistoric affiliations tonorth and south (Wood, MS.) . In Kansas, aside from the Oneota complex at Fanning, noticemust be taken of materials in other localities which have at one timeor another been termed Oneota. These include materials from theBlue River valley (Gumming, 1958), from Wliite Rock Creek inJewell County, and from Glen Elder in Mitchell County. Among thelimited materials I have seen from these localities, there are smallthin sherds with trailing and fine line incising on shoulder areas,usually in patterns of parallel slanting lines and sometimes in-cluding, as at 14P012 and 14P013 in Tuttle Creek reservoir area, theuse of small punctates, indentations on the lip, and finely crushedshell tempering. At White Rock Creek and Glen Elder (Wedel, 1935,p. 227; Cooper, 1955, p. 13; Rusco, MS.), the specific ceramic resem-blances to Oneota are much less close and there are notable differ-ences : the almost exclusive use of grit instead of shell for tempering,and the moderate abundance of simple stamping. Locally, the wareis readily distinguishable from that on neighboring Upper Repub-lican and other sites, earlier and later; but it is a far cry from thepottery at Leary and Fanning. Nonpotteiy remains are inadequatelyrepresented ; but such characteristic Iowa Oneota items as stone diskpipes have been reported from Wliite Rock Creek (pi. 84, d).Grooved mauls, four-edged beveled knives, triangular unnotchedprojectile points, catlinite, "rib-edge" awls, and other items indicatea fairly late prehistoric or protohistoric time level; but no tradematerials have been reported from either site. An Upper Mississippi Wedbl] AN INTRODUCTION TO KANSAS ARCHEOLOGY 613 1800 1600 1400 1200 1000 800 600 400 200 CD WesternHIGH PLAINS CHEYENNE, ARAPAHOCOMANCHE, KIOWADISMAL RIVERI4SCI I4LAI(AJ Central KansasSouth North Flint Hills- MissouriOsage Plains Bluffs PAWNEEI4RPI uj^ little river* >3 Focus^ ^ LOWER WALNUT "15 Focus^^ GLEN ELDERS Focus htSAII40T5 I4PH45:K I4LAI(B) H Eastern I4P024^KANSA -7I4DP2I4WNI-2 I4DPZ \ RENNER,I4WYI,etc. P A E O ) N D ) AN Nebraska HISTORIC PAWNEE,etc. RedbirdFocusLear/ r^ BluestoneLynch 1400 1200 600 200 A.D.Bc' -5000 Figure 106.?Suggested chronology of certain archeological sites and complexes in Kansas,including provisional correlations with archeological sequences in the Rio Grande puebloarea and in Nebraska. (See pp. 615-620.) (314 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174 connection is certainly strongly indicated, and there may be somerelationship with Oneota; but the relationships to the named foci inIowa, Minnesota, and Wisconsin are clearly of a very different sortthan is implied in the Fanning, Leary, or Stanton materials. Onpresent evidence, inclusion of White Rock Creek and Glen Elder inthe Oneota Aspect is not warranted.In summary, we still know too little about the time of arrival ofthe earliest Oneota in the trans-Missouri region and the directionfrom which it came- The traces of Oneota at Lynch and St. Helena,however, seem like very good evidence that its influences were operat-ing west of the Missouri during the final stages of Upper Republicanoccupation, possibly before A. D. 1500; and Leary, as elsewheresuggested, probably indicates the presence in essentially classic pro-portions of an Oneota community in the region before A. D. 1600.If these dates are substantially correct, it seems remarkable that thereis so little evidence of Oneota in the Lower Loup pottery complex. Iknow of no unmistakably Oneota sherds from any Lower Loup sites ; and there are no Lower Loup sherds from Leary, though several werenoted at Fanning. It would appear that Oneota influence on easternPlains pottery was less pronounced than many of us have supposed.Alternatively, it may be that Leary antedates the Lower Loup sitesfrom which extant collections have been gathered, which propositionmight help explain the presence here of a square Upper Republicantype of earth lodge instead of a circular structure of Lower Loupform.Potterywise, probably the chief traits that have led to classificationof materials as Oneota are one or more of the following: shell tem-pering, the combination of straight rims with globular vessel bodies,decoration on the inner rim or lip surface, trailed or incised decora-tion and punctates on shoulder areas irrespective of design patterns,and presence of strap handles. Some of these features, it is true,make their first important appearance or become markedly moreabundant in late prehistoric and early protohistoric times?at roughlythe time Oneota appears in the area. Straj) handles, however, arecommon on Nebraska Aspect pottery, probably pi-edating developedOneota in the eastern Plains; shell-tempering characterizes MiddleMississippi pottery, which is known to have reached the southeasternNebraska and northeastern Kansas region; and the free use of rec-tilinear shoulder decoration is not limited in its occurrence to Oneotapottery. Distinctly Oneota design patterns are rare or absent onLower Loup pottery, and apparently also on such late prehistoric andprotohistoric manifestations as Scalp Creek, Arzberger, Lynch, Red-bird, and White Rock. More plentiful is a basically triangular ar-rangement of contiguous blocks of oppositely slanted parallel lines re- Wbdel] an introduction TO KANSAS ARCHEOLOGY 615peated about the vessel shoulder. This has a wide distribution onpottery of the Eastern United States, apparently in both Middleand Upper Mississippi contexts. Whatever its derivation, this is notOneota; but its wide spread along the eastern Plains and in theMissouri valley offers an interesting problem whose closer examina-tion might well contribute to solution of other problems on this generaltime level in the region. The same motif occurs, curiously enough, onPotsuwii Incised pottery from the Chama Valley in north-central NewMexico at approximately the 1400 A. D. time level (Jeancon, 1923,pp. 54-57 ; Wendorf, 1953, pp. 55, 98) . That the relatively few Oneota sites now known west of the Missourirepresent a continuous residence in the region by Oneota peoplesseems improbable. They impress me rather as repeated incursionsfrom the east, probably varying in strength and duration, beginningperhaps before A. D. 1500 and continuing for an undetermined periodthereafter. The final stages of Oneota in the trans-Missouri regionmust await further investigation of late iTth and 18th century sitesof the Chiwere and Dhegiha groups in the area.TIME PERSPECTIVEChronological arrangement of the various Kansas sites and culturecomplexes discussed in the preceding pages can be done by severallines of evidence and in varying degrees of preciseness. This prob-lem has two facets: (1) relative dating, or the determination of thetime of a site or culture with reference to earlier and later sites andcultures; and (2) absolute dating in terms of our own calendar. Inboth respects, the present task has been materially lightened by themarked advances made in surrounding areas and also in datingmethods since completion of the 1937-40 fieldwork of the U. S. Na-tional Museum.For some of the late sites that have come under investigation, thereis historical documentation by White explorers and other visitors thatmakes possible determination not only of the time of occupancy butalso the tribal identity of the native inhabitants. Other sites showvarying degrees of Euro-American contact in the form of metal,glass, and other materials of nonaboriginal origin; and, since suchalien materials usually became increasingly important and plentifulamong the Indians after their introduction about the 16th century,their relative amount in a given site inventory permits some infer-ences as to age. This method, of course, gives relative rather than ab-solute time determinations. For pre-White contact sites, stratigraphyor the superimposition of culture layers is the most reliable indicatorof the succession of cultures, but again does not tell precisely where aspecific site or layer belongs in time. 616 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174An approximation of absolute dating has been possible in a fewinstances. The finding in Kansas sites of Southwestern pottery frag-ments representing types whose span of existence in New Mexico hasbeen accurately dated through tree rings is an important clue tochronological equivalence of certain Central Plains and Southwesterncultures. The application of dendrochronological methods to Plainssites has not been carried vei-y far, but a few 18th century sites of theDismal River Aspect have been plausibly so dated in Nebraska ; andat Ash Hollow Cave an early 14th-century date for an Upper Repub-lican layer rests on the same sort of evidence. Finally, there areradiocarbon dates for several prehistoric sites that offer helpful cluesto the probable position of some of the earlier cultures now recognized.The time relationships of the Kansas materials, as I see them atthe moment, have been charted in figure 106, to which the present dis-cussion is supplementary and explanatory. Further details have beenbrought together in the descriptions of individual sites and in thereviews of the various complexes represented. The sections of thechart dealing with Rio Grande sequences have been prepared withthe generous assistance of Mr. Stanley Stubbs, Museum of New Mex-ico (letter of March 25, 1957) . As is to be expected, the most recent sites are also the most firmlydated. The Manhattan or Blue River site (14P024) is the locationof the main Kansa village of ca. 1800-30, where the tribe was visitedin 1819 by Professor Say of the Long exploring expedition. TheKansas Monument site (14RP1) can be safely identified with aPawnee village dating from approximately the last quarter of the18th century, and in all likelihood abandoned before 1800. The con-curi'ent occupation in western Kansas was entirely by equestrian bisonhunters?the Comanche, Kiowa, Cheyenne, Arapaho, and other tribes,whose legacy to archeology was of the most meager sort.Three other sites are assigned to an 18th-century dating, ante 1750.In northeastern Kansas, the late component burials and caches atDoniphan (14DP2) are accepted as afiiliated with the site of theGrand village of the Kansa visited by Bourgmond in 1724. The "El Cuartelejo" site in Scott County, 14SC1, is assigned to a slightlyearlier date?partly because of Tewa Polychrome sherds and incisedPueblo pottery pipe fragments identified by Mera as belonging to the "last two decades of the 17th up to the early years of the 18th cen-turies," and partly because of the essential similarity of the site in-ventory to other Dismal River sites in Nebraska dated to the early18th century by dendrochronological and other methods. Moreover,as I have indicated in the discussion of this site, there seems to be areasonable likelihood that it was one of the Cuartelejo Apache ran-cherias from which Ulibarri in 1706 rescued a number of Picuris cap- m\ Wedbl] an introduction to KANSAS ARCHEOLOGY 617tives or fugitives. In soutlieastem Kansas, the Neodesha "fort"(14WN1) and nearby village site (14WN2) are tentatively allocatedto the first quarter of the 18th century on what I cheerfully concedeis inadequate evidence. I have elsewhere indicated my views regard-ing the possible identification of these remains as Wichita of perhapsthe time of Du Tisne.The Fanning site, 14DP1, is clearly a relatively late Oneota mani-festation?certainly later than the nearby Leary site. It probablydates fi-om the latter half of the 17th century, perhaps even from thefinal quarter. The probability seems good that it is identifiable witha Kansa occupancy.The Great Bend Aspect is appreciably earlier than any of the above,though in part certainly post-White contact. The partly inclusivepresence at several sites in Rice and Cowley Counties of Rio Grandeglaze paint sherds dating from circa 1525-1650, and surface findsin the same localities of other Pueblo pottery types running from thelate 15th to the early 18th centuries, probably bracket the occupationfairly closely. The historical and archeological evidence linking thesesites with the Quivira villages visited by Coronado in 1541 and Oiiatein 1601 has been summarized previously in this paper and elsewhere.The Great Bend aspect is probably approximately contemporaneouswith the Lower Loup Focus in Nebraska, as indicated by broad andconsistent similarities in the cultural inventories of the two complexes,and the evidence of direct contact shown by cross finds of potsherds.These include the discovery by Udden (1900, fig. 10) in one of thePaint Creek middens of a large vessel fragment done in unmistakableLower Loup style, and the finding of shell-tempered, flat-based vesselfragments of Lower Walnut Focus type in a house ruin at the Graysite in Colfax County, Nebraska.The Pratt complex, though very inadequately known as yet, isdated earlier than Great Bend because of (1) the presence of RioGrande glaze paint and biscuit ware sherds attributed to the 1425-1550 period, thus preceding slightly the dating indicated for GreatBend; and (2) the general nature of the artifact inventory, whichincludes a significantly higher proportion of cord-roughened grit-tempered pottery than Great Bend and also shows other similaritiesto late prehistoric cultures of Oklahoma. The finding of a glass beadis not wholly consistent with such a dating, nor, perhaps is the pres-ence of what seem to be Dismal River sherds. Since the availablematerials are all surface finds, it is quite likely that there has beensome mechanical mixing of specimens from two or more separateoccupations. I recognize the urgent need for systematic excavationat this site, and in any related ones that may be found in the southern 018 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174Kansas area ; and meanwhile record my feeling that it precedes typo-logically, and may be in part ancestral to, the Great Bend Aspect.The Glen Elder Focus may belong to the same general time levelas Pratt, but appears to me to find more congenial company on theGreat Bend-Oneota level. The general absence of White contactmaterial suggests a dating not later than the early 16th century; ifthe complex had been significantly later in time, I would expect inits trait inventory (Rusco, MS.) some recognizable ceramic or otherevidence of direct contacts witli the Lower Loup Focus and the GreatBend Aspect, these cultures having flourished in some strength notfar to the north and south, respectively, of the Glen Elder locality.For the Upper Republican and Nebraska Aspects I have suggestedan earlier dating than those usually given heretofore (Wedel, 1947 a,fig. 51; Champe, 1946, fig. 17; Lelimer, 1954 b, fig. 1). The reasonsfor this, and for proposing a new Smoky Hill Aspect on the same gen-eral time level, have been discussed elsewhere herein. There are,unfortunately, no records of datable Southwestern sherds found in-clusively in any sites of this general culture stage in Kansas andNebraska. This contrasts with the situation in the related AntelopeCreek Focus of the Texas Panhandle, where "potsherds centeringabout the Rio Grande glazes of periods I and II, but also including St.Johns polychrome" have impelled Krieger (1947, p. 143) to set a dateof circa A. D. 1300-1450. The absence of Southwestern pottery in Up-per Republican sites in western Kansas may well reflect the extremelylimited excavation done so far there; but the same explanation can-not be invoked for southwestern Nebraska, where the recent large-scale operations and much previous digging on a smaller scale atMedicine Creek have failed to turn up a single identified sherd ofPueblo origin. Geographically, the Medicine Creek locality is nomore remote from the upper Rio Grande than is the Great Bend re-gion of central Kansas, to which puebloan influences penetrated re-peatedly in the 15th century and later. This may mean that theUpper Republican hold on the High Plains between the Arkansasand Platte Rivers was already slipping by the time glaze-paint deco-rated pottery was establishing itself on the Rio Grande, and that theearth-lodge village-dwellers were moving northeastward toward theMissouri River.Note has been taken elsewhere of a radiocarbon date of A. D. 1176?150 years for "an Upper Republican site in the middle period . . .of this culture" (Crane, 1956, Sample M-113) in Clay County,Kans. The associated cultural materials here have not been de-scribed in print, so far as I am aware ; nor do I know just whatfeatures differentiate a "middle period" Upper Republican site fromany other assignable to this aspect. The location of the site on theI Wedel] an introduction TO KANSAS ARCHEOLOGY 619Lower Republican suggests that it may fall into what I propose tocall the Smoky Hill Aspect rather than into Upper Republican asknown from farther west and northwest. In any case, the date, ifvalid, points to an earlier time level than has been indicated by mostprevious estimates.The presence of Crockett Curvilinear Incised pottery at anotherSmoky Hill site (14SA1) suggests contemporaneity with some GibsonAspect sites of the lower Arkansas-Red River area. This ware occursin the Spiro Focus where, as we have noted, radiocarbon dates fallinto two widely disparate groups?one at circa 330 B. C.? 200 years,the other at A. D. 1316?250 years. Krieger (Suhm, Krieger, andJelks, 1954, p. 262) assigns the type to "more or less entire time spanof Gibson Aspect, some part of 500-1000 A. D." ; but in his charac-terization of the Haley Focus, where Crockett Curvilinear Incised islisted, he gives the chronological position as "some part of span 800-1200 A. D." (ibid., p. 182) . In this light, the placement of site 14SA1in figure 106 may be conservatively late.I know of no radiocarbon or other "precise" dating for the NebraskaAspect. Its partial equivalence with Upper Republican on one hand,and with some variant of Middle Mississippi on the other, suggestthe placement shown on the chart.Tlie entries for Plains Woodland and Hopewellian also representa considerable downward revision from previous suggested chronol-ogies. The revision is based, in part, on recent radiocarbon deter-minations. Not enough of these are at present available for theCentral Plains area to permit satisfactory evaluation, but those nowat hand suggest that some of tlie Plains Woodland complexes mayhave to be allowed a greater time span, an earlier appearance in thearea, or both, than has been indicated in figure 106. For Kansas CityHopewellian, for example, there are four radiocarbon determinationsfrom the Renner site: A. D. 8 ?250 years (M-572) ; A. D. 108 ?200years (M-571) ; A. D. 438 ?200 years (M-573) ; A. D. 687 ?250 years(M-454) . It is in the highest degree improbable that the Renner sitewas occupied for six centuries, or that Hopewellian peoples reoccupiedit at different times during a period approaching that length. I haveassumed that a date in the second or third centuries of the ChristianEra probably approximates the time of occupation. Other relatedsites in the district maj^ be earlier or later than Reimer.As to other Plains Woodland complexes, the Keith Focus and sitesthat appear to be fairly closely related to it are represented by thefollowing C14 determinations: Woodruff ossuary (14PH4), A. D.611 ?240 (C-928); Massacre Canyon, Nebr. (25HK13), B. C.122 ?250 (M-181, Crane and Griffin, 1958, p. 1121) ; site 25FT18,Frontier County, Nebr., A. D. 828 ?200 (M-841). These figures 620 BUREAU OF AMERICAN ETHNOLOGY IBull, 174have a much wider spread (950 years), and suggest a much earlieroccurrence, than I have shown in figure 106. The oldest date, M-181,is based on fresh-water shell, the others on charcoal.There is but one date for the Valley Focus, based on fresh-watershell from the type site, 25VY1. Reported as "regarded by thecollector as much too old," this date is B. C. 1872?300 (M-182, Craneand Griffin, 1958, p. 1121). It is entirely possible that Valley Focusshould be allowed a much longer time span than I have indicatedin figure 106; but I, too, would prefer further confirmation beforeaccepting the one date now at hand.^?Since almost nothing is known of the preceramic occupations ofKansas, I have not attempted their placement on the chronologicalchart. If, as present radiocarbon datings suggest, the Woodland andHopewellian manifestations were present in the area in the earlycenturies of our era, any Archaic materials whose existence and exactnature here may be established in future will probably be acceptableas belonging to a pre-Christian time horizon. That man was presenthere several millennia ago is certainly suggested by the recurrentsurface finds of paleo-Indian materials, including Plainview and Fol-som types of projectile points, elsewhere dated by radiocarbon ateight to ten thousand years before present. The association at TwelveMile Creek of a projectile point with skeletons of Bison occidentalisestablishes man's presence here at a time when faunal forms nowextinct roamed the area; and similar associations in adjacent Statesstrengthen the likelihood of further finds of this sort in Kansas.Development of this problem, however, is for the future, since, towardthese early stages of man's existence in Kansas, our own work con-tributed little. I can add only my conviction that the prospects forfurther and more significant finds along this line are excellent.INTERPRETATIONSThe data on Kansas archeology gathered by the United States Na-tional Museum in 1937-40 are now before us. So, also, are most ofthe essential available archeological observations on the State thathave been published, summarily or otherwise, in various journals andother papers. So far as seems warranted, the sites have been classi-fied and assigned to the larger regional groups to which they seemrelated (table 18) ; and a provisional scheme of chronological relation-ships has been offered (fig. 106). There remains now the matter ofinterpreting and integrating the materials ; and this can perhaps bestbe done by a provisional historical reconstruction. "For permission to Include here Plains Woodland and Hopewellian dates not yet pub-lished when the proof sheets left my hands, I am obligated to James B. GrlflBn. James H.Howard, Carl H. Chapman, and Robert !>. Stephenson. These determinations weremade by Crane of the University of Michigan. Wkdkl] an introduction TO KANSAS ARCHEOLOGY 621The tentative nature of any such attempted reconstruction at thistime must be emphasized at the outset. Geographically, the materialat hand affords only a very spotty and incomplete coverage of theKansas area. Large sections of the State remain unsurveyed byarcheologists?or, if surveyed, the results have not yet been madegenerally available. Again, the operations of the U. S. NationalMuseum were concerned almost wholly with Indian remains of thepotterymaking cultures of the past 1,500-2,000 years?very largely,indeed, with those of the last half of that period. Even here, thesampling has been necessarily limited and often inconclusive. Stillmisampled and untested are several thousand years of time precedingour era, and many hundreds of miles of river and creek valleys in allparts of the State.Partially offsetting these deficiencies is the obvious fact that the ab-original cultures of Kansas did not originate or develop in completeisolation. Throughout, they participated in various ways in develop-ments that also involved peoples and cultures in adjacent areas. It ispossible, therefore, to evaluate the available Kansas survey data inlight of the much better-known materials in Nebraska and otherneighboring States, where the range and sequence of aboriginal cul-tural development have been emerging with increasing clarity formore than two decades.It seems entirely reasonable to postulate the presence of man inthe plains and stream valleys of Kansas at least as early as eight toten thousand years ago. By this time, the Kansas landscape hadprobably acquired substantially its present form with respect totopography, drainage pattern, and native flora and fauna. The moistcool climate attending the last Wisconsin (Mankato) glaciation ofthe Midwest was giving way to a drier climate, apparently to a climaticregime characterized, like that of the present, by "extremes of aridityand high temperature in summer, followed by cold dry winters . . ."(Frye and Leonard, 1952, p. 180). Climatic fluctuations from timeto time can be inferred, and some of these were probably of consider-able magnitude and duration ; but their details remain to be workedout. There is no sound reason to suppose that the vegetative coverdiffered materially from that of the present, with tree-borderedstreams and broad grassy mterfluvial uplands. The environmentalsetting, in short, was essentially that which the first white explorersencountered some four centuries ago.Evidence for man's presence here at this early date is of the mostmeager sort. It includes no dated campsites, game kills, or humanskeletal remains. As we have noted elsewhere, however, scatteredfinds of well-fashioned Folsom points have been made in the westernand northeastern portions of Kansas ; and in the northeast, also, there 622 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174have been finds of points resembling Plainview types. Such surfacefinds, without definable geologic or faunal associations, are admittedlynot vei-y strong evidence. They assiune added significance, however,in light of dated finds of similar artifact types in nearby Nebraskaand other Plains States. The discovery on Twelve Mile Creek inwestern Kansas of a projectile point in direct association with severalskeletons of Bison occidentcdk is also relevant, for it establishes con-tact between early Kansans and a fauna now extinct. Insofar asthere is anything diagnostic in the point types thought to representthis early time level, the relationships appear to have been with west-era Plains manifestations rather than with early eastern complexes.As to the probable nature of this postulated early occupancy, nodetails can be offered on the basis of the Kansas data alone. Byanalogy with other finds in adjacent areas, however, it is possible tosuggest what may be expected when Early Man or Paleo-Indian siteseventually come to light in Kansas. Throughout the western Plains,the remains of these early big-game huntei-s are associated with bonebeds, evidently the results of game kills or butchering activities, andcampsites. From such locations comes little or no information re-garding the size, composition, and organization of the original com-munities. There are usually no structural evidences, posthole con-figurations, or food-storage devices, such as pits. Hearths may occui',but they are usually scarce and rarely give evidence of long-continueduse. The artifact inventory is almost wholly limited to items of stone,but sometimes includes a few bone implements. The implements areprunarily those designed for the chase and for processing the kill ? projectile points, scrapers, knives, and choppers. There are few orno grinding tools or other items suited to the gathering and process-ing of vegetable foods, nor is there evidence of the dog.Inferentially, the population was a sparse and scattered one, essen-tially nomadic, and possessed a subsistence economy closely linked tothe large grass-eating herd animals. The local groups were prob-ably of very limited size, perhaps consisting for most of the time ofsmall family imits. The habitations were veiy likely light, imper-manent dwellings of poles covered with skins or grass thatch, andfrequent shifts of residence doubtless took place in accord with avail-ability of game and other foods, the needs of the seasons, and othercircmnstances. Although most associations of early artifact typesare with skeletal remains of large mammals, it is veiy probable thatthe yearly round of economic activities included also the harvestingof roots, seeds, and berries in season. Winter residence in shelteredlocalities where game, water, and fuel were obtainable can beinferred. Wedel] an introduction TO KANSAS ARCHEOLOGY 623The subsistence economy outlined above must be characterized asone of food coUectmg, not food producing. A more or less similarbasic economy midoubtedly was followed also by the peoples who,inferentially, followed the western Paleo-Indians in the Kansas re-gion before beginning of the Cliristian Era. In the Eastern UnitedStates, these later peoples, nonhorticultural and lacking also the pot-ter's art and momid burial, have been assigned to the Archaic stage.Subsisting primarily by hmiting, but doubtless with some gatheringand perhaps fishmg, these people made a variety of side-, base-, andcorner-notched projectile points, lanceolate points, blades, haftedscrapers, and choppers. From sites of later phases have come suchground stone implements as celts, grooved axes, boatstones, and gor-gets. Bonework includes awls, needles, tubes, shaft wrenches, andlong slender hairpins with expanded butts. Antler and shell arti-facts are also known. Cave deposits in central Missouri have yieldedburned clay with impressions of twined textiles and coiled basketry.Deep accmnulations of refuse suggest that these communities some-times persisted for long periods of time, presimiably by relying onhunting in forested areas rich with game that moved about less thandid the bison herds on which the Paleo-Indians of the West depended.For the Archaic level, there is only meager and uncertain evidenceat present in Kansas. I have suggested elsewhere the possibility thatthe coarsely chipped points, choppers, scrapers, and other tools of thenorthern Flint Hills region?perhaps including the "paleolitlis" de-scribed by Winchell?may represent an Archaic level. Long bonehairpins with expanded butts, and other elements found at the Younkinsite near Junction City, have also been interpreted as a survival ofArchaic traits. The demonstrated existence in Missouri of a longand varied Archaic occupancy and the occurrence of other Archaicmaterials in northeastern Oklahoma makes highly probable the ulti-mate recognition of related manifestations in the once heavily timberedriver valleys of eastern Kansas. The Nebo Hill complex of the Kan-sas City locality may well have extended across the Missouri Valleytrench into northeastern Kansas, though I am not aware that the char-acteristic lanceolate blades of this manifestation have yet been foundor recognized in the State. Possibly "Lansing Man" should be allo-cated to an Archaic level, but this must be regarded as an inconclusivepoint. Farther west, it seems very probable that preceramic camp-sites and artifacts will ultimately be recognized in alluvial streamdeposits and that some of these materials will be assignable to anArchaic hunting-gathering level.If we can accept the accuracy of radiocarbon datings, present evi-dence suggests that pottery making and maize horticulture reachedthe eastern Kansas area by or during the early centuries of the Chris-484172?59 41 624 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174tian Era. The oldest dated remains with which corn has been foundassociated are the Hopewellian communities of the Kansas Citylocality, inhabited probably before A. D. 400.As represented at the Renner and Trowbridge sites, these Hopewel-lian settlements appear to have been fairly stable communities. Nodefinable structural remains have been found but there are numerousstorage pits of small to medium size. The sites are generously mantledwith village refuse, including potsherds, worked stone and bone, andconsiderable quantities of refuse animal bone consisting chiefly ofdeer and other woodland forms. There is no information as to thetype of dwelling or the arrangement of dwellings within the occupiedarea. Burial mounds of earth, containing square or rectangular drymasonry chambers with entranceways, occur on the bluffs near thesevillages and are believed to be associated with them.Except in their lack of discoverable house remains, the Hopewellianpeoples seem to have possessed a type of life much like that of latersemihorticultural Village Indians of the eastern Plains. The artifactinventory is a comparatively rich and varied one, at least by compari-son with preceding and contemporary residents of the region. Itincludes heavy stemmed and corner-notched projectile points, endand side scrapers, flake and other knives, and other chipped tools;grooved axes, celts, mammiform objects, and other items of groundstone ; metapodial and deer ilium beamers, turkey metatarsal and mam-mal bone awls, large and small needles, longitudinally pierced deerphalanges, imitation bear canines, and various unidentified objects ofbone; dressed subcylindrical and curved sections, socketed projectilepoints, and other artifacts of deer antler; and some native copper.Pottery occurs in quantity ; all of it is grit tempered. At least twowares are represented : a cord-roughened minority ware with punchedexterior bosses below the rims of tall conical-bottomed jars; and amore carefully made smoothed ware, including jars and bowls, com-monly bearing rocker-roughening, dentate-stamping, or rouletteimpressions which are sometimes in zoned arrangement.The closest relationships of these people appear at the moment tohave been with the Illinois Hopewellians. The westward thrust rep-resented by the sites at Kansas City spread only a little way fartherup the Missouri, apparently stopping short of the Nebraska line ; butthere was also a movement up the Kansas River to Junction City orbeyond, as well as a southward spread from the Kansas River valleyinto northeastern Oklahoma. The details of these movements,whether cultural, ethnic, or both, remain to be worked out. Essen-tially, it seems that the Hopewellian occupation of the eastern Plainswas restricted to the eastern third of Kansas, and that village sites Wbdbl] an introduction TO KANSAS ARCHEOLOGY 625 of any size or permanence are not likely to be found west of the areaassigned by physiographers to the Central Lowland province.Apparently contemporaneous in part with the Hopewellian com-munities of the eastern Plains were other peoples with a much ruderculture who inhabited the Plains to the west and northwest. Thesewere the bearers of what has long been recognized as Plains Woodlandculture. There are several recognized variants, all with pottery butmost without clear evidence of horticulture. Plere the sites areusually small and inconspicuous, often deeply buried in stream ter-races, and with a wide distribution along the smaller rivers and creeksof Kansas and Nebraska. In Nebraska, comprehensive excavationshave disclosed shallow basins from 15 to 18 feet in diameter, some-times with hearths, and suggesting habitation units. Comparablefeatures have not been reported from Kansas Woodland sites. Thehabitations are thought to have been of light construction, perhapsof poles covered with grass thatch, bark, or skins, and seldom num-bering more than half a dozen single-family units per site. Burialsinclude both single primary and massed secondary interment, thelatter in what appear to have been communal cemeteries resultingfrom prolonged, but not necessarily uninterrupted, localized residencenearby.The artifact inventory of these western Woodland peoples is muchsimpler and less varied than that found at Hopewellian sites. Itincludes stemmed and corner-notched projectile points, end scrapers,unbeveled knives, sandstone sharpening blocks, split metapodial andsplinter awls, bone tubes, hollowed bison phalanges, and sometimesbone "knives" fashioned from scapula fragments. Pottery is char-acteristically grit temT)ered and cord roughened, with large jarstypical, but apparently much less plentiful than in Hopewellian andlater sites and complexes of the ceramic period. Missing are suchcommon lat?r Plains artifact types as bison scapula hoes, diamond-shaped beveled chipped knives, and longitudinally grooved sandstoneshaft smoothers; and these items are also absent from the westernHopewellian assemblages.The chronological and other relationships between the variousWoodland complexes now recognized in the Central Plains and themore highly developed Hopewellian culture in their eastern portionare not entirely clear, A minority ware at the Hopewellian sites, aswe have noted, is a heavy, gravel-tempered, cord-roughened potteryreminiscent of that which characterizes many of the westerly Wood-land sites. Partial contemporaneity or overlap is suggested. Theseminority sherds at Hopewellian sites resemble those attributed in Ne-braska to the Valley Focus type, and it thus seems likely that thismanifestation or variants of it were present in the area at least as 626 BUREAU OF AIVIERICAN ETHNOLOGY [Bull. 174 early as Hopewellian. Similar sherds have also been found in asso-ciation with the later Loseke Creek Woodland variant, suggesting fur-ther that the Valley type may represent a basic and long-lived Wood-land pottery that arrived in the Central Plains as early as Hopewel-lian, if not indeed before, and survived after the latter culture haddisappeared. Until radiocarbon dating has been applied to ValleyFocus materials or direct stratigraphic evidence is found, the questionof priority in time as between Valley Focus and Kansas City Hope-wellian must remain conjectural.Other than the Hopewellian, most Woodland sites in the Kansas-Nebraska region suggest a simple creek-valley hunting and gatheringeconomy, with relatively small population aggregates. Squash andgourds in the Stems Creek variant suggest horticulture, as doescharred corn in Loseke Creek, neither of which manifestations has yetbeen reported from Kansas. The bones of deer, elk, and smallerwoodland margin mammals and birds often predominate over thoseof bison in the refuse deposits; but it is not clear whether this reflectsfood preferences or, alternatively, hunting and butchering methods.This point I have discussed at greater length elsewhere in the presentpaper.The time span of the Middle Woodland-Hopewellian occupation ofthe Central Plains is still undetermined. The few scattered radio-carbon dates now available suggest that both Keith Focus and ValleyFocus may predate the beginning of the Christian Era and also theappearance of Hopewellian peoples on the eastern margin of theCentral Plains. Two of the three Keith Focus dates, on the otherhand, are later than three of the four Renner dates; and the third,like the only extant Valley Focus date, was derived from fresh-watershell rather than charcoal. Loseke Creek has not been dated, but itwould appear to be certainly post-Hopewellian and probably morerecent than either Keith or Valley Focus. It seems probable that theHopewellian sites in the Kansas City-Kansas River valley arearepresent a relatively short-lived thrust, in the early centuries of theChristian Era, westward from the Mississippi Valley into an areadominated before and probably afterward by a varied assortment ofsimpler hunting-gathering and, later incipient horticultural, peoplesgenerally classed as having Middle Woodland affiliations. I see noreal evidence of continuity in the Central Plains between the relativelystable Hopewellian communities and the almost certainly much latersemisedentary village dwellers to whom we come next.Some time after circa A. D. 1000, but still several centuries beforearrival of the first Wliite men in the Central Plains region, the MiddleWoodland-Hopewellian complexes were succeeded by a group of semi-horticultural pottery-making cultures implying a much more stable Wkdkl] .\N introduction TO KANSAS ARCHEOLOGY 627pattern of settlement. Here there is direct and convincing evidenceof cultivation, with subsistence based in part on a small scale hoe-linked maize-bean-squash-sunflower horticulture and in part on hunt-ing, fisliing, and gathering. Included in these semisedentarycomplexes were several prehistoric manifestations featuring small,loosely arranged, unfortified settlements?the Nebraska Aspect ofnortheastern Kansas and eastern Nebraska, the Upper Republicanaspect of western Kansas and Nebraska, and the related Smoky HillAspect sites in central Kansas.Highly characteristic of these cultures?collectively subsumed underthe designation of Central Plains phase or tradition?are remains ofsubstantial earth-covered lodges. Square to rectangular in floor plan,with rounded corners and extended entranceways, these were generallymuch larger than the dwellings inferred for the earlier Woodlandpeoples, and were probably deigned for larger extended familygroups. In the Nebraska Aspect, largely restricted in Kansas to theimmediate bluffs of the Missouri, the settlements consisted of scatteredhouses usually irregularly disposed along the tops of narrow ridgesand bluffs ; terrace locations seem less characteristic, though they maywell exist along some of the short creek valleys draining into the Mis-souri. Deep pits marking these sites indicate that the lodges wereoften semisubterranean, with the floors from 2 to 4 feet or more be-neath the gromid surface. Farther west. Upper Republican villagesand apparently those of the Smoky Hill aspect in the lower KansasRiver drainage, stood on terraces or on bluffs iromediately overlookingthe lesser streams and creeks. For the most part, these houses stoodon or but a few inches beneath the ground surface and are not prop-erly called pithouses. Single lodges were scattered randomly at inter-vals of a few yards to several hundred feet; in other instances,clusters of two to four lodges were similarly separated from othersmall clusters or single units. One or more small subfloor cache pits,cylindrical or bell shaped, were to be found in each of these habita-tions; and small refuse dumps often occur nearby. The higlily dif-fuse pattern of settlement found at village sites of the Central Plainsphase was ill adapted to fortification; nowhere has any evidence ofdefensive works been noted ; and relative freedom from enemy raidsseems to be generally indicated.For these small scattered rural communities, probably seldom ex-ceeding 100 souls and often much smaller, the nearby creek bottomsdoubtless provided ample garden lands, and game could be obtainedby foot hunters in the brushy valleys and along the upland margins.Inferentially, though the scale of horticulture here was more limitedthan in protohistoric times, the basis of life would seem to have beena more secure one than in the preceding Woodland stage; and this is 628 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174 reflected in the greater amounts of refuse and more abundant arti-facts of pottery, stone, bone, and other materials. Thus, for example,the floors of single Upper Kepublican house sites have been knownto yield fragments of more pottery vessels than have come from thethree or four probable habitation units comprising an entire Wood-land site. The nature of the pottery and other artifacts from thevarious complexes included in the Central Plains Phase has been re-viewed in a preceding section.The westward distribution of the small earth-lodge communities ofthis period in Kansas remains to be worked out in detail. It hasalready been shown in Nebraska that such settlements were wide-spread along the smaller streams tributary to the Elkhom, Loup,Platte, and Eepublican Rivers, and that their distribution extended toalmost all localities where climate and soils permitted creek-bottomgardening, dependable water for domestic use could be found, andtimber was available for building purposes. There seems to havebeen a preference for residence on small streams, perhaps becausetimber was more readily available here than on the larger and moreopen river valleys. The upper reaches of such streams as tlie Solo-mon, Smoky Hill, and Saline, as also the northeast-flowing tributariesof the upper Republican, would seem likely localities for searchingout additional evidences of similar village sites in Kansas.The origins of the Central Plains tradition remain to be workedout. That it is basically of eastern or southeastern derivation seemsclear. The square earth lodge is well known from prehistoric culturesfarther south, in eastern Oklahoma and Arkansas; and as we haveseen, there is direct evidence of contacts between Smoky Hill valleysites and the lower Arkansas valley. JNIore accurate determination ofchronology in the two areas is needed before we can be certain of thesignificance of the contacts. Meanwhile, the general prevalence ofcord-roughened grit-tempered pottery, the hoe complex and asso-ciated domesticates, the common use of curved-bone fishhooks, and thebroad parallels in other items of material culture between the prehis-toric Central Plains and Late Woodland materials in the easternUnited States all point, I think, to the direxjtion in which most of theimportant Central Plains traits were rooted. Evidences of South-western contacts with the prehistoric Central Plains tradition of Kan-sas and Nebraska are lacking, despite some fairly extensive investiga-tions in the western Upper Republican area; and if the UpperRepublican people derived their house type from the northeasternPueblo area (Jennings et al., 1956, p. 87) , it seems strange that so littleelse in their material culture points in that direction. I have alreadyindicated my agreement with Spaulding's suggestion that a simplifiedversion of the Central Plains tradition may ultimately be recognized, Wedkl] an introduction TO KANSAS ARCHEOLOGY 629 out of which the Nebraska and Upper Republican Aspects specializedto the east and west, respectively, perhaps from something like whatI have termed the Smoky Hill Aspect.The relationships of the complexes to later cultures of the CentralPlains also await further investigation. No direct connection hasyet been demonstrated between the Nebraska Aspect and the laterSiouan peoples of the eastern Plains, or between Upper Republicanand the Pawnee who followed it in the region farther to the west.There are numerous similarities between Upper Republican and LowerLoup Focus, which is generally regarded as Pawnee; but no transi-tional or intermediate sites have been foimd in Kansas that can beregarded as part of a cultural sequence clearly connecting the pre-historic cultures with historic tribes. There is a small percentageof cord-roughened sherds in Great Bend Aspect sites in central Kan-sas ; but it has not been shown that they are a survival from Upper Re-publican times or, even if they were, that they establish an UpperRepublican ancestry for the Great Bend peoples.Transitional sites may yet be found, of course ; but there is growingevidence that they lie outside, or peripheral to, the Central Plains.Many of the Upper Republican sites are associated with old soilhorizons believed to represent former land surfaces subsequentlymantled by a few inches to two feet or more of wind-blown deposits.The overlying materials may reflect diminished rainfall, decreasedvegetative cover, and increased soil movement which, in turn, suggestsa possible climatic basis for abandonment of the old villages. TheUpper Republican communities, as we have seen, were pushed farwestward into a region of low uncertain precipitation which, even innormal years, borders on the minimum required for successful comgrowing ; and they must have been vulnerable to the intense droughtsand hot winds that are a characteristic feature of the regional climate.Even the creek bottom gardening of the Indians was not necessarilysafe from such climatic hazards, as I have pointed out elsewhere(Wedel, 1941 a, 1953 a) ; and prolonged droughts, with widespreaddrying up of springs and watercourses, might well have been animportant factor in forcing the Upper Republican horticultural com-munities out of their habitat in western Kansas and Nebraska, per-haps toward the east or north. That such droughts occurred inprehistoric times is indicated by tree-ring studies in western Nebraska(Weakly, 1940) ; and the one thus dated at 1439-1468,^^ if very wide-spread, may account for abandonment of the westerly Upper Republi-can area well before the arrival of the first Spanish in the central " "I am inclined to bellere that the 1459-68 and 1439-54 periods are not actuallyseparate droughts but constitute a single protracted one with not quite so severe conditionsbetween 1454 and 1459. These years were certainly subnormal . . ." (Weakly to Wedel,letter of March 28, 1957). 630 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174Kansas region. The strong element of Upper Republican materialculture manifested in late prehistoric sites in northeastern Nebraska,such as Lynch and St. Helena, as well as farther north at Arzberger,suggests where the bearers of Upper Republican culture went aftertheir presumed departure from the earlier habitat in the westCentral Plains.It is not possible to state with certainty to what extent or for howlong the inferred northward drift of Upper Republican peoples intlie 15th century left the Kansas region unoccupied or when newpeoples moved into the area. In the southern part of the State,the still very imperfectly known Pratt complex suggests a middleor late 15th-century northward thrust from some of the late pre-historic complexes of Oklahoma. So far as available sherd sam-ples, stone, and bone artifacts are indicative, this complex apparentlysliares more with the Custer and Washita River Foci than withUpper Republican. There are also resemblances to the PanhandleAspect, but perhaps less pronounced than those between the latterand Upper Republican. A semihorticultural subsistence economyis indicated, but nothing is known of the house types or settlementpatterns. Of some interest is the fact that here occur the earliestclear-cut instances of Pueblo pottery in association with a pre-Wliitecontact Kansas complex. Further investigations at Pratt, or in anyrelated sites that may be found south of the Arkansas River inKansas, are urgently needed.Farther north, inferentially also in the late 15th century, appeareda different complex represented by the Glen Elder Focus. Housetypes and the settlement pattern are again uncertain. Ceramically,an Upper Mississippi culture is indicated, but with a little simplestamping and cord roughening also present ; and the meager artifactinventory otherwise points to participation in widespread traitscommonlj^ found in other late prehistoric and protohistoric complexesof the Central Plains. The possibility that Glen Elder representsa Kansa occupancy suggests itself, but this cannot be demonstratedby any evidence now in hand.The earliest archeological remains in Kansas showing contact withwhite men are those around the great bend of the Arkansas Riverin the central part of the State. These have been assigned to theLittle River Focus, in turn identified with the Quiviran (Wichita)Indians met by Coronado in 1541. A slightly deviant manifesta-tion farther down the Arkansas, near its juncture with the Walnut,has been designated the Lower Walnut Focus. This and the LittleRiver Focus at present comprise the Great Bend Aspect,There are marked differences between these sites and those of theearlier Central Plains phase. Great Bend Aspect sites are frequently Wedbl] an introduction TO KANSAS ARCHEOLOGY 631larger, may be more compactly arranged, and include refuse moundsand great numbers of storage pits. The dwelling type has not beendetermined archeologically, and it seems doubtful that the earth-lodge, either square or circular in form, was present. At the sametime, there appear to have been specialized structures ("council-circles") in some of the villages for which no counterpart is knownin the prehistoric sites. Hunting was probably fully as importantas in earlier times, with bison bones especially plentiful in the refusedeposits. There is also evidence of a well-developed and intensivemaize horticulture. Charred ears of corn, well filled out and bear-ing 10, 12, and 16 rows of kernels, have been found; and the abun-dance of storage pits, ranging up to 8 feet in diameter and 10 feetin depth, suggests crop surpluses much larger than those of earlierpeoples. Numerous bison scapula hoes and grinding implementsseem confirmatory.The material culture inventory shows notable differences from thatof the prehistoric peoples. Probable absence of the earth lodge hasalready been noted. Missing also from Great Bend Aspect sites arebone fishhooks, bent tubular clay and equal arm stone pipes, andsmooth-bladed fleshing tools of mammal leg bone. The longitudi-nally grooved sandstone shaft smoother continues in large numbers,as does the rib arrowshaft wrench, beveled diamond-shaped knife,and the lanceolate knife. The cord-roughened pottery and globularto round-shouldered vessels of the Central Plains phase have givenway almost entirely to smoothed and simple-stamped wares, featur-ing high-rimmed vessels in which height tends to exceed diameter,handles are relatively plentiful, and flat disk bases are often present.Chipped stonework is more abundant and often better made. It in-cludes large numbers of finely made small end scrapers ; decided pre-ponderance of small triangular unnotched projectile points; nu-merous diamond-shaped and other beveled knives, including espe-cially a slender two-edged form with lateral notches and roundedbutt ; large numbers of expanded-base and plain shafted drills ; abun-dant use of fossiliferous Florence flint; and limited use of obsidian.In ground stone, such common Great Bend forms as the followingare rare or absent in Central Plains phase sites: grooved mauls ofvarious shapes ; large metates and one- or two-hand manos ( but com-pare Smoky Hill Aspect) ; fine-grained stone shaft polishers; L-shaped, high-bowled stone pipes; perforate and imperforate sand-stone disks ; and use of catlinite. The following bone artifact typesmay be added to the above list: socketed bison scapula hoes; flatsplit-rib and "rib-edge" awls; transversely scored mammal ribs;stemmed projectile points (both round and four-sided in cross sec-tion) ; wedge-shaped cancellous paint applicators ; and epiphyseal hide 632 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174dressers. Glaze-paint sherds, fine-grained stone shaft polishers, andturquoise indicates Southwestern contacts. There is no evidence ofthe horse or horse gear; but large dogs are indicated, and these doubt-less served as draft animals.Some of the traits just listed, as we have already noted, are wide-spread throughout the Plains in protohistoric times. They occurat Dismal Kiver sites in western Kansas and Nebraska, dated atcirca A. D. 1700 (?50 years), in Glaze levels at Pecos, at LowerLoup sites in east-central Nebraska, and yet farther to the north inthe Dakotas. In part, they include items that appear to be ofnorthern derivation, such as simple stampmg on pottery, use ofcatlinite, flat split-rib awls, and perhaps grooved mauls and epiphys-eal hide dressers. The toothed flesher, if eventually established forGreat Bend, is probably also northern. Others have a more decidedlysouthern distribution: the socketed or grooved scapula hoe of theLower Walnut Focus, foimd also in Sanders, Fort Coffee, and WashitaRiver Foci; L-shaped, high-bowled stone pipes, as in Sanders andFort Coffee; flat-based pottery vessels, occasionally with handles,and including shell-tempered wares; and the metate-mano complex,with back and forth grinding motion.The marked increase in abundance and variety of chipped-stoneartifacts, and their generally superior workmanship, as comparedto the prehistoric chipped-stone industry of the Central Plainsphase, has been pointed out elsewhere. For a number of the artifacttypes that appear at this time, such as "rib-edge" awls, scored ribs,cancellous bone paint applicators, 2-edged flint knives with lateralnotches, and round or 4-sided bone projectile points with or withoutstem, the available comparative data give little or no informationregarding time depth and possible source. Determination of theantecedents of the Great Bend Aspect, and relationships of the com-plex to such earlier manifestations as Pratt, Custer, Washita River,and Antelope Creek, clearly merit careful investigation when moreadequate samples are available from these manifestations.At present, what seems clear is that a diverse series of traits, in-cluding several associated with hunting and skin dressing, werecombined with older northern and southern items into an assemblagethat by the 16th century had spread widely over the south-centralPlains and had also come into direct contact with such frontier pueb-loan communities as Pecos. The similarities between the CentralPlains and eastern Pueblos are certainly far more striking andobvious at this level than in any previous one of which we haveknowledge at present. The Great Bend Aspect, moreover, looksmore like late prehistoric southern (Oklahoma and northern Texas) Wbdbl] an introduction TO KANSAS ARCHEOLOGY 633 cultures than does Lower Loup ; and the Little River Focus perhapsrepresents the northernmost extension of Wichita tribes from thesouth who still retained, ceramically and otherwise, a culture withunmistakable southern flavor. At the same time, to the extent thattheir subsistence pattern and material culture were oriented towarda bison-hunting livelihood, they were participating wholeheartedlyin practices widespread throughout the Plains. By contrast, theirPawnee kindred on the Loup River in Nebraska, possessed a ceramictradition, house types, village complex, and other traits that linkedthem with a more northerly pattern. So viewed, the Great BendAspect may represent a sort of southern coalescent tradition analogousto that of the historic Village Indians in Nebraska and northward(Lehmer, 1954 a).The Great Bend Aspect appears at present to be one of the mostdistinctive manifestations of native culture in Kansas. Systematicwork aimed at definition of its geographic extent and variations,and of course the determination of chronology within the complex,are greatly needed. That many more sites exist than we have con-sidered is well known ; their examination ought to throw importantlight on the antecedents of the culture, as well as on its later stages.Touching on the latter problem, perhaps, is the question of identityand relationships of several sites in nearby areas, as for example,the Deer Creek site in Kay County, Okla., and perhaps alsothe Neodesha "fort" in Wilson County, Kans. Both are apparentlylater in time than the Great Bend Aspect sites considered in thepresent paper; whether they represent stages in the withdrawalsouthward of the Wichita, perhaps under increasing pressures fromsuch alien peoples as the Plains Apache to the west and the SiouanOsage to the east, is one of several intriguing problems for which Ido not have the answer.Of other native Kansas peoples who were contemporaneous withthose of the Great Bend Aspect, we have at present no certain evi-dence. There are, however, two excellent prospects?the Oneota andthe Plains Apache. It is entirely probable, I think, that Oneotapeoples had already arrived in the trans-Missouri region while theGreat Bend Aspect still flourished in central Kansas. The Leary siteat the mouth of the Nemaha perhaps represents one such earlylocation. In Kansas, however, the only presently known Oneota siteis Fanning, which is pretty certainly late in the picture and may notbe directly in the line of descent from Leary. The frequency withwhich small quantities of European-made materials turned up inthe Fanning excavations suggests actual trade contacts of the late17th or early 18th century, rather than a chance early visit by white 634 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174men. The possibility must be recognized, of course, that other andearlier Oneota sites will yet be identified in the prairies of eastern ornortheastern Kansas, as along the Blue, Kepublican, or Osage Rivers.As will be apparent by reference to discussions elsewhere in thispaper, the Fanning complex differs widely from that in Great Bendsites. It is clearly of eastern derivation. Ceramically, its strongestaffiliations are with Oneota sites scattered from central Missourithrough Iowa to southern Minnesota and Wisconsin. Into this basic-ally eastern complex were incorporated various traits more character-istic of the Plains?some use of the earth lodge, deep undercut cachepits, bison scapula hoes, abundant small end scrapers, and boat-shaped sandstone shaft smoothers. Apparently lacking are many ofthe bone tools highly characteristic of the protohistoric Plains, in-cluding "rib-edge" and flat split-rib awls, scored ribs, cancellous bonepaint applicators, and hide-dressing tools, as well as the diversifiedand competent flint-chipping industry found in the Great BendAspect sites at the dawn of the White contact period and among theDismal River peoples in later times. It is tempting to see in theFanning complex, with its basically eastern affiliations, an earlyhistoric site of the Kansa Indians who, in the not long-distant past,were probably a prairie or even a marginal woodland people.The Plains Apache are represented in Kansas by scattered sherdoccurrences and one known village site, all situated in the westernthird of the State. The single village site, 14SC1, is from a latertime than the Great Bend Aspect sites considered in this paper, ifwe may trust the evidence of Rio Grande glaze paint sherds andpuebloan pipe fragments associated with it. A settlement of somesize centering around a stone pueblo is indicated, but nothing isImown of the usual house forms or of the arrangement of houseswithin the occupied area. The subsistence economy was probablybased primarily on the chase, with some cultivation of maize, squash,and presumably beans. Pottery of rather distinctive and unexcitingtype was made in some quantities ; and there are definite indicationsof contacts with Southwestern peoples on the Rio Grande, as wellas with Europeans. The extent to which Dismal River materialculture reflected the widespread Plains pattern, with respect to stone,bone, and other categories of artifacts, has been discussed elsewherein this paper.Of the Plains Apache prior to their appearance in Dismal Rivergarb, I know of no archeological evidence in the Kansas-Nebraskaregion. That they were present in the southern Plains by 1541, and inconsiderable numbers, is clear from the Spanish documents pertain-ing to the upper Rio Grande settlements and their surroundings. In Wedel] an introduction TO KANSAS ARCHEOLOGY 635the 16th century, the Apache groups east and northeast of New Mexicowere reportedly nonhorticultural ; and it is not clear just how suchnomadic groups as the Querecho and Teya were related to the latersemihorticultural Apache found north of the Arkansas Kiver around1700?that is, after the Pueblo Revolt and several recorded flights ofPueblo Indians to the Plains?and now identified by archeologistswith the Dismal Eiver culture. This is part of the larger problem ofAthabascan movement into the southern Plains and Southwest, towardwhich a notable contribution has recently been made. Drawing onethnohistorical, archeological, linguistic, and historic data, D. Gunner-son (1956) has suggested that the southern Athabascans, including thePlains Apache, arrived in the Texas-New Mexico plains not long be-fore Coronado, perhaps about 1525. Discovering the eastern pueblos,they attempted their conquest ; but, unsuccessful in this, they becameentrepeneurs of trade between the buffalo plains and the pueblos. Theabrupt appearance of a series of nonpuebloan stone and bone artifacttypes at Pecos in Glaze IV and V times has long been recognized(Kidder, 1932, pp. 43-44, 235, 238) as indicating easterly contactswith the Plains, and at about the time called for by Gunnerson'sthesis. Most of these types, as we have noted, are common in DismalRiver and Great Bend, and some occur also in other late prehistoriccomplexes of the southern Plains which precede in time the postulatedarrival of the Athabascans. If these artifacts were carried to Pecosby the Teyas and Querechos, as Gunnerson suggests, they were prob-ably acquired by these wandering hunters from earlier residents ofthe southern Plains represented archeologically by such semisedentarycomplexes as Antelope Creek, the Custer and Washita River Foci, andthe Great Bend Aspect and its forerunners. In any case, it is stilltrue that archeological sites which might be attributable to the early16th-century Apache "dog-nomads" are unknown or have not beenrecognized.The Dismal River sites apparently represent the terminal stage inthe Plains Apache occupation of the Central Plains. Those for whichacceptable dates are available fall within 25 to 50 years before and/orafter the year 1700. By this reckoning, Dismal River culture devel-oped principally in pre-horse days; and the animal which was tobecome a symbol of the Plains Indians for the next century or morewas acquired not long before the final southward displacement of theApache by the invading Comanche on horseback. This shifting ofpeoples in the High Plains between the Platte and Arkansas Rivers,as we have seen, appears to have been completed during the secondquarter of the 18th century; and by midcentury, the Comanche dom- 636 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174inated western Kansas. No archeological sites or materials attribut-able to the Comanche are known from the region.At about the time the Apache were being dispossessed by the Com-anche in western Kansas, the Kansa in the northeastern part of theState were coming into recorded contact with the French. In 1724,this tribe had its principal town on the Missouri, at present Doni-phan. The site is a scant 15 or 18 miles from the Fanning Oneotasite, for which a pre-1700 Kansa authorship and French contactshave been suggested. Our excavations at Doniphan yielded only afew cache pits and burials of the White contact period, some of thelatter being clearly intrusive into older Nebraska Aspect house sites.No satisfactory evidence of direct relationships to Fanning?or forthat matter with any other earlier sites or cultures?were noted.Since our work may not have been in the main village area, arche-ology can add little at present to the meager picture that has comedown to us from the French records. The settlement was appar-ently a large one, reportedly with 150 lodges and more than 1,000inliabitants. There is no indication, historical or archeological, asto the type of habitation, nor is there any evidence of stockades,ditches, or other defensive works. The cultivation of maize andother crops was practiced. There were few horses; and on the greattribal buffalo hunts to the west and southwest, the majority of theIndians depended on dogs or on their own backs for transport ofbaggage and meat. Very little pottery was found by us, nor wasthere much else in the way of native material culture. I am unableto say whether this is because we were on the edge of the main occu-pied area or, alternatively, reflects the early disappearance of nativearts and industries in the face of growing White penetration. Pos-sibly further search in the Doniphan locality, or at the later Kansasite north of Leavenworth to which the tribe moved when Fort Cav-agnolle was established, would disclose more adequate traces of nativeKansa culture. Our limited investigations near the mouth of BlueRiver, where a large compact, but unfortified, Kansa village of 100or more houses, some of them clearly earth lodges, stood from circa1800-1830, likewise disclosed very little Indian bone or stoneworkand no pottery. This looks like further evidence that the Kansa,situated nearer the sources of French contacts, lost their native craftsat an appreciably earlier time than did, for example, the more remotePawnee in Nebraska.In central Kansas, between the Comanche of the short-grass plainsand the prairie-dwelling Kansa, no village sites of the mid-18th cen-tury have yet been identified. The Great Bend Aspect sites hadapparently been given up and their erstwhile Wichita inhabitants Wbdbl] an introduction TO ELANSAS ARCHEOLOGY 637were withdrawing southward. Sometime during the third quarterof the 18th century, a portion of the Pawnee established themselvesin a fortified village on the Republican River not far south of theNebraska line. No contemporary description of the inhabited com-munity has come down to us; but the remaining traces include some30 or more house sites, cache pit depressions, and modest earth-walldefensive works. Excavations have disclosed circular earth lodgeswith 6 and 8 primary roof supports, small quantities of native pot-tery, stonework, and bone artifacts, and considerable amounts ofEuropean (especially French) contact materials. As has been indi-cated in a previous section, this is the only recorded Pawnee villagesite in the State, and its abandonment probably preceded the year1800. A larger village site also attributed to the Republican Pawneeand inhabited until about 1811, stood some 30 miles farther up theRepublican in present Webster County, Nebr. The outlines ofPawnee archeology during and after the period represented are fairlywell known, and the data and artifact materials from these two sitesaccord well with the historic Pawnee material culture complex.So far as the Kansas data are concerned, the known Pawnee andKansa sites in the State indicate that here, as elsewhere throughoutthe eastern Plains, the historic native populations were concentratedin a very few large towns, compactly arranged and sometimes forti-fied, and situated on the banks of the larger streams. Under constantpressure from the west by the footloose Comanche and, later, by theKiowa, Cheyemie, Arapaho, and Sioux, and from the east by Whiteswith trade goods, whiskey, and new diseases, the Village tribes wereattempting with indifferent success to combine their older horticultureheritage with the newer equestrian bison-hunting economy. Theirfailure and the resultant cultural collapse in the 19th century,reflected also in the archeological record, are matters of history.CONCLUSIONReviewing our materials in terms of the major objectives listed inthe introduction, it must be admitted that only partial success wasachieved. Clearly established was the fact that an abundance ofarcheological materials exists within the State, and the further factthat these show measurable differences from area to area and fromone time period to another. Sufficient data were gathered, despitethe limited time spent, to permit a tentative but plausible integrationof our findings with those made previously in adjacent States in everydirection. A major disappointment lay in the fact that the remainsof such historic tribes as the Kansa were found in very small quantities 638 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174and under such circumstances as to preclude any real insights intotheir antecedents and earlier cultural relationships.It is abundantly clear that the dearth of published informationon Kansas prehistory reflects principally the absence of any syste-matic or sustained attack on the archeology of the State, not a lackof suitable subject matter. This, of course, should have been clearto all decades ago, both from the carefully drawn reports of suchscientists as Udden and Williston, and from the brief but provoca-tive notices by a host of nonprofessional observers. Our investiga-tions, in point of fact, were concerned very largely with sites alreadyknown for years, either from the scanty literature or else from localsources. It is apparent, moreover, that we barely scratched the sur-face. Village and campsites exist in all pai'ts of the State ; they areparticularly numerous in the eastern and northeastern sections, butoccur also along most or all of the stream valleys westward into theHigh Plains. Large areas of considerable promise still await pro-fessional attention?in the Kepublican and Smoky Hill drainages,on the Neosho, Verdigris, and Osage Rivers, and in the southern andwestern areas. All that is required to tap this reservoir of informa-tion is a sustained program participated in by nonprofessionals aswell as by professionals. The combined efforts of both are needed toexpand further the present sketchy framework of human prehistoryhere and to fill in details.The results of the survey indicate further an interesting and im-portant variability in remains through time and space, and theseserve to document a long story of human endeavor. At the moment,it is undoubtedly true that there are more gaps than continuities inthe story, but the outlines are emerging with increasing clarity.Basically, it is a story of several millennia of food-collecting (i. e.,hunting-gathering) subsistence economies, followed by several cen-turies of food producing (i. e., horticulture), with hunting again com-ing into the ascendancy in the last century or two, but this time withthe horse as an important adjunct. So far as there is any evidence atall, the story begins with men who were living, perhaps ten thousandyears ago or more, when large game animals of species now extinctroamed the Kansas grasslands and furnished the principal subsistencebasis. Later, came the hunters and gatherers of Archaic and of EarlyWoodland times, about whom we still know almost nothing from theKansas region. The introduction of horticulture in Woodland times,probably within the Christian Era, and its subsequent intensificationgave rise sometime after A. D. 1000 to several semisedentary small-village cultures. From these developed larger communities whose Wedel] an introduction TO KANSAS ARCHEOLOGY 639economy rested increasingly on an intensive maize-bean-squash hor-ticulture and whose descendants were, in part, the Village Indiansof historic times. In Kansas, as in Nebraska, concentration of thehistoric tribes?the Kansa, Pawnee, and others?in one or two largevillages or towns for each tribe, completed a long sequence of chang-ing settlement patterns. From small camps and loosely arrangedrural hamlets, widely scattered and situated on small creeks as oftenas on the larger rivers, these communities appear to have grown pro-gressively larger in size and fewer in number until they culminatedin compact towns, sometimes fortified, of many hundreds of personsresiding along the major streams. With these changes, which maysafely be inferred from the archeological record, must have come pro-found shifts in social and political organization, and doubtless in othernonmaterial aspects of life as well. On these points, unfortunately,the archeological record has little to say.In many respects, the prehistory of Kansas as now seen parallelsthat of Nebraska. Two notable differences seem to be indicated, how-ever. One is the relatively greater abundance of demonstrable con-tacts between the prehistoric Kansans and the Pueblo peoples of theupper Eio Grande region. These appear to have been strongest fromthe 15th century to the 18th. There is much we still need to learnabout these contacts, an especially important point being the extentto which they may have involved groups possibly newly arrived inthe Southern Plains, such as the Apache. The probability of earliercontacts, which may have left less easily discernible traces, must alsobe considered. In any case, there is at present no evidence that anymarked deflection of Central Plains culture resulted from these inter-areal relationships. A second point of difference from Nebraska is thestronger southerly orientation of late prehistoric and protohistoricaboriginal culture in Kansas as far north as the great bend of theArkansas. Wliereas the earlier prehistoric remains of Kansas, as inNebraska, are clearly rooted in the east, influences from the Oklahomaregion are certainly present in southern and central Kansas in latertimes. This matter richly deserves far more attention than it has sofar received.With respect to the relationships of native man in Kansas to thenatural environment, there are also interesting leads in the availabledata. So far as we now know, no part of the State was continuouslyuninhabitable or unusable to man within the time span of his inferredpresence here. The manner in which he utilized the landscape, how-ever, clearly varied from time to time and from section to section. Inpart, this certainly reflected environmental stresses on his culture ; but 4&4172?59 42 640 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174in part it may also be viewed as an indication of man's growing abilityto adapt to varying environmental opportimities and limitations ashis cultural equipment improved. The earliest hunters of the Kansasarea appear to have been representatives of the western Paleo-Indiantradition ; and the discovery of Folsom and Plainview type points innortheastern Kansas suggests that these early groups may have spread,if thinly, over most or all of the State. For them, the inferred abun-dance of large and small game and of other natural food resources,would have made the entire Kansas area a prime hunting ground.To the later Archaic and Early Woodland hunters and gatherers, thegame-rich stream valleys running eastward to the oak-hickory forestsof Missouri, would similarly have offered a most inviting huntingground. Later, with introduction or adoption of maize, the regionaldifferences of climate, soils, and topography became increasingly sig-nificant. Settlement during the maize-pottery period was undoubt-edly always heaviest in the well-watered stream valleys of easternKansas, then as now the most favorable corn-growing sector of theState. There is evidence, however, that the small prehistoric UpperRepublican communities, at least, probably spread in northern Kansasfar to the westward where, owing to more effective precipitation, theCom Belt margin bends sharply toward the west. It may be suggestedthat the tributaries of the Smoky Hill, Solomon, and Saline Riverswould probably reward further search for prehistoric village sites ofthese early-day corn-growing peoples. Our work in Scott and LaneCounties has direct bearing on this point, for it brought to light sug-gestions of native maize horticulture far west of the area normallyoccupied by horticultural peoples in later times.In Kansas, as in Nebraska, there seems to have been a marked re-traction in the territory held by these native food producers beforethe white man arrived; and by the 16th century, corn-growing In-dians were largely restricted in Kansas to the region east of the 99thmeridian. The erstwhile Upper Republican range farther west wasin the hands of wandering hunters who only later acquired a feebleinterest in raising com. The probability that climatic instability anddrought were partly or largely responsible for this retraction of ter-ritory has been discussed elsewhere. In any case, the stage was setfor the observed historic alinement of Siouan, Algonkian, and otherhunting peoples in the western short-grass plains, of semisedentarymaize-growing Siouans to the east in the tall grass prairies, and ofsemihorticultural Caddoans occupying for a time the intermediatemixed-grass area in central Kansas. Further work may show more Whdbl] an introduction TO KANSAS ARCHEOLOGY 641 clearly to what extent the climatic factor, as contrasted to the his-torical and cultural, was instrumental in bringing about the observedhistoric groupings. As Kroeber (1939) has observed, "the immediatecauses of cultural phenomena are other cultural phenomena." Therecan be little doubt, however, that a technologically retarded societyattempting to subsist in any significant degree on maize growing inthe western Plains would have found its economic basis gravely im-periled by the recurrent and often devastating droughts that haveprobably always been a characteristic of the region.That further systematic researches in Kansas archeology, properlyplanned and adequately supported, can be expected to yield fruitfulresults, is apparent. So, also, is the fact that these can be made tocontribute significantly to better understanding of wider anthropo-logical problems in the Plains, particularly through full utilizationwherever possible of all relevant etlmohistorical and ethnographicmaterials. It seems possible that the simpler nature of much of theprehistoric Kansas (and Nebraska) material, as compared to the rel-atively rich and abundant remains on the Middle Missouri, may re-flect the fact that we are dealing here with the earlier antecedentsof Plains culture, the initial stages perhaps of man's efforts to de^velop a food-producing society adapted to the peculiar environmentof the Plains region. Manifestly, the geographic location of Kansashas exposed its prehistoric and later inhabitants to cultural influencesand ethnic movements from a number of directions and from severalhighly developed centers of cultural differentiation. There are strongsuggestions in the very limited data available of marked physicaldifferences between the bearers of some of the different material cul-ture inventories observed. In the long-range view, of course, con-sideration must also be given to the environmental variability and to acertain climatic instability that characterize the area. All these con-siderations lead to the final observation that the region with whichthis paper has been primarily concerned offers an excellent oppor-tunity to enrich further our understanding of culture growth andchange, and of hiunan ecology, in the Great Plains. 642 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174LITERATURE CITEDAbel, Anna H.1904. Indian reservations in Kansas and the extinguishment of their title.Trans. Kansas State Hist. See, vol. 8, pp. 72-109.1939. Tabeau's narrative of Loisel's expedition to the upper Missouri.Norman, Okla.Adams, F. G.1904. Reminiscences of Frederick Chouteau. Trans. Kansas State Hist.Soc, vol. 8, 1903-4, pp. 423-434.1906. Kansas mounds. Trans. Kansas Acad. Science, vol. 6, pp. 51-54.Adams, George I.1902. Physiographic divisions of Kansas. Bull. Amer. Geogr. Soc, vol. 34,No. 2, pp. 89-104.AlTON, A. S.1939. Coronado's muster roll. Amer. Hist. Rev., vol. 44, No. 3.Allen, J. A.1876. The American bisons, living and extinct. Mem. Geol. Surv. Ken-tucky, vol. 1, pt. 2.Allison, V. C, and Randolph, V.1927. Prehistoric inhabitants of Crawford County, Kansas. Amer. Anthrop.,n. s., vol. 29, No. 3, pp. 258x-261x.Anonymous.1881. The Kansas Indians. Trans. Kansas State Hist. Soc, vols. 1-2, pp.280-301.Ayer, Mrs. Edward E. (transl. ) . 1916. The memorial of Fray Alonso de Benavides, 1030. Chicago.Baebour, E. H., and Schultz, C. B.1932. The mounted skeleton of Bison occidentalis, and associated dart-points. Bull. Nebraska State Mus., vol. 1, No. 32, pp. 263-270.Barnes, Lela, Editor.1936 a. Journal of Isaac McCoy for the exploring expedition of 1828.Kansas Hist. Quart, vol. 5, No. 3, pp. 227-277.1936 b. Journal of Isaac McCoy for the exploring expedition of 1830.Kansas Hist. Quart., vol. 5, No. 4, pp. 339-377.Beabdsley, Gretchen.1940. The groundnut as used by the Indians of eastern North America.Pap., Michigan Acad. Sci., Arts, and Letters, vol. 25 (1939), pp.507-515.Beauchamp, W. M.1902. Metallic implements of the New York Indians. New York State Mus.Bull 55. Albany.Beckwith, E. G.1855. Report of exploration of a route for the Pacific Railroad, near the38th and 39th parallels of latitude, from the mouth of the Kansasto Sevier River, in the Great Basin. Washington.Bell, Earl H., and Gilmore, G. H.1936. The Nehawka and Table Rock foci of the Nebraska Aspect. Chap,in Nebraska Archeol., No. 4, pp. 301-355. Univ. Nebraska, Lincoln.Bell, Robert E., and Baebreis, David A.1951. A survey of Oklahoma archeology. Bull. Texas Archeol. and Paleont.Soc, vol. 22, pp. 7-100. Wbdbl] an INTRODUCTTION TO KANSAS ARCHEOLOGY 643Bennett, J. W.1945. Archaeological explorations in Jo Daviess County, 111. Univ. ChicagoPubl. Anthrop., Archaeol. Series.Berry, Brewton ; Chapman, Carl ; and Mack, .John.1944. Archeological remains of the Osage. Amer. Antiq., vol. 10, No. 1,pp. 1-11.Billings, M. O.1883. Ancient remains in Marion County, Kansas. Kansas City Rev. Sc. andInd., vol. 6, pp. 211-212.Bolton, H. E.1911. The Jumano Indians in Texas, 1650-1771. Texas Hist. Assoc. Quart.,vol. 15, No. 1, pp. 66-84. July.1914. Athanase de Mezieres and the Louisiana-Texas Frontier, 1768-1780. 2vols. Cleveland.1915. Texas in the middle eighteenth century. Univ. California Publ. inHist., vol. 3. Berkeley.1916. Spanish exploration in the Southwest, 1542-1706. New York.1949. Coronado: Knight of Pueblos and Plains. Whittlesey House andUniv. New Mexico Press.Bradbury, John.1817. Travels in the interior of America in the years 1809, 1810, and 1811.Liverpool and London.Brower, J. V.1898. Quivira. Mem. Explorations in the Basin of the Mississippi, vol. 1.St. Paul.1899. Harahey. Mem. Explorations in the Basin of the Mississippi, vol. 2.St. Paul.1904. Mandan. Mem. Explorations in the Basin of the Mississippi, vol. 8.St. Paul.Beown, Wm. L., and Anderson, Edgab.1947. The northern flint corns. Annals Missouri Botanical Garden, vol. 34,No. 1, pp. 1-29. St. Louis.Bryan, F. T.1857. Explorations for road from Fort Riley to Bridger's Pass, 1856. App.H, Rep. Chief, Topographical Engineers, Ann. Rep. War Dept., 1857.35th Congr., 1st sess.. Exec. Doc. vol. 2, No. 2.Cableton, J. Henry.1943. A dragoon campaign to the Rocky Mountains in 1845. The PrairieLogbooks, pp. 1-53-280. (Edited and with an introduction by LouisPelzer. Caxton Club, Chicago. ) Chamberlin, T. C.1902. The geologic relations of the human relics of Lansing, Kansas,Joum. Geol., vol. 10, No. 7, pp. 745-779.Champe, John L.1936. The Sweetwater Culture Complex. Chap. Nebraska Archeol. No. 3,pp. 249-297, Univ. Nebraska, Lincoln.1946. Ash Hollow Cave. Univ. Nebraska Stud., n. s., No. 1. Lincoln.1949 a. White Cat Village. Amer. Antiq., vol. 14, No. 4, pt. 1, pp. 285-292,April.1949 b. A report for the Laboratory of Anthropology, Univ. Nebraska,1940-1947. Proc. Fifth Plains Conf. for Archeol. Note Book No.1, Lab. Anthrop., Univ. Nebraska, pp. 14-17. 644 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174Chapman, Cabl H.1946-48. A preliminary survey of Missouri archeology. Missouri Archeol.,vol. 10, pts. 1^ BuU. 20-23, pp. 1-164. Columbia.1952. Culture sequence in the Lower Missouri Valley. In Archeology ofEastern United States, pp. 139-151. J. B. Griffin, ed. Univ. ChicagoPress.Chouteau, Auguste.1816. Notes. Manuscript in National Archives, Records of the Bureauof Indian Affairs, Ancient and Miscellaneous Surveys. 1. vol. 4.Washington.COCKRUM, E. L.1952. Mammals of Kansas. Univ. Kans. Publ. Mus. Nat. Hist, vol. 7,No. 1, pp. 1-303.Cole, Fannie E.1904. The Kansas Indians in Shawnee County after 1855. Trans. KansasState Hist. Soc, vol. 8, 1903^, pp. 481-483.CoLLOT. Victor.1924. A journey in North America. 2 vols., 1 atlas. Paris, 1826.(Translated and reprinted, Firenze, 1924.)CONNELLEY, W. E.1918. Notes on the early Indian occupancy of the Great Plains. Coll. KansasState Hist. Soc. ( 1915-18 ) , vol. 14, pp. 438-470.Cooper, Paul1936. Archeology of certain sites in Cedar County, Nebraska. Chap, inNebraska Archeol. No. 1, pp. 11-145. Univ. Nebraska, Lincoln.1940. Report of explorations (1938). Nebraska Hist., vol. 20, No. 2, pp.94-151.1955. The archeological and paleontological salvage program in the MissouriBasin, 1950-51. Smithsonian Misc. Coll., vol. 126, No. 2.Crane, H. R.1956. University of Michigan Radiocarbon Dates I. Science, vol. 124,No. 3224, pp. 664-672. October 12.Crane, H. R., and Griffin, J. B.1958. University of Michigan Radiocarbon Dates III. Science, vol. 128,No. 3332, pp. 1117-1123. November 7.CuLiN, Stewart.1907. Games of the North American Indians. 24th Ann. Rep., Bur. Amer.Ethnol., 1902-03. Washington.CuMMiNG, Robert B., Jr.1958. Archeological investigations at the Tuttle Creek Dam, Kansas. Bur.Amer. Ethnol. Bull. 169, Riv. Bas. Surv. Pap. No. 10.Curtis, F. S., Jr.1927. Spanish arms and armor in the Southwest. New Mexico Hist. Rev.,vol. 2, No. 2, pp. 107-133.Dean, Bashford.1930. Handbook of arms and armor. Metropolitan Mus. Art. New York.Delanglez, Jean.1939. Tonti letters. Mid-America, vol. 21, No. 3, pp. 209-238. Inst. JesuitHistory, Loyola Univ., Chicago.1943. Franquelin, mapmaker. Mid-America, vol. 25, No. 1, pp. 29-74.Inst Jesuit History, Loyola Univ., Chicago. WEDKLl AN INTRODUCTION TO KANSAS ARCHEOLOGY 645Dellingee, S. C, and Dickinson, S. D.1942. Pottery from the Ozark bluff shelters. Amer. Antiq., vol. 7, No. 3,pp. 27&-289, January.DiNSMOEE, R. S.1912. A visit to an old Indian village site near Sparks, Kansas. Archeol.Bull., vol. 3, No. 2, pp. 53-54.Dodge, Richabd I.1877. The plains of the great west and their inhabitants. New York.DoBSEY, Geoege a.1904. Mythology of the Wichita. Carnegie Inst. Washington, Publ. No. 21.DoBSEY, J. Owen.1886. Migrations of Siouan tribes. Amer. Nat., vol. 20, No. 3, pp. 211-222.Du Lao, Peebin.1807. Travels through the two Loulsianas in 1801, 1802, and 1803. London.DuNBAE, John B.1880. The Pawnee Indians : their history and ethnoloyy. Mag. Amer. Hist.,vol. 4, No. 4, pp. 241-281.DuNLEVY, MaeION LtJCILE.1936. A comparison of the cultural manifestations of the Burkett (NanceCoimty) and the Gray-Wolfe (Colfax Coimty) sites. Chap, in Ne-braska Archeol., No. 2, pp. 147-247, Univ. Nebraska, Lincoln.Du Peatz, Le Page.1774. The history of Louisiana. London.Eiseley, L. O.1939. Evidences of a pre-ceramic cultural horizon in Smith County, Kansas.Science, vol. 89, p. 221.Emoby, W. H.1848. Notes of a military reconnaissance, from Fort Leavenworth, in Mis-souri, to San Diego, in California, including parts of the Arkansas,del Norte, and Gila Rivers. Made in 1846-47, with the advancedguard of the "Army of the West." New Yoi*k.EWEES, J. C.1957. Hair pipes in Plains Indian adornment. Bur. Amer. Ethnol. Bull.164, Anthrop. Pap. No. 50, pp. 29-86.Byman, Chables.MS. The Schultz Focus: A Middle Woodland mound complex In the lowerRepublican Valley, Kansas. MS. (n. d.), Univ. Kansas Mus. Nat.Hist.Fenneman, N. M.1931. Physiography of western United States. New York.Fletchee, Alice, and La Flesche, Feancis.1911. The Omaha tribe. 27th Ann. Rep. Bur. Amer. Ethnol., 1905-06.Floea, S. D.1948. Climate of Kansas. Rep. Kansas State Board of Agric, vol. 67, No.285. June.Fobd, J. A., and Willey, G. R.1940. Crooks site, a Marksvllle period burial mound In La Salle parish,Louisiana. Dept. of Conserv., Louisiana Geol. Sur., Anthrop.Study No. 3. 148 pp., illua.Fowke, G.1922. Archeologlcal investigations. Bur. Amer. Ethnol., Bull. 76. 646 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174Fbeed, Maby Louise.MS. The Lynch site, 25BD1. M. A. thesis, Dept. Anthrop., Univ. Nebraska,Lincoln. 1954.Fkemont, J. C.1845. Report of the exploring expedition to the Rocky Mountains in theyear 1842, and to Oregon and California in the years 1843?i4.House Doc. 166, Washington.Fbench, B. F.1852. Historical collections of Louisiana, embracing translations of manyrare and valuable docimients relating to the natural, civil, andpolitical history of that state . . . Vol. 4. Redfield, N. Y.Fkye, J. C.1946. The High Plains surface in Kansas. Trans. Kansas Acad. Sci., vol.49, pp. 71-86.Feye, J. C, and A. B. Leonard.1952. Pleistocene geology of Kansas. State Geol. Surv. Kansas, Bull. 99.Lawrence.FuLLBX, H. M., and Hafen, L. R., Editors.1957. The journal of Captain John R. Bell . . . 1820. Glendale, Calif.Gates, F. C.1937. Grasses in Kansas. Rep. Kansas State Board Agric, vol. 55, No.220-A, pp. 1-349.1940. Flora of Kansas. Kansas State Coll., Contr. Dept. Botany, No. 391.George, Wm. H., and George, Edythe L.1945. Villages on the banks of the Missouri River in South Dakota. Minn.Archeol., vol. 11, No. 3, pp. 61-67. Minneapolis.Gilmoee, M. R.1919. Uses of plants by the Indians of the Missouri River region. 33d Ann.Rep., Bur. Amer. Ethnol., pp. 43-154.Gould, C. N.1898 a. The timbered mounds of the Kaw Reservation. Trans. Kansas Acad.Sci. vol. 15, pp. 78-79.1898 b. Prehistoric mounds in Cowley County. Trans. Kansas Acad. Sci.,vol. 15, pp. 79-80.1899. Additional notes on the timbered mounds of the Kaw Reservation.Trans. Kans. Acad. Sci., vol. 16, p. 282.Gray, P. L.1905. Gray's Doniphan County history. Bendena, Kans.Griffin, James B.1937 a. The archaeological remains of the Chiwere Sioux. Amer. Antiq.vol. 2, No. 3, pp. 180-181.1937 b. Culture identity of the Ozark "Top Layer." Amer. Antiq., vol. 2,No. 4, pp. 296-297. April.1943. The Fort Ancient Aspect. Univ. Michigan Press, Ann Arbor.1958. The chronological position of the Hopewellian culture in the EasternUnited States. Mus. Anthrop., Univ. Mich., Anthrop. Pap. No. 12.Griffin, J. W.MS. The Upper Mississippi occupations of the Fisher site, Will County,Illinois. Univ. Chicago. (Unpublished M. A. thesis, n. d.)Geiffing, W. J.1904. Committee on explorations. Trans. Kansas State Hist. Soc, 1903-04,pp. 133-135. Wkdel] an introduction TO KANSAS ARCHEOLOGY 647Geinnell, George B.1893. Pawnee hero stories and folk-tales. New York.1920. Who were the Padouca? Amer. Anthrop., n. s., vol. 22, No. 3, pp.248-260.1923. The Cheyenne Indians. Vols. 1 and 2. New Haven.GuNNERSON, Dolores A.1956. The southern Athabascans: Their arrival in the Southwest. ElPalacio, vol. 63, Nos. 11-12, pp. 346-365. Santa Fe.Gunnerson, James H.1952. Some Nebraska Aspect pottery types. Plains Archeol. Conf., NewsLetter, vol. 5, No. 3, pp. 39-48. Lincoln.1956. Plains-Promontory relationships. Amer. Antiq., vol. 22, No. 1, pp.69-72.1959. An introduction to Plains Apache archeology?the Dismal RiverAspect. Bur. Amer. Ethnol. Bull. 173, Anthrop. Pap. No. 58.Haas, Mart.1942. Comments on the name "Wichita." Amer. Anthrop., n. s., vol. 44,No. 1, pp. 164-65.Hackett, Chas. W., Editor.1923. Historical documents relating to New Mexico, Nueva Vizcaya, andapproaches thereto, to 1773. Carnegie Inst. Washington, Publ.330, vol. 1.Hamilton, Henry W.1952. The Spiro mound. Missouri Archeol., vol. 14. Columbia, Mo.Hamilton, Raphael N.1934. Early maps of the Missouri Valley. Amer. Hist. Rev., vol. 39, No. 4,pp. 645-662.Hammond, G. P., and Rey, A.1940. Narratives of the Coronado Expedition. Coronado Hist. Ser., vol. 2,Univ. New Mexico Press.Harrington, John P.1940. Southern peripheral Athapaskawan origins, divisions, and migra-tions. Smithsonian Misc. Coll., vol. 100, pp. 503-532.Harrington, M. R.1920. Certain Caddo sites in Arkansas. Indian Notes and Monogr., Misc.Ser., No. 10. New York.Harris, John.1705. Naviganteum atque Itinerantium bibliotheca, or a complete collec-tion of voyages and travels. 2 vols. London.HaWORTH, Erasmus.1897. Physiography of western Kansas. Kansas Univ. Geol. Surv., vol. 2,pp. 11-49.Hill, A. T., and Cooper, Patil.1937. The Schrader, Champe, and Fremont 1 sites. Nebraska Hist. Mag.,vol. 17, No. 4, pp. 221-292. Oct.-Dec. 1936.1938. The archeological campaign of 1937. Nebraska Hist. Mag., vol. 18,No. 4, pp. 237-359. Oct.-Dec. 1937.Hill, A. T., and Kivett, M. F.1941. Woodland-like manifestations in Nebraska. Nebraska Hist. Mag.,vol. 21, No. 3, pp. 147-243. July-Sept. 1940.Hill, A. T., and Metcalf, George.1942. A site of the Dismal River Aspect in Chase County, Nebraska. Ne-braska Hist. Mag., vol. 22, No. 2, pp. 158-226. April-June 1941. 648 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174Hill, A. T., and Wedel, Waldo R.1936. Excavations at the Leary Indian village and burial site, RichardsonCounty, Nebraska, Nebraska Hist. Mag., vol. 17, No. 1, pp. 2-73.HODQE, F. W.1900. Pueblo ruins in Kansas. Amer. Anthrop., n. s., vol. 2, No. 4, p. 778.1910. The Jumano Indians. Amer. Antiq. Soc, Proc, n. s., vol. 20, pp.249-268.1920. Hawikuh bonework. Mus. Amer. Indian, Heye Found., Indian Notesand Monogr., vol. 3, No. 3, pp. 65-151.HoDQE, F, W., Editor.1907-10. Handbook of American Indians north of Mexico. Bur. Amer.Ethnol. Bull. 30. 2 pts.HOLDEN, W. C.1933. Excavation of Saddle-Back ruin. Bull., Texas Arch, and Paleont.Soc, vol. 5, pp. 39-52.Holmes, W. H.1902. Fossil human remains found near Lansing, Kansas. Amer. Anthrop.,n. 8., vol. 4, pp. 743-752.1903. Aboriginal pottery of the eastern United States. 20th Ann. Rep.,Bur. Amer. Ethnol., pp. 1-201.Hooton, E. a., and Willoughbt, C. C.1920. Indian village site and cemetery near Madisonville, Ohio. Pap. Pea-body Mus. Amer. Archeol. and Ethnol., vol. 8, No. 1. Cambridge.HOENADAY, W. T.1889. The extermination of the American bison, vs^ith a sketch of its discov-ery and life history. Ann. Rep. Smithsonian Inst, for 1887, pt. 2,pp. 367-548.HoucK, Louis.1909. The Spanish regime In Missouri. 2 vols. Chicago.HovTABD, Edgar B.1935. Evidence of early man in North America. Mus. Journ., vol. 24, Nos.2-3, pp. 61-175. Philadelphia.HbduSka, A.1907. Skeletal remains suggesting or attributed to early man in NorthAmerica. Bur. Amer. Ethnol. Bull. 33.1927. Catalogue of human crania in the United States National Museumcollections. (The Algonkin and related Iroquois; Slouan, Caddoan,Sallsh and Sahaptin, Shoshonean, and California Indians.) Proc.U. S. Nat. Mus., vol. 69, art. 5, Publ. 2631, 127 pp.HULBERT, A. B.1933. Southwest on the Turquoise Trail : The first diaries on the road toSanta Fe. Overland to the Pacific, vol. 5. Stewart Comm. Colo-rado College and Denver Public Library.Hunt, C. B.1954. Pleistocene and Recent deposits in the Denver area, Colorado. U. S.Geol. Surv., Bull. 996-C. Washington.Hurt, Wesley R.1952. Report of the Investigation of the Scalp Creek Site 39GR1 and theEllis Creek Site 39GR2, Gregory County, South Dakota. SouthDakota Archeol. Comm., Archeol. Stud., Cir. No. 4. Pierre, S. D.1953. Report of the investigation of the Thomas Riggs Site 39HU1, HughesCounty, S. D. South Dakota Archeol. Comm., Archeol. Studies,Clr. No. 5. Pierre, S. Dak. WedelJ an introduction TO KANSAS ARCHEOLOGY 649Hyde, George E.1934. The Pawnee Indians. Pts. 1-2. Denver.1951. Pawnee Indians. Univ. Denver.Jackson, A. T.1933. Indian pipes of East Texas. Bull. Texas Archeol. and Paleont. See,vol. 5, pp. 69-86.1938. Picture-writing of Texas Indians. Univ. Texas Publ. No. 3809,Anthrop. Pap. vol. 2.Jahes, Edwin.1S23. Account of an expedition from Pittsburgh to the Rocky Mountains,performed in the years 1819-20 . . . under the command of MajorStephen H. Long. 2 vols. Philadelphia.Jeancon, J. A.1923. Excavations in the Chania Valley, New Mexico. Bur. Amer. Ethnol.Bull. 81.Jenkins, Meble T.1941. Influence of climate and weather on growth of corn. In Climate andMan, U. S. Dept. Agric. Yearbook of Agriculture, 1941, pp. 308-320.Jenks, A. E.1932. The problem of the culture from the Arvilla gravel pit. Amer.Anthrop., n. s., vol. 34, No. 3, pp. 455-466.Jennings, J. D. ; Reed, E. K. ; et al.1956. The American Southwest: A problem in cultural isolation. Mem.,Soc. Amer. Archeol., No. 11, pp. 59-127. Salt Lake City.Johnson, C. P.1897. Mound in Cowley County, Kansas. Antiquarian, vol. 1, pt. 4, pp. 95-96.Johnson, W. D.1900. The High Plains and their utilization. U. S. Geol. Survey, 21st Ann.Rep., pt. 4, pp. 601-741.Johnston, C. Stewabt.1939. A report on the Antelope Creek ruin. Bull. Texas Archeol, and Paleont.Soc, vol. 11, pp. 190-202.Jones, Horace.1928. The story of early Rice County. Wichita, Kans.Jones, Paul A.1929. Quivira. Wichita, Kans.1937. Coronado and Quivira. Lyons, Kans.Jones, Volney H.1949. Maize from the Davis site, its nature and interpretation. Soc. Amer.Archeol., Mem. 5, pp. 239-249.Jones, V. H., and Fonner, R. L.1954. Plant materials from sites in the Durango and La Plata areas, Colo-rado. Carnegie Inst Washington, publ. No. 604, pp. 93-115.KLansas State Board op Agbicultube.1929. Corn in Kansas. Rep. Kansas State Board of Agric, vol. 48, No. 191.Topeka.Keixogg, Royal S.1905. Forest Belts of western Kansas and Nebraska. U. S. Dept. Agric,Forest Service, Bull. 66.Keyes, Charles R.1927. Prehistoric man in Iowa. Palimpsest, vol. 8, No. 6, pp. 185-229. IowaCity. (Reprinted, with revisions, in The Palimpsest, vol. 32, No.8, 1951). 650 BUREAU OF .\MERICAN ETHNOLOGY [Bull. 174KEYES, Chables R.?Continued1949. Four Iowa archeologies with Plains aflaiiations. Proc. Fifth PlainsConf. for Archeol., Notebook No. 1, Lab. Anthrop., Univ. Nebraska,pp. 96-97.KiDDEB, A. V.1932. The artifacts of Pecos. Pap. Southwestern Exi^edition No. 6, PhillipsAcademy, Yale University Press, New Haven.1958. Pecos, New Mexico : archeological notes. Pap. R. S. Peabody Found.Archaeol., vol. 5. Andover, Mass.KiNNAiED, Lawrence, Editor.1946. Spain in the Mississippi Valley. Part 2 : Post war decade, 1782-1791.Ann. Rep. Amer. Hist. Assoc., 1945, vol. 3, pt. 2. Washington.1949. Spain in the Mississippi Valley, 1765-1794. Pt. 1, The RevolutionaryPeriod, 1765-1781. Ann. Rep. Amer. Hist. Assn., 1945, vol. 2.KrvETT, M. F.1949. Archeological investigations in Medicine Creek Reservoir, Nebraska.Amer. Antiq., vol. 14, No. 4, pt. 1, pp. 278-284. April.1950. Archeology and climatic implications in the Central Plains. Proc.Sixth Plains Archeol. Conf., 1948, Univ. Utah, Dept. Anthrop.,Anthrop. Pap. No. 11.1952. Woodland sites in Nebraska. Nebraska State Hist. Soc, Publ. An-thropology, No. 1. Lincoln.1953. The Woodruff ossuary, a prehistoric burial site in Phillips County,Kansas. Bur. Amer. Ethnol. Bull. 154, Riv. Bas. Surv. Pap. No. 3.Keiegeb, Alex D.1946. Culture complexes and chronology in northern Texas, with extensionof puebloan datings to the Mississippi Valley. Univ. Texas Pub-lication No. 4640. Oct. 22.1947. The eastward extension of puebloan datings toward cultures of theMississippi Valley. Amer. Antiq., vol. 12, No. 3, pp. 141-148.1949. The southern limit of Central Plains cultural complexes. Proc. FifthPlains Conf. for Archeol., pp. 98-99, Note Book No. 1, Lab. Anthrop.,Univ. Nebraska.Kboebeb, a. L.1939. Cultural and natural areas of native North America. Univ. Cali-fornia Publ. Amer. Archeol. and Ethnol., vol. 38.La Flesche, Francis.1924. Omaha bow and arrow-makers. Annaes do XX Congresso Interna-cional de Americanistes, vol. 1, pp. 111-116. Rio de Janeiro.1928. The Osage tribe, two versions of the child-naming rite. 43d Ann. Rep.,Bur. Amer. Ethnol., pp. 23-164.Lange, Charles H.1953. A reappraisal of evidence of Plains influences among the Rio GrandePueblos. Southwestern Journ. Anthrop., vol. 9, No. 2, pp. 212-230.Lehmeb, Donald J.1954 a. Archeological investigations in the Oahe dam area. South Dakota,1950-51. Bureau of Amer. Ethnol., Bull. 158, Riv. Bas. Surv. Pap.No. 7.1954 b. The sedentary horizon of the Northern Plains. Southwestern Journ.Antlirop., vol. 10, No. 2, pp. 139-159. Summer. Univ. New MexicoPress, Albuquerque.Leightt, C. E. ; Warburton, C. W. ; Stine, O. C. ; and Baker, O. E.1922. The com crops. U. S. Dept. Agric. Yearbook of Agriculture, 1921,pp. 161-226. WedeM an introduction to KANSAS ARCHEOLOGY 651Lessee, A., and Weltfish, G.1932. Composition of the Caddoan linguistic stock. Smithsonian Misc. Coll.,vol. 87, No. 6.Lobby, W. F.1954. Chicago Radiocarbon Dates V. Science, vol. 120, No. 3123, pp. 733-742. November 5.Logan, Wilfred D.1952. Graham Cave, an Archaic site in Montgomery County, Mo. Mem.,Missouri Archeol. Soc, No. 2. Columbia.LowEKY, Woodbury.1912. A descriptive list of maps of the Spanish possessions within thepresent limits of the United States, 1502-1820. Washington.LowiE, Robert H.1953. The relations between the Kiowa and the Crow Indians. Bull. Soc.Suisse des Americanistes, No. 7. Geneva.Malleby, Garbick.1893. Picture-writing of the American Indians. 10th Ann. Rep., Bur.Ethnol., pp. ^807.Mabgey, Pierre.1886. Exploration des affluents du Mississippi et decouverte des MontaguesRocheuses (1679-1754). Decouvertes et establissements desFrancais dans I'ouest et dans le sud de I'Amerique Septentrionale(1614^1754). Memoires et documents originaux. Pt. 6. Paris.1875-86. Decouvertes et establissements des Francais dans I'ouest et dansle sud de I'Amerique Septentrionale (1614-1754). Memoires etdocuments originaux. Pts. 1-6. Paris.Martin, H. T.1909. Further notes on the pueblo ruins of Scott County. Kansas Univ.Sci. Bull., vol. 5, No. 2, pp. 11-22.Mason, O. T.1891. Aboriginal skin-dressing?a study based on material in the U. S. Na-tional Museum. Ann. Rep. U. S. Nat. Mus. for 1889, pp. 553-589.Mason, Silas C.1882. The pictured rocks of Pipe Creek. Trans. Kansas Acad. Sci., vol. 8,p. 4 (title only).McClung, C. E.1908. Restoration of the skeleton of Bison occidentalis. Kansas Univ. Sci.Bull., vol. 4, pp. 249-252.McCoy, Isaac.1840. History of Baptist Indian Missions. Washington.McDebmott, J. F., Editor.1940. Tixier's travels on the Osage prairies. Univ. Oklahoma Press,Norman.McGee, W. J.1897. The Siouan Indians. 15th Ann. Rep. Bur. Amer. Ethnol., pp. 153-204.McKebn, W. C.1945. Preliminary report on the Upper Mississippi Phase in Wisconsin.Bull. Public Mus. Milwaukee, vol. 16, No. 3, pp. 109-285.Mead, J. R.1890. Camps of prehistoric people in Sedgwick County, Kansas. Trans.,Kansas Acad. Sci., vol. 12, pp. 64-65.1899. Some Natural History Notes of 18.59. Trans., Kansas Acad. Sci., vol.16, pp. 280-281. 652 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174 Meiaj), J. R.?Continued1904. The Wichita Indians in Kansas. Trans. Kansas State Hist. Soc.,1903-4, pp. 171-177.1905. Notes on the archeology of Butler County. Trans., Kansas Acad.Sci., vol. 19, pp. 329-330.1906. The Saline River country in 1859. Trans. Kansas State Hist. Soc.,1905-6, vol. 9, pp. 8-19.Meleen, E. E.1938. A preliminary report of the Mitchell Indian village site and burialmounds. Univ. South Dakota Mus., Archeol. Stud., Circ. 2, pp. 1-34.Melish, John.1818. United States of America. Compiled from the latest and best Authori-ties. Entered according to Act of Congress 10th July 1818, and pub-lished by Murray Draper Fairman & Co. Philadelphia.MicHELSON, Truman.1921. Who vrere the Padouca? Amer. Anthrop., vol. 23, No. 1, p. 101.Miller, Edward.1869. Explanatory memoir on a series of photographic representations ofremarkable rocks, on Smoky River, Kansas. Amer. Phil. Soc., Proc,vol. 10, No. 79, pp. 382-384.MooNET, James.1896. The Ghost-Dance religion and the Sioux outbreak of 1890. 14th Ann.Rep., Bur. Ethnol., pt. 2.1898. Calendar history of the Kiowa Indians. 17th Ann. Rep. Bur. Amer.Ethnol., pp. 129-445.MOOREHEAD, W. K.1931. Archeology of the Arkansas River Valley. Tale Univ. Press, NewHaven.Morehouse, G. P.1908. History of the Kansa Indians. Trans. Kansas State Hist. Soc.,1907-8, vol. 10, pp. 327-368.1910. Archeology of Diamond Valley, southwestern Morris County, Kansas.Archeol. Bull., vol. 1, No. 2, pp. 34r-37.Morse, jEDn)LA.H.1822. A report to the Secretary of War of the United States, on Indian af-fairs, comprising a narrative of a tour performed in the summer of1820. New Haven.Mott, Mildred.1938. The relation of historic Indian tribes to archeological manifestations inIowa. Iowa Journ. Hist, and Politics, vol. 36, No. 3, pp. 227-314.MUDGE, B. F.1896. Traces of the moundbuilders in Kansas. Trans. Kansas Acad. Sci.,vol. 2 (for 1873), pp. 69-71.MuLLOY, William.1942. The Hagen site: a prehistoric village on the Lower Yellowstone.Univ. Montana Publ. in the Soc. Sci., No. 1. Missoula.Mter, W. E.1922. Archeological fieldwork in South Dakota and Missouri. Explor. andField-Work of the Smithsonian Inst, in 1921, pp. 117-125. Smith-sonian Misc. Coll., vol. 72, No. 15.Nasatir, a. p., Editor.1952. Before Lewis and Clark: Documents illustrating the history of theMissouri 1785-1804. 2 vols. St. Louis Hist. Doc. Found. St. Louis. Wbdbl] an ESTTRODUCTION to KANSAS ARCHEOLOGY 653Neweix, H. p., and Kbiegee, A. D.1949. The George C. Davis site, Cherokee County, Texas. Mem. No. 5, Soc.Amer. Archeol. Menasha, Wis.Newlon, W, S.1894. Archeology of the lower Neosho valley. Areheologist, vol. 2, No. 6,pp. 192-193. Waterloo, Ind.Orb, Kenneth G.1946. The areheologieal situation at Spiro, Oklahoma ; a preliminary reportAmer. Antiq., vol. 11, No. 4, pp. 228-256.Over, W. H.1940. Stone mauls of the northwest Plains region. Amer. Antiq., vol. 5, No.4, pp. 336-337.Parker, H. N.1911. Quality of the water supplies of Kansas. U. S. Geol. Survey, Water-Supply Paper 273. Washington.Parker, J. D.1887. Mounds in Davis County. Trans., Kana. Acad. Sci., vol. 10, pp. 72-73.Pasco, Geo. L., and Ritzenthaler, Robt. E.1949. Copper disks in Wisconsin, Wisconsin Archeol., n. s., vol. 30, No. 3,pp. 51-52.Patrick, G. E.1906. The Great Spirit Spring. Trans. Kansas Acad. Sci., vol. 7, pp. 22-26.Pauulin, C. O.1932. Atlas of the historical geography of the United States. Carnegie Inst.Wash, and Amer. Geogr. Soc. of New York.Pike, Zebulon M.1810. An account of expeditions to the sources of the Mississippi, andthrough the western parts of Louisiana, to the sources of the Arkan-sas, Kans, La Platte, and Pierre Jaun Rivers ; performed by orderof the Government of the United States during the years 1805, 1806,and 1807. And a tour through the interior parts of New Spain,when conducted through these provinces. Philadelphia.Plank, Pryob.1910. Old village sites in Doniphan County, Kansas. Archeol. Bull., vol. 1,No. 4, pp. 97-99.Price, Raymond S.MS. Early ceramic period sites in northeastern Nebraska. M. A. thesis,Univ. Nebraska, Dept. Anthrop., Lincoln. 1956.PUTNAM, F. W.1880. Report of the curator. 13th Ann. Rep. Peabody Mus. Amer.Archaeol. and Ethnol. (for 1879), Harvard Univ., vol. 2, No. 4, pp.709-755.Remsburq, Geo. J.1893. Two mounds near Atchison. Areheologist, vol. 1, No. 3.1894. Indian graves. Areheologist, vol. 2, No. 2, p. 61.1909. Areheologieal notes on northeastern Kansas. Archeol. Bull., vol. 1,No. 1, pp. 1-3.1910. Notes from my areheologieal field book. Archeol. Bull., vol. 1, No. 2,pp. 42-46.1911. Recent examinations in Salt Creek valley in eastern Kansas.Archeol. Bull., vol. 2, No. 3, pp. 66-67.1912 a. Random areheologieal notes. Archeol. Bull., vol. 3, No. 1, pp. 23-24. 654 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174Remsbubg, Gbx). J.?Continued1912 b. Abstracts of correspondence relative to Kansas archeology.Archeol. Bull., vol, 3, No. 4, pp. 121-122.1915 a. Waconda Spring. Archeol. Bull., vol. 6, No. 1, pp. 8-10.1915 b. Notes relative to Kansas archeology. Archeol. Bull., vol. 6, No. 6,pp. 11-12.1919. An old Kansas Indian town on the Missouri. Plymouth, Iowa(privately printed). Also printed in Gray's Doniphan Countyhistory, by P. L. Gray, 1905, pp. 148-155, Bendena, Kans.Renaud, E. B.1936. Pictographs and petroglyphs of the high western plains. 8th Rep.Archeol. Surv. of the high western plains. Univ. Denver.RiCHEY, W. E.1904 a. Report of committee on mounds and village sites. Trans. KansasState Hist. Soc., 1903-4, vol. 8, pp. 135-136.1904 b. Early Spanish explorations and Indian implements in Kansas.Trans. Kansas State Hist. Soc, vol. 8, 1903-4, pp. 152-168.ROBEBTS, F. H. H., Jb.1940. Developments in the problem of the North American Paleo-Indian.Smithsonian Misc. Coll., vol. 100, pp. 51-116.RoBEBTS, Helen H.1936. Musical areas in aboriginal North America. Yale Univ. Publ.Anthrop., No. 12, 41 pp.Robinson, A. P.1881. In Abstracts of the Smithsonian correspondence relative to aboriginalremains in the United States, p. 446. Ann. Rep. Smithsonian Inst,for 1880.Roe, F. G.1951. The North American buffalo. Toronto.ROMEB, A. S.1933. Pleistocene vertebrates and their bearing on the problem of humanantiquity in North America. In The American aborigines, theirorigin and antiquity. 5th Pacific Sci. Congr., Toronto.Royce, C. C.1899. Indian land cessions in the United States. 18th Ann. Rep. Bur. Amer.Ethnol., pt. 2, pp. 521-964.Rusco, Maby Kiehl.MS. The White Rock Aspect. M. A. thesis, Dept. Anthrop., Univ. Nebraska,Lincoln. 1955.Sayles, E. B.1935. An archeological survey of Texas. Medallion Pap., No. 27. GilaPueblo, Globe, Arizona.SOHMITT, Kael.1950. The Lee Site, Gv3, of Garvin County, Oklahoma. Bull., Texas Archeol.and Paleont. Soc, vol. 21, pp. 69-89. Lubbock, Texas.ScHMiTT, Karl, and Schmitt, Iva Osanai.1952. Wichita Kinship, past and present. University Book Exchange, Nor-man, Okla.Schmitt, K., and Toldan, R.1963. The Brown site, Gd-1, Grady County, Oklahoma. Bull. Texas Archeol.Soc, vol. 24, pp. 141-176. Wedhl] an introduction to KANSAS ARCHEOLOGY 655SCHOEWE, W. H.1949. The geography of Kansas, pt. 2?physical geography. Trans., Kan-sas Acad. Sci., vol. 52, No. 3, pp. 261-333.1951. The geography of Kansas, pt. 3?hydrogeography. Trans., KansasAcad. Scieuce, vol. 54, No. 3, pp. 263-329.SCHOLES, F. V.1930. The Supply Service of the New Mexican Missions in the seventeenthcentury. Pt. Ill, 1663-1680. New Mexico Hist. Rev., vol. 5, No.4, pp. 386-404.ScHOLEs, France V., and Mera, H. P.1940. Some aspects of the Jumano problem. Carnegie Inst. Washington,Publ. 523, pp. 265-299.SoHULTz, C. B., and Eiseley, L. C.1935. Paleontological evidence for the antiquity of the Scottsbluff bisonquarry and its associated artifacts. Amer. Anthrop., n. s., vol. 37,No. 2, pp. 306-319.ScHULTZ, Flotd, and Spaulding, Albert C.1948. A Hopewellian burial site in the Lower Republican Valley, Kansas.Amer. Antiq., vol. 13, No. 4, pp. 306-313.SCXXTT, C. A.1938. Kansas trees and their uses. Rep. Kansas State Board Agric, vol.47, No. 186-A, pt. 1, pp. 15-147.Secoy, Frank R.1951. The identity of the "Paduca" ; an ethnohistorical analysis. Amer.Anthrop., vol. 53, No. 4, pt. 1, pp. 525-542; October-December.Sellards, E. H.1940. Early man in America : Index to localities, and selected bibliography.Bull. Geol. Soc. Amer., vol. 51, pp. 373-432.1952. Early man in America. Austin, Tex.Serviss, E. F.1883. Mounds near Edwardsville, Wyandotte County, Kansas. Ann. Rep.Smithsonian Inst, for 1881, p. 528.Shetrone, H. O.1941. The Mound-builders. New York and London.Shimkin, D. B.1940. Shoshone-Comanche origins and migrations. Proc. 6th Pacific Sci.Congr., vol. 4, pp. 17-25.Shippee, J. M.1948. Nebo Hill, a lithic complex in western Missouri. Amer. Antiq., vol.14, No. 1, pp. 29-32.1953. A Folsom fluted point from Marshall County, Kansas. Plains Archeol.Conf. News Letter, vol. 5, No. 4, p. 54.Sibley, George C.1922. "From a journal to the Pawnee and Kansas Villages, undertaken byan officer of the Factory on the Missouri." Louisiana Gazette, May16, 1912; reprinted in Nebraska Hist. Publ., vol. 20, pp. 5-11,Lincoln.1927. Extracts from the diary of Major Sibley. Chronicles of Oklahoma,vol. 5, pp. 196-218. Oklahoma City.Skinner, A.1921. Material culture of the Menomini. Mus. Amer. Indian, Heye Found.,Indian Notes and Mouogr., vol. 20.484172?59 43 656 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174Skinneb, Mokbis F., and Kaisen, O. C.1947. The fossil bison of Alaska and preliminary revision of the genus.Amer. Mus. Nat. Hist. Bull., vol. 89.Smith, Caelyle S.1949 a. Archeological investigations in Ellsworth and Rice Counties, Kan-sas. Amer. Antiq., vol. 14, No. 4, pt. 1, pp. 292-300. April.1949 b. Fieldwork in Kansas, 1949. Plains Archeol. Conf. News Letter,vol. 2, No. 4, pp. 5-6.1950. European trade material from the Kansas Monument site. PlainsArcheol. Conf. News Letter, vol. 3, No. 2, pp. 2-9.Smith, Hale G.1951. The Crable site, Fulton County, 111. Anthrop. Pap., No. 7, Mus. An-throp., Univ. Michigan,Smith, H. T. U.1938. Geomorphic evidence relating to the antiquity of man in north cen-tral Kansas. (Abstract) Geol. Soc. Amer. Bull., vol. 49, p. 1901.1940. Geologic studies in southwestern Kansas. Kansas Geol. Surv.,Bull. 34.Snow, F. H.1872. Catalogue of the birds of Kansas. Trans., Kans. Acad. Sci., vol. 1,pp. 21-29.SoLECKi, Ralph.1953 a. A Plainview point found in Marshall County, Kansas. PlainsArcheol. Conf., News Letter, vol. 5, No. 4, pp. 52-53.1953 b. Appraisal of the archeological and paleontological resources ofthe Tuttle Creek Reservoir, Marshall, Pottawatomie, and RileyCounties, Kansas. Missouri Basin Project, Lincoln. (Mimeogr.)Spauldinq, Albert C.1949. The Middle Woodland period in the Central Plains. Proc. FifthPlains Conf. for Archeol., Note Book No. 1, Lab. Ajithrop., Univ.Nebraska, pp. 105-111.1956. The Arzberger site, Hughes County, South Dakota. Mus, Anthrop.,Univ. Michigan, Occas. Contrib. No. 16.Steen, Charlie R.1953. Two early historic sites on the southern Plains. Bull. TexasArcheol. Soc, vol. 24, pp. 177-188. Abilene.Stephenson, Robert L.1954. Taxonomy and chronology in the Central Plains?Middle MissouriRiver area. Plains Anthrop., No. 1, pp. 15-21. Lincoln.Sterns, F. H.1914. Ancient lodge sites on the Missouri in Nebraska. Amer. Anthrop.,n. s., vol. 16, No. 1, pp. 13.5-137.MS. The archaeology of eastern Nebraska, with special reference to theculture of the rectangular earth lodges. 2 vols. (Unpublished Ph. D.thesis. Harvard University, 1915,)Steward, J. H.1937. Ancient caves of the Great Salt Lake region. Bur. Amer. EthnoL,Bull. 116.Stewart, T. D.1943. Skeletal remains from Platte and Clay Counties, Missouri. U. S.Nat. Mus. Bull. 183, pp. 245-273.Stockton, J. B.1883. In Abstracts from anthropological correspondence. Ann. Rep.Smithsonian Inst, for 1881, p. 685. Wedbl] an introduction to KANSAS ARCHEOLOGY 657Stbong, W. D.1933. Plains culture area in light of archeology. Amer. Anthrop., vol. 35,No. 2, pp. 271-287.1935. An introduction to Nebraska archeology. Smithsonian Misc. Coll.,vol. 93, No. 10.1940. From history to prehistory in the northern Great Plains. SmithsonianMisc. Coll., vol. 100, pp. 353-394.Studeb, Floyd V.1934. Texas Panhandle Culture ruin No. 55. Bull. Texas Arch, andPaleont. Soc, vol. 6, pp. 80-96.SuHM, Dee Ann ; Kkieqeb, A. D. ; and Jelks, E. B.1954. An introductory handbook of Texas archeology. Bull. Texas Archeol.Soc, vol. 25. Austin.SwANTON, John R.1952. The Indian tribes of North America. Bur. Amer. Ethnol. Bull. 145.Thobuen, J. B.1927. An important archeological and historical discovery. Chronicles ofOklahoma, vol. 5, No. 4, pp. 407-413.Thomas, A. B.1929. San Carlos : A Comanche Pueblo on the Arkansas River, 1787. Colo-rado Mag., vol. 6, No. 3, pp. 7&-91. May.1932. Forgotten frontiers. A study of the Spanish Indian policy of DonJuan de Bautista de Anza, Governor of New Mexico, 1777-1787.Norman, Okla.1935. After Coronado : Spanish exploration northeast of New Mexico 1696-1727. Univ. Oklahoma Press, Norman.1940. The Plains Indians and New Mexico, 1751-1778. Univ. New MexicoPress, Albuquerque.Thomas, Ctbus.1891. Catalogue of prehistoric works east of the Rocky Mountains. Bur.Ethnol. Bull. 12.1894. Report on the mound explorations of the Bureau of Ethnology. 12thAnn. Rep., Bur. Ethnol., pp. 17-730. Washington.Thornthwaite, C. W.1941. Climate and settlement in the Great Plains. In Climate and Man,U. S. Dept. Agric. Yearbook of Agriculture, 1941, pp. 177-187.Thwaites, Reuben Gold, Editob.1904-5. Original journals of the Lewis and Clark Expedition, 1804-6. 8vols. New York.TowNSEND, John K.1839. Narrative of a journey across the Rocky Mountains to the ColumbiaRiver, and a visit to the Sandwich Islands, Chili, . . . With ascientific appendix. Philadelphia.Teoike, Rudolph C.1955. Anthropological theory and Plains archeology. Bull., Texas Archeol.Soc, vol. 26, pp. 113-143.Tucke:k, Sara Jones.1942. Indian villages of the Illinois Country. Vol. II, Sci. Pap., IllinoisState Museum, Pt. I, Atlas. Springfield, 111.Udden, J. A.1900. An old Indian village. Augustana Library Publ., No. 2, Rock Island. 658 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174Van Royen, Whuem.1937. Prehistoric droughts in the Central Great Plains. Geographical Rev.,vol. 27, No. 4, pp. 637-650.ViLLIEBS D0 TebBAQE, MaRO DE.1925. La decouverte du Missouri et I'histoire du Fort Orleans (1673-1728).Paris.Wallace, E., and Hoebel, E. A.1952. The Comanches. Univ. Oklahoma Press, Norman.Wabben, Gouveeneub K.1859. Memoir to accompany the map of the territory of the United Statesfrom the Mississippi River to the Pacific Ocean, giving a briefaccount of each of the exploring expeditions since A. D. 1800, witha detailed description of the method adopted in compiling thegeneral map. Washington.Watson, Vibqinla.1950. The Optima Focus of the Panhandle Aspect: Description and analy-sis. Bull. Texas Archeol. and Paleont. Soc., vol. 21, pp. 7-68. Lub-bock, Tex.Weaklt, H. B.1940. Tree-rings as a record of precipitation in western Nebraska. Tree-ring Bull., vol. 6, No. 3, pp. 18-19. Tucson.1943. A tree-ring record of precipitation in western Nebraska. Journ. For-estry, vol. 41, No. 11, pp. 816-819.Webb, C. H,, and Dodd, Monboe.1939. Further excavations of the Gahagan mound ; connections with aFlorida culture. Bull. Texas Archeol. and Paleont. Soc., vol. 11,pp. 92-128. Abilene.Wedel, Waldo R.1934. Preliminary notes on the archeology of Medicine Valley in south-western Nebraska. Nebr. Hist. Mag., vol. 14, No. 3, pp. 144-166.1935. Reports on field work by the archeolojacal survey of the NebraskaState Historical Society, 1934. Nebraska Hist. Mag., vol. 15, No. 3,pp. 132-255.1936. An introduction to Pawnee archeology. Bur. Amer. Ethnol. Bull.112.1938 a. Inaugurating an archeological survey in Kansas. Expl. and Field-work of the Smithsonian Inst, in 1937, pp. 103-106.1938 b. Some problems and prospects in Kansas prehistory. Kansas Hist.Quart., vol. 7, No. 2, pp. 115-132.1938 c. The direct-historical approach in Pawnee archeology. SmithsonianMisc. Coll., vol. 97, No. 7.1940 a. Culture sequence in the central Great Plains. Smithsonian Misc.Coll., vol. 100, pp. 291-352.1940 b. Archeological explorations in western Kansas. Explor. and Field-work of the Smithsonian Inst, in 1939, pp. 83-86.1941. a. Environment and native subsistence economies in the Central GreatPlains. Smithsonian Misc. Coll., vol. 101, No. 3.1941 b. In search of Coronado's province of Quivira. Explor. and Field-work of the Smithsonian Inst, in 1940, pp. 71-74.1942. Archeological remains in central Kansas and their possible bearingon the location of Quivira. Smithsonian Misc. Coll., vol. 101, No. 7. Wedei,] an introduction TO KANSAS ARCHEOLOGY 659Wedel, Waldo R.?Continued1943. Archeological investigations in Platte and Clay Counties, Mo. U. S.Nat. Mus., Bull. 183.1946. The Kansa Indians. Trans. Kansas Acad. Sci., vol. 49, No. 1, pp.1-35. June.1947 a. Culture chronology in the Central Great Plains. Amer. Antiq.,vol. 12, No. 3, pp. 148-156. January.1947 1>. Prehistory and environment in the Central Great Plains. Trans.,Kansas Acad. Sci., vol. 50, No. 1, pp. 1-18. June.1049. Some Central Plains sherd types from Kansas. Proc. Fifth PlainsConf. for Archeology, Note Book No. 1, Lab. Anthrop., Univ. Ne-braska, pp. 86-90. Lincoln.1950. Notes on some Plains-Southwestern contacts in light of archeology.In For the Dean, Essays in Anthropology in Honor of Byron Cum-mings on his 89th Birthday, Sept. 20, 1950, pp. 99-116. HohokamMuseums Assn., Tucson, and Southwestern Monuments Assn.,Santa Fe.1953 a. Some Aspects of human ecology in the Central Plains. Amer.Anthrop., vol. 55, No. 4, pp. 499-514. October.1953 b. Prehistory and the Missouri Valley development program : Stim-mary report on the Missouri River Basin archeological surveyin 1948. Bur. Amer. Ethnol. Bull. 154, Riv. Has. Surv. Pap., No. 1,pp. 1-59.1955. Archeological materials from the vicinity of Mobridge, South Da-kota. Bur. Amer. Ethnol. Bull. 157, Anthrop. Pap. No. 45, pp. 69-188.1956. Changing settlement patterns in the Great Plains. In Prehistoricsettlement i)atterns in the New World, G. R. Willey (ed.). VikingFund Publ. in Anthrop., No. 23, pp. 81-92. Wenner-Gren Found,for Anthrop. Research, Inc.1957. The central North American grassland: Man-made or natural? InStudies in Human Ecology, pp. .39-69. Soc. Sci. Monogr., Ill, PanAmerican Union, Washington.Wedel, Waldo R., and Geiffenhagen, G. B.1954. An English balsam among the Dakota aborigines. Amer. Journ.Pharmacy, vol. 126, No. 12, pp. 409-415. PhUadelphia.Wedel, Waldo R., and Hnx, A. T.1942. Scored bone artifacts of the Central Great Plains. Proc. U. S.Nat. Mus., vol. 92, No. 3141, pp. 91-100.Wedel, Waldo R., and Kivett, Maevin F.1956. Additional data on the Woodruff ossuary, Kansas. Amer. Antiq.,vol. 21, No. 4, pp. 414-416. Salt Lake City.Wedel, Waldo R., and Tbowbeidge, Habey M.1940. A prehistoric roulette from Wyandotte County, Kansas. Proc. U. S.Nat. Mus., vol. 88, pp. 581-586.WEa^TFisH, Gene,1930. Prehistoric North American basketry techniques and modem dis-tributions. Amer. Anthrop., vol. 32, No. 3, pt. 1, pp. 454-495.Wendoef, Feed.1953. Salvage archeology in the Chama Valley, New Mexico. School ofAmerican Research, Monogr. No. 17. Santa Fe, New Mei. 660 BUREAU OP AMERICAN ETHNOLOGY [Bdll.174West, G. A.1934. Tobacco, pipes, and smoking customs of the American Indians. Bull.Public Mus. City of Milwauliee, vol. 17, pts. 1 and 2, pp. 1-994.Westebvelt, Frances A.1924. The final century of the wampum industry in Bergen County, NewJersey. Bergen County Hist. Soc, Hackensack, N. J.Whitefobd, G. L.1937. Prehistoric Indian excavations in Saline County, Kansas. Salina,Kansas. (Privately printed.)1941. Indian archeology in Saline County, Kansas. Salina, Kansas.(Privately printed.)WiLTORD, Lloyd A.1955. A revised classification of the prehistoric cultures of Minnesota.Amer. Antiq., vol. 21, No. 2, pp. 130-142.Wnx, Geoege F., and Hecker, Thad C.1944. The Upper Missouri River Valley aboriginal culture in North Dakota.North Dakota Hist. Quart., vol. 11, Nos. 1 and 2, pp. 5-126. Jan-uary-April. Bismarck.Will, George F., and Spinden, H. J.1906. The Mandans. Harvard Univ., Peabody Mus. Amer. ArchaeoL andEthnol. Pap., vol. 3, No. 4.Williams, A. C.1877. In Annual Report of the Commissioner of Indian Affairs for theyear 1877. Washington.WiLLISTON, S. W.1879. Indian figures in western Kansas. Kansas City Rev. of Sci., vol. 3,p. 16.1899. Some prehistoric ruins in Scott County, Kansas. Kansas Univ.Quart., Ser. B, vol. 7, No. 4, pp. 109-114.1902 a. An arrowhead found with bones of Bison occidentalis Lucas,in western Kansas. Amer. GeoL, vol. 30, pp. 313-315.1902. b. A fossil man from Kansas. Science, n. s., vol. 16, pp. 195-196.1905 a. On the occurrence of an arrow-head with bones of an extinctbison. Intern. Cong. Americanists, Proe. 13th sess., pp. 335-337.1905 b. On the Lansing Man. Intern. Cong. Americanists, Proc. 13thsess., pp. 85-89.WiLLISTON, S. W., and Martin, H. T.1900. Some pueblo ruins in Scott County, Kansas. Kansas Hist. Coll.,vol. 6, pp. 124-130.WlNCHELL, N. H.1903. The Pleistocene geology of the Concannon farm, near Lansing, Kan-sas. Amer. Geol., vol. 31, pp. 263-308.1913. A consideration of the paleoliths of Kansas. No. 1, The weatheringof aboriginal stone artifacts. Minnesota Hist. Soc. Coll., vol. 16,pt. 1.WiNSHip, George P.1896. The Coronado expedition, 1540-1542. 14th Ann. Rep. Bur. Ethnol.,pp. 329-613.WiSSLER, C.1912. North American Indians of the Plains. Amer. Mus. Nat. Hist.,Handbook Ser. No. 1. New York. Wedkl] an introduction TO KANSAS ARCHEOLOGY 661Withers, Arnold M.1954. University of Denver archeological fieldwork, 1952-1953. Southwest-ern Lore, vol. 19, No. 4, pp. 1-3.WiTTE, A. H.1947. Certain archeological notes on the High Plains of Texas. Bull.,Texas Archeol. and Paleont. Soc, vol. 18, pp. 76-82.Wood, W. Raymond.1956. Settlement patterns of the Redbird focus. Plains Anthrop., No. 7,pp. 3-9. Lincoln.MS. The Redbird focus. M. A. thesis, Dept. Anthrop., Univ. Nebraska,Lincoln. 1956.WORMINGTON, H. M.1957. Ancient man in North America. 4th ed. Denver.Wright, Muriel H.1951. A guide to the Indian tribes of Oklahoma. Univ. Oklahoma Press,Norman.Zimmerman, M. E.1918. The ground-house Indians and stone-cist grave builders of Kansasand Nebraska. Coll. Kansas State Hist. Soc, vol. 14, pp. 471-487. APPENDIXTHE FAUNAL LISTS FROM KANSASIn table 19, I have summarized the occurrence of identified mam-mal, bird, and molluscan remains, chiefly unworked refuse material,from sites discussed in the present report. The lists must be viewedwith caution; they are suggestive rather than definitive. For onething, the extent of excavation at the several sites varied greatly, sothat the base from which the sample is drawn is very uneven. Inno instance can it be said that the site represented was more thansampled; and I am certain that more extended excavation wouldappreciably lengthen the faunal inventories. Moreover, as any stu-dent of Plains archeology will immediately realize, the samplingis decidedly one sided with regard to the time factor. Only inLane County did we obtain identifiable famial materials from sitesor levels that can be termed prehistoric ; the Nebraska Culture housesites at Doniphan and the house sites and caches at the Grifing site,all of pre-White age, yielded virtually no bone or shell refuse andnone that could be identified.Despite these deficiencies, it seems to me that the inventories areof some interest. In the first place, the sites from which the ma-terials were collected are scattered throughout the State in sucha way as to sample various sections with dissimilar environmentalcharacteristics, varying from the deciduous forests of the east to theshort-grass country of the west. These environmental variationsinclude, of course, difference in the biotic assemblages. Moreover,with one exception, the sites represented antedate the year 1800, andthus may throw a little light on faunal distributions and popula-tions prior to the period of white settlement and of large-scale agri-cultural and commercial development. If it can be reasonably as-sumed that the Indians, for the most part not yet supplied withhorses, drew most lieavily upon the fauna near their villages, thenthe refuse from their hunting and food-gathering activities shouldreflect the general nature of that fauna at the time of the Indianoccupation. Here again any inferences drawn are subject to thelimitations of our sampling, as well as to the limitations on ourknowledge of the exact nature of hunting methods and butcheringtechniques during the long period of time preceding the comingof the white man and the horse.662 Wbdbl] an introduction to KANSAS ARCHEOLOGY 663The sites from which identifiable faunal materials were taken havebeen grouped geographically in table 19. They include three eachfrom the northeastern (Doniphan and Pottawatomie Counties), cen-tral (Rice County), south central (Cowley County), and western(Lane and Scott Counties) sections of the State. The northeasternsites are situated in or near heavily timbered river valleys with tallgrass prairies nearby. The south-central sites are near the junctionof the Walnut and Arkansas Rivers, both with well-timbered valleys,and near the western edge of the oak forest but with easy access totreeless plains in the southern Flint Hills and across the Arkansas tothe west. Sites of the central group are in proximity to moderatelyor sparsely timbered small creek valleys in the mixed prairie and shortgrass country. The western sites lie well beyond the deciduous forestarea of the State, on sparsely timbered creeks hemmed in closely byshort grass plains. Some 300 miles separate the Fanning and Doni-phan sites of the northeast group from those of the western group ; the south central sites are more than 200 miles distant from DoniphanCounty.With reference to mammal remains, only two species?^bison andwhite-tailed deer?were represented at all sites. Their relative abun-dance from east to west, however, varies inversely. Bison bonestend to occur with increasing frequency as the short grass plains areapproached, whereas deer bones diminish westward as the forest andbrush cover dwindles. As regards these two forms, the indicatedeast-to-west changes in frequency of occurrence undoubtedly rest inlarge part, though not exclusively so, on ecological factors. In pass-ing, it may be noted that the 1937 excavations of the U. S. NationalMuseum at the Renner site in Platte County, Mo., situated in theMissouri valley some 40 miles downriver from Doniphan County,Kans., yielded a vastly larger number of deer bones than of bison(Wedel, 1943, p. 27). The unexpectedly high incidence of deerremains at the Scott County site (14SC1) is very likely due to anunusually favorable ecological situation resulting from the conver-gence witliin a few miles of the archeological zone of several canyonsand deep ravines, originally no doubt with thick brush cover and ad-jacent to excellent water supplies.Bones of the dog occur with less frequency ; but their absence fromcertain sites undoubtedly reflects inadequate sampling rather thanactual lack of the animal. This is surely the case for the historicKansa site (14P024) near Manhattan, where such contemporai^observers as Sibley reported the animal in 1811. Absence of dogbones at the Elliott site, despite their presence at two immediatelyadjacent sites that are apparently chronologically equivalent and cul-turally similar, can also be explained as inadequate sampling. The 664 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174 ?H 1 iM IN ii-i i>o 1 ie IPS i^w (ZYl.n) JBlIBAi. niB}J8onn 90U3ra3A0J w r: w O O CJ4 . - . C C O ? 5 SC^.g"c: feo ^ ^ ^ =5 ? i: sS ? i; f= ? 5 s-5E.S2 pQ goli V oOpqfi Wedbl] an introduction TO KANSAS ARCHEOLOGY 665 xxxx XX XXX XX : :xXXX xxxx XX xxxxxx XXXXX IX XXXX XX XXX XXX lis ^tt) Sxi t:^ ^ s s a.o^-sst',,^ &?. jHS "SO?2ScS52?-?.SgRial's K5 55S g ?-^ -5 ? i? ^C io:S-glgg 152^^2 Is^iS^i -3T3 k fc22 ? BXi^j^Stir'^M t- P P P !> bj:j ttis o 53 *- S o 7^ t* !2Se-S O '^ S < 1^ 1^J- aq tea, a; a; ?550 tf"S S c9 ?=3' F ^ ! :m~ i'ShJIS^- - ? - ct, ""i; o-oo-E^ ft, a, bqo 1^ '^j &2oo "3 ? s s1113-2 "^f-lft,ft, SoilnilssgssSo_e s s s g'Spa ?&& o c a a sa, 666 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174dog bones await full treatment by a qualified observer ; but it may benoted that among the remains large animals are indicated for practi-cally all of the central, south central, and western sites. This sug-gests the possibility of large draft animals perhaps developed in lateprehistoric times for purposes that were later accomplished by thehorse, that is, the transport of baggage and other burdens.Otherwise, the assemblages indicated in table 19 conform prettymuch to what might be expected, except that there are a number ofabsences most readily explained in terms of incomplete sampling.The black bear, puma, beaver, wildcat, and woodchuck are recordedonly for the northeastern sites. They reflect the presence of heavytimber and brush cover ; all are woodland species, though their distri-bution in the State was undoubtedly much wider than our scantyevidence indicates. Antelope, wolf, badger, and prairie dog werefound by us only in the central and western sites ; the coyote, kit fox,and wood rat only in the western sites. Elk bones were recoveredfrom the northeastern, central, and south-central sites, at all of whichsome open woods and adjacent grasslands were to be found. Raccoonbones occurred in sites in all sections. The occurrence of the opossumis of interest; it was represented in two sites at the edge of the Mis-souri valley in the extreme northeast. Both occurrences are fromIndian villages of the early contact period (pre-1750) ; and they sug-gest that the animal may have been just beginning its spread from theeastern woodlands into the trans-Missouri plains.Bird bones occurred so infrequently and scatteringly as to be of littleusefulness. It is of some interest, however, to note that the wildturkey is represented only in the northeastern and south-central sites ? in other words, in and at the margin of the oak forests of easternKansas. Actually, the turkey ranged much further westward alongthe timbered stream valleys; and it would be rather surprising iffuture excavation failed to disclose its remains at sites in our centralgroup and to the north.So far as this evidence goes, it suggests that environmental condi-tions during the period represented by our sites were essentially likethose of the recent past, say after 1800. The earlier Indian popula-tions with which our investigations were primarily concerned gen-erally pursued a semihorticultural subsistence economy, supplement-ing their domestic crops with such animal and other foods as they wereable to get in the vicinity of their settlements. In the eastern part ofthe State, the abundant fauna of the forested valleys was constantlydrawn upon; and this was augmented from time to time by bison,killed either in the nearby prairies or in course of longer huntingexcursions to the west. The residents of the central, south central, Wedel] an introduction TO KANSAS ARCHEOLOGY 667and western sections presumably drew much more regularly on thebison, in and near whose range their settlements were located ; and tothis were added such items as deer and other smaller mammals as thelimited timber of the stream valleys offered.It has already been indicated that only the Lane County data intable 19 relate to pre-White sites. They are of added interest becausethey suggest a phenomenon that has been noted in other Central Plainslocalities. As the table shows, the later occupation here, attributedto the Upper Republican culture, shows a much higher proportion ofbison to other animal bone than does the earlier Woodland occupation.Conversely, the Woodland level yielded much more antelope, and moredeer, prairie dog, and rabbit than did the Upper Eepublican horizon,despite the fact that our tests included an appreciably smaller pro-portion of the Woodland zone than of the Upper Republican. Thedata seem to hint at a greater reliance on smaller mammals by theWoodland peoples and, conversely, a greater emphasis on bison by thelater Upper Republican peoples. Similar differences between Wood-land faunal assemblages and those from later archeological horizonshave been noted in Nebraska, especially on Medicine Creek in FrontierCounty and at the Kelso site in Hooker County (Kivett, 1949, p. 283 ; 1952, p. 39). The recurring indications in widely separate localitiesthat small mammals, birds, and other lesser animals are relativelymore plentiful than bison in Woodland sites, and that this situationis reversed in later manifestations in the same localities, poses aninteresting problem.The solution to this problem is not immediately apparent. It iseasy to suggest that differences in hunting methods or butchering tech-niques may be responsible, but this is not enough. In what way didthey differ? At the Pottorff site in Lane County, both Upper Re-publican and Woodland peoples lived on the same creek terrace,separated only by an interval of time whose length is unknown ; bothhad equal access to thinly wooded creek bottoms and to nearby grass-lands ; both hunted on foot ; both transported their kill on their ownbacks or on dogs. Here, as at the Nebraska localities specified, onlyfoot hunters are involved; and I see no reason to suppose that thelater peoples, whether of Upper Republican or of Dismal River cul-ture, would have been any more likely than the earlier Woodlandhunters to carry back to camp any appreciable quantities of uselessbison bone. Why, then, the indicated differences in the animal inven-tory as shown by the refuse bone ?Several possible explanations may be suggested. One which im-presses me as rather remote is that the bison were for some reason lessplentiful in Woodland times than later, and thus furnished a much 668 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174less dependable food source. Another is that the western Woodlandsites in which most of our work has been carried on to date representmainly winter camps, situated in sheltered stream valleys, where bisonwere less easily obtained than were deer and other small game animalsand birds associated with brush or tree cover. Again, it is possiblethe Woodland peoples actually preferred small game to bison ; but Istrongly doubt that any Indian group would live long in proximity tothe bison herds without realizing, and taking advantage of the fact,that here was a vastly more rewarding quarry for the hunter than adeer, an antelope, or a rabbit. On the other hand, the latter could bemore easily secured by individual effort than could the bison, which aremost effectively taken by organized communal drives. Archeologicalevidence suggests that the Woodland populations, especially in thewestern and possibly earlier variants, were small, scattered, and un-settled ; and it may be that for such groups effective large-scale bisondrives could not be mounted as successfully or as frequently as amongthe later and more populous groups.My convictions on this matter are not particularly strong at present.If, however, future work bears out the indicated faunal differenceswe have just discussed, I suggest that the likeliest explanation maylie in the sparse and unstable nature of the Woodland occupations ascompared to later semihorticultural manifestations of the region ; thatthe small communities of this period probably found it less profitableto pursue the bison ; and that their bone refuse reflects a subsistenceeconomy based primarily on creek-bottom hunting and gathering,and perhaps only slightly above the foraging level. DESCRIPTION OF THE SKELETAL REMAINSFROM DONIPHAN AND SCOTT COUNTIES,KANSAS By T. D. StewaktTwo lots of material will be described: (1) The remains of 10burials collected by Dr. Wedel in the summer of 1937 at the Doniphansite, and (2) the remains of 3 burials collected by Dr. Wedel in thesummer of 1939 at the Young site in Scott County. The former datesfrom historic times and probably represents the Kansa tribe ; the lat-ter seems to represent the prehistoric people of the Early Woodlandculture. As will be shown, the skulls of these two lots offer a decidedcontrast in shape, those of the Kansa being artificially deformed andhence quite round, and those of the Woodland people being extremelylong and narrow. I. DONIPHAN SITEThe skeletal remains recovered at this site consist of 7 subadultsand 5 adults (3 males, 1 female, and 1 female?). The skeletons arewell preserved but vary considerably in completeness, owing appar-ently to the action of burrowing animals, to farming activities, and tothe curiosity of man in recent times. Detailed measurements of theadults are assembled in tables 1 to 4. Otherwise each specimen isconsidered separately in the order of the burial numbers, and thegroup is discussed on p. 679.Burial No. 1 {USNM No. 378811).?ChWdi about 8 years of age.Lacks mainly the leg bones, left ulna, pubes, ischia, hands and feet, afew cervical vertebrae, and most of the lumbar and sacral vertebrae.The skull shows asymmetrical occipital compression (more on leftthan right) with a resultant cranial index of 98.7. Eruption of thepermanent teeth has advanced to the stage where the first molars andall of the incisors have reached the occlusal plane, but the secondmolars and the canines have not yet emerged from the alveoli. Thecusps of the deciduous first molars are worn down so that the dentinis exposed ; those of the deciduous second molars are not worn downso far.The base of the skull and the bones of the upper arms, shoulders,and chest are stained green by copper salts.BuHal No. 2 {USNM No. 378812).?ChiU about 8 years of age.Skull damaged. Skeleton lacks mainly the radii, left ulna, fibulae,pubes, ischia, hand and foot bones, and several vertebrae. 669 670 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174The skull shows asymmetrical occipital compression (more on leftthan right) with a resultant cranial index close to 91.5, Tooth erup-tion is in the same stage as in No. 1. However, wear of the deciduousmolars is a little more advanced.Some of the long bones have been gnawed by animals. In addi-tion, the anterior midshafts of the tibiae, especially the left, show anearly stage of periostitis with slight diffuse swelling. Another sucharea occurs in the right ulna along the whole posterior surface of theshaft. The nature of these inflammatory processes is unknown.Burial No. 3 [USNM No. 378813).?A^LvM female ( ?). In manyrespects this is an intermediate case from the standpoint of sexidentification. Size suggests a female, but some of the sex charac-ters of the pelvis suggest a male. All of the bones are present exceptthe right radius, right clavicle, part of the sacrum, and a few ribs,vertebrae, hand and foot bones. Most of the bones present have beengnawed by animals, but particularly the borders of the nose andorbits, the coronoid processes of the mandible, and certain prominentridges on the long bones. The occiput of the skull has been slightlymore compressed on the right than on the left (pi. 94).In considering age, the symphyseal surface of the pubis appears tohave begun to break down after having reached a complete plateauwith rim. Probably this means an age of 40-1- years. In keepingwith this age is the advanced ectocranial closure of the main vaultsutures, including the inferior portion of the masto-occipital suture.Also, the molar teeth were being lost at the time of death and theremaining teeth are moderately worn. On the other hand, osteophy-tosis (lipping of the borders of the vertebral bodies) does not exceed2+ (on a 4+ scale).A pathological process appears on the anterior border of the lefttibia at midshaft and consists of a small area of swelling and alteredsurface texture. Probably this is healed periostitis following trauma.Among the anomalies present are the following: A small para-occipital process on the left side of the foramen magnum; a smallthird trochanter on each femur. It should be noted also that thehumeri lack septal apertures.Burial No. J^ (USNM No. 3788U) .?Adult male. Consists mainlyof the left humerus, right ulna, femora, tibiae, innominates, sacrum,13 vertebrae, and some small bones. Animals have gnawed thezygomatic processes, right orbit, and the shafts of the left humerus,right ulna, right tibia, and right fibula. The occiput of the skullhas been slightly more compressed on the right than on the left(pi. 95).The symphyseal face of the pubis has reached a plateau, but therim is incompletely formed. This stage probably represents an agearound 35 years. Ectocranial closure of the vault sutures, which Wbdel] an introduction TO KANSAS ARCHEOLOGY 671is advanced in the region of bregma, but nowhere else, supports thisdetermination. Also, tooth wear is only second degree, and vertebralosteophytosis is limited to three lumbar segments (2+ ) . No pathological changes were noted. Large third trochantersare present on the femora. The himieri lack septal apertures.Burial No. 5 {USNM No. 378815).?KdiVLlt male. The lower jawis present but not the skull. Although some of the postcranial bonesare damaged, very few are missing, namely, the left clavicle andfirst two cervical vertebrae. Green stains appear on the forearm andpelvic bones.An age of 30-35 years is indicated by the symphyseal surface ofthe pubis which has reached the plateau stage but retains traces ofbillowing and still lacks a rim. Other indicators of age are : Second-degree wear of the teeth of the mandible ; beginning ( 1+ ) vertebralosteophytosis limited to two points in the spine.No pathological changes were noted. There is an extra malformedright lower incisor between and behind Ii and Ij. A moderate-sizedthird trochanter is present on the left femur; a small one is presenton the right. The humeri lack septal apertures.BuHal No. 6 (USNM No. 378816) .?Child. Partial skull only (noface) . In size tliis specimen is comparable to No. 2, hence probablyit is about the same age. There is no asymmetry of the vault, andthe cranial index is 93.3.Burial No. 7 {USNM No. 378817).?Chim about 8 years of age.The lower jaw is present but not the skull. The postcranial skeletonlacks the right radius, clavicles, left femur, left tibia, fibulae, leftpubis, left ischium, some ribs, some vertebrae, and the hand andfoot bones. All of tlie bones except the vertebrae are stained green.Some of the long bones have been gnawed by animals.The age is shown by the eruption of the permanent lower first mo-lars and incisors which have reached the occlusal plane. All othermandibular permanent teeth are still deep in the alveoli.Burial No. 8 {USNM No. 378818).?Male between 15 and 20 yearsof age. The specimen consists of a few pieces of skull, the lowerjaw, 5 incomplete long bones, damaged innominates and sacrum, andtwo lumbar vertebrae. Unfortunately the parts most critical forageing are not present. However, it appears that the elbow and hipepiphyses are united, the crest of the ilium is ununited. Si and S2vertebrae are incompletely united, and the epiphyses of the lumbarcentra are incompletely united. Also, the third molars are not erupted.Some of the long bones have been gnawed by animals. The left sideof the mandible shows green stain.Burial No. 10 {USNM No. 378819).?Adult male. Lacks mainlythe left radius, left ulna, left tibia, fibulae, sternum, some ribs and484172?59 14 672 BUREAU OF AJMERICAN ETHNOLOGY [Bull. 174 vertebrae, and most of the hand and foot bones. Animal tooth marksare limited to the foramen magnum of the skull, the right ulna, andthe pelvic bones. The occiput of the skull has been very slightly morecompressed on the left than on the right (pi. 96).The symphyseal surface of the pubis has reached a plateau but isstill lacking a rim. This stage, which probably corresponds to anage of 30-35 years, is consistent with the amount of tooth wear andectocranial suture closure present : second degree wear and closure ofthe coronal and sagittal sutures chiefly around bregma. Beginningosteophytosis appears in the upper thoracic vertebrae.No pathological changes were noted. A small third trochanter ispresent in each femur. Septal apertures are not present in thehumeri.Bunal No. 11 {VSNM No. 378820).?K^wlt female. Althoughsome bones are damaged, only a few parts are missing : left clavicle,sternum, and various smaller bones. All of the bones present are freefrom animal tooth marks. The occiput has been symmetrically com-pressed (pi. 97).As so often in Indian women of middle age, the symphyseal surfaceof the pubis is distorted (Stewart, 1957). On the other hand, theteeth are moderately worn and ectocranial suture closure is advancedin the region of bregma. Osteophytosis is general throughout thevertebral column, but still not above 2+ at any point. All of thissuggests an age of 35+ years. It is noteworthy also that large inter-proximal caries are present in the upper right second and thirdmolars. Less advanced interproximal caries are present in some ofthe other teeth.Death may have been due to a penetrating wound in the posteriormedial angle of the left parietal. In contrast to the sharply definedexternal opening, which is rectangular (15X3 mm.), the internalopening is craterlike, due to the flaking of the internal table over acircular area 2.4 to 2.7 cm. in diameter. The area of breakage lacksthe fresh look which characterizes damage made in the process ofexcavation. A second small wound (5X3 mm.), obviously made bythe point of the same instrument, is located just medial to the leftparietal boss. Here penetration is through the external table intothe diploe. In each case the shape of the external opening andstraightness of the edges of the opening suggest a metal instrument.No pathological processes were seen. The fifth lumbar vertebrahas an arch defect through the pars interarticularis on the right side;on the left the arch is intact but the articular facets are distorted bymarginal lipping. A small third trochanter is present in the leftfemur ; none is present on the right. Pinpoint-sized septal aperturesare present in both humeri. Wedbl] an nSfTRODUCTION TO KANSAS ARCHEOLOGY 673n. YOUNG SITEThree burials, one including at least two individuals, were recoveredfrom this early site in Scott County. All individuals appear to beadult females. In contrast to the good state of preservation of thebones from the Doniphan site, those from the Yoimg site are veryfragile. Only one skeleton, No. 3, approaches completeness. What-ever measurements could be taken are assembled in tables 1 to 4. Therelationships of this group are discussed on page 680.Burial No. 1 {JJSNM No. 379758).?At least two adult females.The following parts are present: Right mastoid process, 2 upper jaws,left half of a lower jaw, 6 loose teeth, 2 small rib fragments, distalend of a left humerus, a right femur lacking distal end, and part of aright innominate. The sciatic notch of the innominate and the lineaaspera of the femur have been gnawed by animals.If one of the upper jaws belongs with the lower jaw?and this isuncertain?at least two individuals are represented. Judging fromthe innominate, one is a female; judging from the size of the jaws,both are females. The sizes of the other bones are consistent withthis interpretation.All of the teeth present show extreme wear. In addition the sacro-iliac joint shows degenerative changes, probably due to childbearing.Thus both individuals are adult and perhaps in middle age.The parts present are not complete enough for measuring. Thefemur lacks a third trochanter. The humerus lacks a septal aperture.Bunal No. 3 {USNM No. 379759).?KdixAt female. The follow-ing parts are present: Nearly complete skull and lower jaw (fig.107) ; C2-3, 5-7, Tl vertebrae; part of the right scapula; distal partof the left clavicle; pair of humeri (head of left damaged) ; pair ofradii (proximal end of left damaged) ; pair of ulnae (distal end ofright damaged) ; left second and third metacarpals; a phalanx fromthe left hand; pair of innominates (right more complete than left) ; pair of femora; pair of tibiae (right damaged at both ends) ; pair offibulae (distal end of right and proximal end of left damaged) ; theright talus ; and the right calcaneus.In the absence of the pubes, age must be stated in general terms onthe basis of tooth wear, suture closure, and joint changes. Tooth wearis moderate, but suture closure is minimal and joint lipping is not evi-dent. The age, therefore, is probably around 25-30 years.The only pathological changes seen are in the leg bones. The lowerhalf of the right fibula exhibits some roughness on the lateral surfaces.An extension of this condition appears on the posterior medial borderof the right tibia in the middle third. Probably such localized rough-ness represents healed periostitis from an old injury. 674 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174 Figure 107.?Four stereographic views of the Young site skull No. 3 (USNM No. 379759),a young adult female. Note the hyperdolichocrany (63.4) and the suggestion of Negroidappearance in the face. Wkdel] an introduction TO KANSAS ARCHEOLOGY 675 Figure 108.?Four stereographic views of the Young site skull No. 5 (USNM No. 379760),a female around 40 years of age. Note the dolichocrany (73.0). 676 BUREAU OF AMERICAN ETHNOLOGY [Boll. 174The femora lack third trochanters and the humeri lack septal aper-tures.Bunal No. 5 {VSNM No. 379760).?KdnM female. The follow-ing parts are present: Nearly complete skull and lower jaw (fig.108), Cl and 2 vertebrae, a right humerus (proximal end missing),a right radius and ulna fused at the proximal end (distal end of ulnamissing), an accessory bone (?) perhaps related to the pathologicalfusion of the radius and ulna, 3 phalanges of the fingers, left in-nominate (damaged), pair of femora (extremities damaged), distalend of the left tibia, the left metatarsal, and 2 phalanges of the toes.The circumarticular area of the ilium has female characters. Thisidentification is supported by most of the sex characters of the skull,although the supraorbital ridges are above average in prominence fora female and the superior orbital margins are rounded as in mostmales.Age determination in this case, as in No. 3, depends largely on toothwear, suture closure, and joint changes. Tooth wear is extreme andendocranial suture closure is complete; but ectocranially the suturesare mostly still visible. In addition, the facet for the dens on Cl ismarkedly lipped. All of this suggests an age of not less than 40 andperhaps above 40.An old injury has led to the fusion of the right radius and ulnajust below the elbow joint. The resulting position of the radius andhand is mid-pronation. Flexion of the elbow appears not to have beenaffected, although on the humerus the joint surface is roughened andthe joint margins are lipped. A loose nodule of bone (13X11X6mm.) was found in the vicinity of the head of the radius. Since thelatter has suffered postmortem damage, the nature of the original in-jury, whether fracture or infection, and the explanation of the bonenodule, are not clear. Also, the middle third of the shaft of the ulnaappears to be more slender than usual. Perhaps, therefore, some boneatrophy has occurred in this area. Wedbl] AN INTRODUCTION TO KANSAS ARCHEOLOGY 677jaddn 'xaptn 'ob^ ie}o; 'xeptn -obj ?XBtn -ziq "raBja%q -sBU-'^d "Aiv ??q noist3u-ao;a9j^ xapm -j^ ?ti]ni "Jj 'niBia; pjoqo -J j; (00)A4IOt3dBO IBtnBJQ 9inponi iBiUBJQ (UnT3)xopnj -m tiBoiM xapni *;q treaj^ xapm iBiuGJo ?;q ?Sojq-'jodtg ?;q ?39Jq--SEa: ?XBtn 'JBI 'nitjia ?XBin?jsod-*;nB ?iiiEia;ON IMNSil ?OJSE puna ?JGo COQO00 ?3 CO MS C rt ? ? I -,"0 ^?X3 CO c: r^ O od Qc ^ ^ CO cc C^ CO MOOCOOO -co o: ot-H C-l lO COcom I- t^GO 00 a: c:t^ t^ r^ t^Mcomco xapm "AjB '^xa ?aq -AiB -^xa ??i -AiTj -^xa xapuT IBSBN ?jq yBSBM ??q IBSBN ueaui'xopui 'qjo ucara '-jq -qJO weava ''^q "qJO (?saa)9l3uB 'loaAiv(?3aa)9l2UB IBlOBj; ?jd "Aicajd ? SB q p n a "jd 'SBn?qns-'SBqopna ?SBn--SBqopnaONPMNSn ?0^v[ IBIJTia ?S-Q 9^ -^'(i CSO MN H" lOTrrJf ?QOiOOOOcc "Q CO eo t>-01000 lO 40 -^ ?c pOgQ 678Table 2.- BUREAU OF AMERICAN ETHNOLOGY [Bull. 174-Kansas femora: Individual measurements {mm.) and indices by site,side and sex Site Whdbl] an introduction TO KANSAS ARCHEOLOGY 679Table 4. ? Kansas long hones: Measurements (mm.) by site, side, and sex Site 680 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174inance of males in both collections, and in part it is due to the round-ness of the skulls, which accentuates the size of the faces. Yet itseems quite likely that more adequate collections will confirm suchan impression.Next it is noteworthy that the roundness of the skulls in both col-lections is due, at least in part, to cradleboard flattening. Just howmuch flattening has occurred in any one case is difficult to say. Also,the point of greatest compression is lower on the occiput in theSteed-Kisker series (Stewart 1943, pis. 49-50) than in the Doniphanseries. Probably this represents a variation in cradling practice.However, the effect of compressing the lower part of the occiput hasbeen to increase the vault height. Thus, the Steed-Kisker skullsappear to have higher vaults than the Doniphan skulls. Using amean height index involving the biporionbregma height, 5 Steed-Kisker skulls (all males) range from 71.6 to 75.5, whereas 4 Doniphanskulls (2 males, 2 females) range from 68.4 to 75.1. More materialwill be needed to tell whether this observation is significant. Yet itseems likely that, if the Doniphan skulls had not been deformed, theywould have been classed as lowheaded. The Steed-Kisker skulls giveno hint as to their original shape.Turning to the face, one is struck by the massiveness alreadymentioned. Beyond this it is difficult to tell which characters repre-sent group differences and which are simply individual variations.It can be said only that both groups tend to the same sort of nasalstructure, and especially to prominence of the nasal bones.In the postcranial skeleton size again is the outstanding feature.Yet, except for one very tall individual in the Steed-Kisker series(femora No. 57) , the difference between the two groups is not striking.The absence of septal apertures of the humerus is perhaps noteworthyalso.Young vs. 'Woodruff and MissouH IlopeweUians.?Before makingthis comparison attention should be called to the contrast in type be-tween the two skulls recovered from the Young site. No. 3 is hyper-dolichocranic (63.4), whereas No. 5 is dolichocranic (73.0). Ob-viously, these skulls are undeformed. No. 3, unlike No. 5, also hasa combination of facial features which reminds one of the AfricanNegroes. The same facial appearance occurs in combination withextreme longheadedness in the cave dwellers of the Big Bend regionof Texas (Stewart, 1935). The face of No. 5, on the other hand,is not so distinctive and does not suggest any particular group.Unlike these two Early (?) Woodland skulls, the single MiddleWoodland skull from the Woodruff ossuary is somewhat deformed andmost of the Missouri Ilopewellians are considerably and distinctivelydeformed. In spite of their altered shapes, the Hopewellians give Wbdbl] an introduction to KANSAS ARCHEOLOGY 681indications of being fairly longheaded normally. On this score theWoodruff skull gives little indication. Thus one is easily persuadedto place the Early ( ? ) Woodland skulls with the Hopewellians,largely on the basis of longheadedness and general gracility.The Woodruff skull was regarded as male by Gumming (Kivett,1953, p. 140), but to me it looks more like a female (the pelvis ismissing) , The published pictures of this skull (ibid., pi. 28) show thedeformity to be like that common to the Doniphan series?high onthe occiput and asymmetrical. Not so evident in the picture is thebroken tip of the nose and the size disharmony between the fore- andback-parts of the mandible. Actually the ascending ramus of themandible is of slight build as in most females, but the symphysealarea is high and heavy as in many males. Probably it is this com-bination of features which reminds one of the Doniphan skulls, ratherthan the earlier Woodland-Hopewell skulls. However, I am im-pressed by the high-headedness of the Woodruff skull, and, havingbut a single specimen, this makes me hesitate to place it with anyparticular group.In view of this discussion of the skeletal evidence, something shouldbe added to Wedei's remarks (see p. 91) about the putatively an-cient Lansing skull found a few miles distant across the Missourifrom the Steed-Kisker site. Wedei's analysis of Hrdlicka's procedurein this case agrees with what I have written (Stewart, 1949, pp. 10-12) about the development of the concept of morphological dating.During the interval from 1902 to 1907 when he was studying theLansing skull, Hrdlicka had very little perspective on cranial typesfrom the Plains. Not until he prepared the portion of his "Catalogof Crania" dealing with this area (Hrdlicka, 1927) did he gain theperspective that enabled him to change some of the earlier attribu-tions. Only then, too, did he realize that there was a lowheaded strainin the late tribes of this area. Characteristically, he did not correcthis earlier statements about the Lansing skull. It is clear to me nowthat he continued to regard this skull as of the "Algonkin" type andhence of no great age.By comparison with the fragile Young site specimens, the Lansingskull seems well preserved?even fresh looking, as was remarked atthe time. However, the Young site material was not deeply buriedand this probably accounts for the difference in preservation. In theabsence of a face on the Lansing skull, extended comparison is outof the question. On the other hand, of the two skulls from the Youngsite only No. 5 is rouglily similar in type to the Lansing skull. Howclosely these skulls of opposite sex compare is shown in figure 109,where stereographic drawings of top and side views appear, one skullsuperimposed on the other. 682 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174 Wbdel] an introduction TO KANSAS ARCHEOLOGY 683In lateral profile the main difference, aside from the disparity insize, is in the conformation of the occiput, the Lansing skull havinga more prominent occiput. As seen from above, the distortion of theLansing skull is very noticeable : the left side of the frontal extendsforward more than the right and the outlines of the parietals areasymmetrical. Although the Lansing skull was assembled from manypieces, the asymmetries do not seem to be a product of the restora-tion. Owing to such defects nothing more can be learned from thiscomparison. Incidentally, the Lansing skull has a minimum frontaldiameter of 94 mm., a frontal chord of 113 mm., and a resultingfrontal index of 83.2. This compares with a frontal index of 83.5for Young skull No. 5.All of this simply bears out the impression of rough similarity withwhich we started and demonstrates the problem of interpreting theaffiliation of a single defective skull. Certainly, if two skulls froma site, such as Young skulls Nos. 3 and 5, can be so different, it is notimpossible that the Lansing skull fits into such a population.LITERATURE CITEDHrdliCka, AleS1907. Skeletal remains suggesting or attributed to early man in NorthAmerica. Bur. Amer. Ethnol. Bull. 33.1927. Catalogue of human crania in the United States National Museumcollections : The Algonkin and related Iroquois ; Siouan, Caddoan,Salish and Sahaptin, Shoshonean, and Californian Indians. Proc.U. S. Nat. Mus., vol. 69, art. 5, pp. 1-127.KlVETT, Maevin F.1953. The Woodruff ossuary, a prehistoric burial site in Phillips County,Kansas. Bur. Amer. JJthnol. Bull. 154, Riv. Bas. Surv. Pap. No. 3,pp. 103-141.Stewart, T. D.1935. Skeletal remains from southwestern Texas. Amer. Journ. Phys.Anthrop., vol. 20, No. 2, pp. 213-231.1943. Skeletal remains from Platte and Clay Counties, Missouri. U. S. Nat.Mus. Bull. 183, appendix, pp. 245-273.1949. The development of the concept of morphological dating in connectionwith early man in America. Southwestern Journ. Anthrop., vol. 5,No. 1, pp. 1-16.1957. Distortion of the pubic symphyseal surface in females and its effecton age determination. Amer. Journ. Phys. Anthrop., n. s., vol. 15,No. 1, pp. 9-18.EXPLANATION OF PLATES 1-97Plate 1Views at the Doniphan site, 14DP2.a, Looking west toward Doniphan, across part of site 14DP2.b, Looking west across site 14DP2, toward house 1 on hilltop at center,c, House 2 after excavation, looking southwest. 684 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174Plate 2Burials 1-4 at Doniphan site, 14DP2, presumably Kansa.0, Burial 1, outside south wall of house 1 ; note leather lying across knees.6, Burial 2, intrusive into entrance passage of house 1 ; note stump of post andworked shell near knees.c, Burial 3, intrusive in northeast side of house 1.d. Burial 4, semiseated adult partially covered with stones.Plate 3Burials, some slab-covered, in situ at Doniphan site.a, Group of stone-covered graves, Nos. 6-8, after clearing of topsoil.6, Burial 8, reclining and partly stone-covered.o, Burial 10, fully extended and without stone covering.d. Burial 11, partly covered with slabs.Plate 4Restored pottery vessels from house 2, Doniphan site.a, Plainware vessel from house 2, cache 2, diameter 24 cm. (USNM 381622).6, Vessel with trailed shoulder decoration from house 2, cache 1, diameter26.5 cm. (USNM 381644).c, Cord-roughened vessel from house 2, cache 2, diameter 22 cm. (USNM381621). Plate 5Shell, bone, and antler artifacts, Doniphan site.o, 6, g, from house sites (Nebraska aspect) ; all others from cache pitspresumed to be 18th century Kansa.Plate 6Catlinite and limestone artifacts, Doniphan site. All from cache pits presumedto be 18th century Kansa. Actual size. (USNM Neg. 44762 H).Plate 7Iron and brass artifacts, Doniphan site. All from cache pits, except i fromburial 5. Length of e, 16.5 cm. (USNM Neg. 44762 B).Plate 8House 1 (o) and restored pottery vessel, Fanning site, 14DP1.a, House 1, excavated, from the northeast ; entrance to southwest.&, Restored pottery vessel with four tapered strap handles, from cache pit12, Fanning site. Diameter 27.5 cm. (USNM 381751).Plate 9Trailed and plainware potsherds and vessel handles. Fanning site.Plate 10Trailed potsherds and miniature bowl, Fanning site.Plate 11Rimsherds of Lower Loup Focus types. Fanning site.Or-f, Grit-tempered, collared, usually with multiple handle scars.g, h. Shell-tempered, with thickened rim and diagonal Incisions. Actualsize (USNM Neg. 44762). Wbdhl] an introduction TO KANSAS ARCHEOLOGY 685PlATE 12Miscellaneous artifacts from Fanning site (USNM Neg. 34230 A) . PI.ATE 13Brass and iron (6, c) objects, Fanning site. Length of a, 16.5 cm. (USNMNeg. 44762 F). Plate 14Burial mound excavations in Doniphan County.a, Guthrie site, burial mound on hilltop at right.6, c, Guthrie site, burial pit before and after removal of slab cover andgrave fill.d, Matherson burial mound, covering slabs stiU in situ.Plate 15Surface finds of early-type projectile points from Doniphan and Seward Counties,Kansas, o, 6, d-g, i, k, Fenn Ward collection ; c, h, j, Shelby Gilmore collec-tion ; Ir-o, USNM 382038.a, Near Highland, Doniphan County; gray chert, blunted proximal edges,thinned base ; 52 x 22 x 6.5 mm.&, Near Highland, Doniphan County; gray mottled chert, blunted proximaledges, thinned base ; 46 x 23 x 5 mm.c, Shelby Gilmore farm, 2^4 miles north of Highland, Doniphan County;gray chert, blimted edges ; 44 x 20 x 4 mm.d, Near Highland, Doniphan County; light gray chert, blunted proximaledges ; 52 X 20 X 7 mm.e, 2 miles west and 3 miles north of Doniphan, Doniphan County; blackjasper, proximal edges blunted, base thinned; 60 x 29 x 6 mm./, 5 miles southeast of Highland; mottled gray and tan chert; proximaledges blunted ; 99 x 34 x 7 mm.g, Near Highland ; light gray chert, proximal edges blunted ; 65 x 27 x 5.5mm.h, Shelby Gilmore farm, 2i/^ miles north of Highland, Kans. ; gray chert,broadly fluted, fine edge chipping, base missing.i, 1 mile east and 2V^ miles south of Troy, Doniphan County; light graychert, broadly fluted, fine edge chipping ; 43 x 20 x 4 mm.;, Shelby Gilmore farm, 2^^ miles north of Highland; light-gray chert,fluted, proximal edges blunted ; 55 x 23 x 6 mm.fc. Near Highland; light-gray chert, proximal edges somewhat blunted;97 X 32 X 8 mm.I, Near Liberal, Seward County ; Alibates dolomite, fluted base fragment,m. Near Liberal, Seward County ; Alibates dolomite, fluted midsection.n, Near Liberal, Seward County ; Alibates dolomite, fluted base fragment.0, Near Liberal, Seward County ; Alibates dolomite, thinned basal fragment.Plate 16Excavations near Manhattan.a, House 1, Kansa village site, 14P024, near mouth of Blue River east ofManhattan,6, Portion of Griffing site, 14RY21, on Wildcat Creek west of Manhattan.c. House 1, Griffing site, looking south.Plate 17Riley Cord-Roughened potsherds from Grifling site. 686 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174Plate 18Miscellaneous objects from Kansa village site 14P024.a-?, from village site excavations; others from surface of burial ground.Length of o, 20 cm. (USNM Neg. 44762 D ) Plate 19Miscellaneous iron and brass objects from Kansa village site 14P024. All fromvillage site excavations. Length of /, 13 cm. (USNM Neg. 44762 C).Plate 20Stone, bone, and shell objects from burial mounds near Manhattan. Chippedpoints, bone beads, and shell pendants from Fremont Point Mound No. 1.Pierced and incised bear canine tooth from Fremont Point Mound No. 2. (Allcollected by Dr. Norman L. Roberts, Manhattan.)Plate 21Tobias site (14RC8), Rice County.a. General view looking north toward timbered valley of Little ArkansasRiver.&, o, Start of excavations in mound 17 complex.PLATE 22Excavations at Tobias site, Rice County.a, Sandstone boulders and human remains in fill of basin 1, mound 17complex,&, Disarticulated human bones in fill of basin 1, mound 17 complex.Plate 23Excavations at Tobias site, Rice County.a, Basin 2, mound 17 complex, under excavation, looking northeast.&, Pottery, charred basket and corn, pipe, and other materials on floor ofbasin 2, mound 17 complex. Plate 24Excavations at Tobias site. Rice County.a. Basin 1, mound 17 complex, looking northwest.b, Basin 2, mound 17 complex, looking northeast ; basin cleared, except fortwo control blocks. Plate 25Miscellaneous features in mound 17 complex, Tobias site.a, Mussel shells, bone awls, and chipped stone on floor of basin 2.&, Glass, bone, and turquoise necklace on floor of basin 1.c. Charred remains of coiled basket, floor of basin 2.d, Charred corn, floor of basin 2.Plate 26Restored pottery vessels of type Geneseo Plain, from Tobias site, Rice County.a, USNM 388754, height 23 cm., diameter 21.5 cm.&, USNM 388622, height 21 cm., diameter 19 cm.c, USNM 388898, height 24 cm., diameter 23 cm. Wbdkl] an introduction TO KANSAS ARCHEOLOGY 687Plate 27Restored pottery vessels from Tobias site, Rice County.a, USNM 388644, height 36 cm., diameter 31 cm., type Geneseo Plain.&, USNM 388590, diameter 31.7 cm., type Geneseo Red-filmed, basal portionmissing.c, USNM 388624, height 5 cm., diameter 6 cm., miniature vessel.Plate 28Restored pottery vessels, type Geneseo Simple Stamped, from Tobias site. RiceCounty.o, USNM 388623, height 24 cm., diameter 29 cm.6, USNM 388983, height 27 cm., diameter 27.3 cm.Plate 29Restored pottery vessels from Tobias site, Rice County.a, USNM 388888, height 30.5 cm., diameter 28 cm.; type Geneseo SimpleStamped.6, USNM 388687, height 30 cm., diameter 29 cm. ; check stamped.Plate 30Rimsherds, handles, and sherd disks from Tobias site, Rice County. Height of/, 10 cm. Plate 31Bone and antler implements from Tobias and Malone (c) sites, Rice County.Plate 32Bone awls and polishing (?) tools from various Little River Focus sites in RiceCounty. (Malone site, c, m; Thompson site, a, b, d, j, o; all others fromTobias site.) Plate 33Bone projectile points and bipointed objects (o-t) from Tobias site, RiceCounty. Plate 34Worked antler strips and fragments, and cut mammal bone fragment (l),from Tobias site, Rice County. Plate 35Bone and antler artifacts from Tobias site. Rice County, and fragment of bonepaint applicator (7;,) from Larcom-Haggard site, Cowley County.Plate 36Bone and iron (k, I) objects from various Great Bend Aspect sites in Rice andCowley (&, c) Counties.Malone site. Rice County, j; Thompson site. Rice County, /, 7i, i; Elliottsite, Cowley County, h, c; Tobias site. Rice County, a, d, e, g, k, I.Plate 37Chipped projectile points from Tobias site. Rice County.Plate 38Stemmed (b-g) and plain-shafted {li-m) chipped drills from Tobias site. RiceCounty.484172?59 45 688 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174Plate 39Expanded-base chipped drills from Tobias site, Rice County.Plate 40Chipped knives from Tobias and Thompson (d, k, I) sites, Rice County.Plate 41Top and side views of chipped end scrapers from Tobias site. Rice County.Plate 42Grooved mauls and shaft polishers from various sites in Rice and CowleyCounties. Length of a, 10 cm.; all others to same scale except h, i. (a, b, d,f, g, i, Tobias site, Rice County ; c, Thompson site. Rice County ; h, PaintCreek site, McPherson County, private collection of J. H. Fries ; e, Larcom-Haggard site, Cowley County, a and ft are Type I mauls ; c and e, Type II ; g, d, Type III; /, Type IV. Diameter of /(, 75-85 mm., thickness 40-43 mm.)Plate 43Grinding stones from Tobias (o, d, f, g, h) and Thompson (6, c, e) sites, RiceCounty. (Not to scale.) Plate 44Sandstone disks and abrading stones from Tobias and Malone (/, h, i) sites.Rice County. Plate 45Stone pipes from Tobias site and vicinity. Rice County.Length of i, 77 mm. ; b-e, h, to same scale.a, f, g, Surface finds in collection of Lowell Peverley ; all others fromTobias site excavations, U. S. National Museum.Plate 46Miscellaneous objects from Rice and Cowley (No. 458) Counties.o, Chipped blade and necklace of glass, bird bone, and turquoise frombasin 1, mound 17 complex, Tobias site.b, Shell ornaments and implements.Plate 47Miscellaneous objects from Rice and Doniphan (d, g) Counties.Upper and Or-c, worked shell from Tobias site, Rice County ; /, miniaturepot from pit 1, Tobias site (USNM 388591) ; d, g, small lead cross andcatlinite pipe, Doniphan site. Plate 48Thompson site, 14RC9, Rice County.a, Portion of village site and mound 1 excavations (center) ; looking south-west toward Little Arkansas River.6, Trenching mound 1 ; looking northwest.Plate 49Pueblo and other exotic potsherds from various sites in Rice and CowleyCounties. From Rice County : a-d, g, Thompson site ; /, Tobias site. FromCowley County : e, ft, ;, k, Larcom-Haggard site ; i, Arkansas City CountryClub site. Wedbl] an introduction TO KANSAS ARCHEOLOGY 689Plate 50Oxidized chain mail fragments, pit 4, Thompson site (USNM 389152). Enlarged.X2. Plate 51Larcom-Haggard site, 14C01, Cowley County.a, From northwest, toward Walnut River ; site in middle foreground, beyondrailroad tracks.&, c, From southeast, showing part of village terrace and excavations.Plate 52Arkansas City Country Club site, 14C03, Cowley County.a. Mound 1 excavations, looking northwest. Note flat-topped mounds inbackground, left of clubhouse.6, Mound 1 excavations, looking south; pit 0, unexcavated, at right inexcavation. Plate 53Restored pottery vessels, type Cowley Plain, from various sites in CowleyCounty, a-o, Arkansas City Country Club, mound 1, pit B ; d, Elliott site,pit 8. Height of a, 28 cm. ; others to same scale.Plate 54Restored pottery vessels, type Cowley Plain, from Larcom-Haggard site. Col-lection of Louis Essex.a, Height 31.8 cm., diameter 26.3 cm. ; 6, height 17.8 cm., diameter 16.8 cm.Plate 55Socketed scapula hoes from Cowley County sites.Plate 56Bone and stone artifacts from Cowley and Lane County sites. Length of &,15.8 cm. ; a, c-g, to same scale, fc, Pottorff site, Lane County.Plate 57Pottorff (14LA1) and Risston (14SC4) sites on Salt Creek, on Lane-ScottCounty line.a, Pottorff site looking southeast across Salt Creek and main site terrace.6, Risston site, center, and Salt Creek beyond, looking southeast.Plate 58Houses 1 and 2 excavations, Pottorff site, 14LA1.a, House 1 from the west, shov/ing posthole arrangement and central fire-place.&, House 2 from the west, entrance of house at lower right,c. Relationship of house 1, right, to house 2 at lower level.Plate 59Excavations at the Pottorff site, 14LA1.a. East face of test pit, between Stakes 019:12 (left) and 020:2, Area C,showing burned areas and refuse bone.&, West face of test pit in Square T27 : 4, showing Upper Republican stratum(A), Woodland stratum (B), unidentified stratum (C), and sterile de-posits (D) intervening. 690 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 174Plate 60Pottery from Upper Republican and Woodland occupations, Pottorff site.?-/, Upper Republican rim and body sherds ; /;, rocker-marked sherd frombeneath floor of house 2 ; T^n, Woodland body sherds from lower stratum(Occupation B). Plate 61Bone and shell objects, Pottorff site. All from Upper Republican stratum, ex-cept m and n. Plate 62Projectile points, drills, and end scrapers, Pottorff site.Plate 63Knives and other chipped forms, Pottorff site.Plate 64Stone and bone objects from Occupation B (Woodland), Pottorff site.Plate 65Scott County State Park and site 14SC1.a, Looking south up the Ladder (Beaver?) Creek valley and across LakeMcBride. Arrow points to pueblo ruin and nearby excavations at site14SC1.b. Trench in test 2 through refuse deposits north of pueblo ruin; 14 SCI.Foundation stones in pueblo ruin visible as light spot in shadowed endof trench, behind kneeling figure.Plate 66Excavations at 14SC1, in Scott County State Park.a, b, Test 1 and cache pit, south of pueblo ruin.c. Test 3 and roasting pit No. 1, lower left.d, Hearth on terrace east of Big Spring, on right bank of Ladder Creek.Plate 67Roasting or baking pit No. 3, Scott County State Park.a. Pit sectioned, half cleared to stones on floor.b, Pit sectioned, fill partly removed to show profile and mouth below surfacedisturbance.c. Pit cleared, burned stones in situ on floor.d, Pit fully cleared, showing burning of wall as light-colored lip belowbroken edge. Plate 68Restored pottery vessels from 14SC1, Scott County.a, Small vessel, restored, from test 2. Height 7 cm. (USNM 386959).b. Large jar collected by A, T. Hill from cache pit near bathhouse duringdraining of Lake McBride, 1944-45. Height 36 cm., diam. 30-35 cm.Dismal River Aspect. Plate 69Pottery objects from Scott County site, 14SC1. Actual size.a, Tewa Polychrome sherd, shore of Lake McBride.b-d, h, Pottery pipe fragments with incised and pricked decoration.g, i-k, Undecorated pottery pipes, all tubular in form except ;. Wedel] an introduction TO KANSAS ARCHEOLOGY 691Plate 70Bone and shell artifacts from 14SC1, Scott County.a, bird-bone whistle ; &-c, bone beads ; d, paint applicator ; e, bone ornamentfragment ; /, worked bison hyoid ; g, needle fragment ; /(, projectile point ; i, polishing tool ; ;, k, shell ornaments ; I, m, awls of split legbone ; n, awlof split rib ; o, p, splinter awls ; q, r, "rib-edge" awls. Length of a, 7.9cm. Dismal River Aspect. Plate 71Bone artifacts from 14 SCI, Scott County.a-d, Bison metapodial fleshing tools or grainers.f-j, Finished and striated rib implements, purpose unknown. Dismal RiverAspect. Plate 72Projectile points, drills, and end scrapers from 14SC1, Scott County. DismalRiver Aspect. Plate 73Stone, iron (g), and pottery (h) objects from 14SC1, Scott County.a-c, Scrapers and cutting or chopping tools ; d-f, knives ; g, iron knife blade ; g', marcasite concretion ; h, miniature pottery vessel ; i, mano ; ;, limestonepipe blank (?). Dismal River Aspect.Plate 74Young burial site, 14SC2, Scott County.A, Burials under excavation ; Lake McBride and Scott County State ParkIn background.B, Stone and bone artifacts from burials 3 and 5. a, stemmed projectilepoint; 6, drill (burial 5) ; c, bone awl; d, bone beads; e, g, top and, /, sideviews of split-pebble quartzite scrapers.Plate 75Burials in situ, Young burial site, Scott County.fl, Burial 1, showing stones in association.&, Burial 3, adult female, with terrapin carapace lying over right elbow,c, Burial 5, adult female, vrith artifacts visible under left tibia and knee.Plate 76Petroglyphs at Inscription Rock, 14EW1, Ellsworth County. Petroglyph localityNo. 1. Photographs by G. L. Whiteford,Plate 77Petroglyphs from various localities in Kansas.a, Inscription Rock, petroglyph site No. 1, Ellsworth County.b, Petroglyph site No. 6, 14RC10, Rice County.c, Petroglyph site No. 7, 14RC11, Rice County.d, Petroglyph site No. 10, 14MY1, near Liberty, Montgomery County. Notewar-bonneted horsemen at lower center to left of watch.Plate 78Petroglyph sites in Rice County.a, Spriggs Rocks near Little River; petroglyph site No. 5, 14RC1.6, Sandstone bluff at spring, petroglyph site No. 6, 14RC10. g92 BUREAU OF AMERICAN ETHNOLOGY [Bull. 174PlATE 79Anthropomorphic and other petroglyphs at Spriggs Rocks, Rice County.Plate 80Potsherds from the Roiiiger site, 14CS1, in Chase County.a-e. Middle Woodland (Hopewellian) sherds, showing dentate stamping,zoned decoration, punched bosses, cord-wrapped dowel impressions, andother features. (From Roniger collection. Bazaar, Kans.)/-;, Smoky Hill Aspect sherds, probably type Riley Cord-roughened, sur-face finds in zone showing burned house wattle, small projectile points,etc. /, g, i, rimsherds. (From Roniger collection, Bazaar, Kans.)Plate 81Potsherds from Pratt site, 14PT1.a, b, d, e, Cord-roughened rim and body sherds.0, /, Simple stamped rimsherds suggesting Dismal River type Scott Plain.g, Shell-tempered rimsherds with pinched-up neck nodes.h, I, m, Rio Grande glaze-decorated sherds, Mera's "group C" or Kidder's "Glaze III," date circa 1450-75.i, Biscuit B (Bandelier Black-on-gray) sherd, Tewa area north of Santa Fe,circa 1425-50 to 152-5-50.k. Unidentified sherd with traces of black paint. All from Eugene Wingcollection, Pratt, Kans. Plate 82Chipped and ground stone objects from Pratt site, 14PT1.o, 6, Projectile points ; c, drilled and polished turquoise ; d-f, drills ; g, pipefragment; h, i, end scrapers; f, l-q. knives; k, reamer or large drill. (Allfrom Eugene Wing collection, Pratt, Kans.)Plate 83Bone and ground stone objects from Pratt site, 14PT1. All from Eugene Wingcollection, Pratt, Kans. Plate 84Miscellaneous specimens from various sites in Kansas.a, Shell hair pipes, 6, glass beads, and c, clay trade pipe, from Indian grave,14PT2, near Pratt, all from Eugene Wing collection; d, replica of stonedisk pipe from White Rock site, 14JW1, courtesy of A. T. Hill ; e, steatitepipe, Wyandotte County (USNM 326648) ; /, indurated clay pipe DoniphanCounty (F. Ward collection, Highland) ; g, boatstone from Labette County(USNM 232588). Plate 85Sallna burial pit, 14SA1, Saline County.a. Portion of burials from southwest corner, 1940.&, Flexed, semiflexed, and child burials, 1940.Plate 86Salina burial pit, 14SA1, Saline County.a. Adult burial 81 and pottery vessel 55, in situ.6, Burials 111 and 113, and pottery vessel 110, in situ. Wedel] an USTTRODUCTION TO ICANSAS ARCHEOLOGY 693Plate 87Pottery vessels from Saliua village and burial site, 14SA1.a, Globular jar (No. 110), 15 cm. diameter, 15 cm. height.&, Restored jar from house 1, diameter 37 cm., height 31 cm.Plate 88Pottery vessels from Salina village site, 14SA1.o. From cache 5, house 1 ; diameter 29 cm., height 26 cm.b, From fireplace, house 1 ; diameter 33 cm., height 33 cm.Plate 89Pottery vessels from Salina burial pit, 14SA1.a, No. 73, diameter 18.5 cm., height 17 cm.6, No. 84, diameter 16 cm., height 13.5 cm.c, No number, diameter 14 cm., height 12 cm.Plate 90Pottery vessels from Salina burial pit, 14SA1.a, No. 28, diameter 8.9 cm., height 8.9 cm.6, No. 88, diameter 14 cm., height 11.5 cm.c, No number, diameter 13.5 cm., height 11 cm.Plate 91Pottery from Salina village and burial site, 14SA1.a, Cord-roughened and incised vessel (cast), diameter 20 cm., height 17 cm.b, Effigy pipe fragment from house 1, cache 1.0, Crockett Curvilinear Incised bowl fragment (see also fig. 95).Plate 92Objects of stone, pottery (i), antler {h), and shell from Salina village and burialsite, 14SA1. Length of &, 25.5 cm. ; a, c-e, to same scale. Length of k, 10 cm. ; /-*, to same scale. Plate 93Neodesha "fort," Wilson County, 14WN1.a, Neodesha "fort" as it is thought to have appeared about 1870 ; sketch byT. M. Galey.6, Remaining segments of Neodesha "fort" ; sketch by T. M. Galey.Plate 94Four views of the Doniphan site skull No. 3.Plate 95Four views of the Doniphan site skull No. 4.Plate 96Four views of the Doniphan site skull No. 10.Plate 97Four views of the Doniphan site skull No. U. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 1 Views at the Doniphan site, 14DP2. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 2 Qt BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 3 BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 4 Restored pottery vessels from house 2, Doniphan site. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 5 Shell, bone, and antler artifacts, Doniphan site. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 6 a . , ^f!^V ysj n Catlinllc and limestone artifacts, Doniphan site. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 7 I Iron and brass artifacts, Doniphan site. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE Q Fanning site, 14DP1. a. House 1. h. Restored pottery vessel. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 9 Trailed and plainware potsherds and vessel handles, Fanning site. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 10 Trailed potsherds and miniature bowl, Fanning site. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 11 a -CK/?? g hRimsherds of Lower Loup Focus types, Fanning site. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 12 Miscellaneous artifacts from Fanniii" site. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE l3 o Brass and iron (b, c) objects, Fanning site. BUREAU OF AMERICAN ETHNOinOY BULLETIN 174 PLATE 14 I\V f? '-4 ^ ..'-- (i?* 3UREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 15 2 IN. Surface finds of early-type projectile points from Doniphan and Seward Counties, Kans. BUREAU OF AMFR1'~AM F~rMNOi r>r:v Ri H I FT IN 174 PLATE 16 fc^ ./ -V - BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 17 Riley Cord-roughened potsherds from Griffing site. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 18 Miscellaneous objects from Kaiisa village site, 14P()24. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 19 Miscellaneous iron and brass objects from Kansa village site, 14P024. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 20 Stone, bone, and shell objects from burial mounds near Ma BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 21 u Di BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 22 JL^-- -; ;>- Excavations at Tobias site, Rice County. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 23 Excavations at Tobias site, Rice County. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 24 %^ .y.-/wM.. ... mm- BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 25 BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 26 a b Restored pottery vessels of type Genesco Plain, from Tobias site, Rice County. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 27 Restured puUcr\- vessels frum 'i'ubias site, Rice County, BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 28 a b Restored pottery vessels, type Genesco Simple Stamped, from Tobias site. Rice County. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 29 Restored pottery vessels from Tobias site, Rice County BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 30 u BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 31 !S5i:Jj*';^feJSiitfi:^^'.^ BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 32 Bone awls and polishing(?) tools from various Lilllc River I'ocus sites in Rice County. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 33 Bone projectile points and bipointed objects {o-t) from Tobias site, Rice County. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 34 o I 5cm.Worked antler strips and fragments, and cut nianinial bone fragment (/), from Tobias site,Rice County. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 35 cm.Bone and antler artifacts from Tobias site, Rice County, and fragment of bone paintapplicator (A) from Larcom-Haggard site, Cowley County. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 36 < BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 37 P^ r-i \ cm.Chipped projectile points from Tobias site, Rice County. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 38 Stemmed (b-g) and plain-sliaflcd (li-m) chipped drills from 'lobias site, Rice County. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 39 Expanded-base chipped drills from Tobias site, Rice County. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 40 Chipped knives from 'I'obias and Tliompson ((/, k, I) sites, Rice County, BUREAU OF AMERICAN ETHNOLOGYw i0 BULLETIN 174 PLATE 41 ! ^>;\\\ Top and side views of chipped end scrapers from Tobias site, Rice County. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 42 Grooved mauls and shaft polishers from various sites in Rice and Cowley Counties. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 43 u^ H H O BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 44 ^im I Sandstone disks and abrading stones from 'iobias and iMalone (/, /;, i) sites, Rice County, BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 45 Stone pipes from Tobias site and vicinity, Rice County. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 46 BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 47 Miscellaneous objects from Rice and Doniphan {d, g) Counties. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 48 ^B^UBSSBFMf' Ihoinpson site, 14RC'>, In.c* L. ... BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 49 Pueblo and other exotic potsherds from various sites in Rice and Cowley Counties. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 50 S%^m*.'tf.^>-^^ii^*i''k'^*-.iL*~ Oxidized chain mail fragments, pit 4, Thompson site. (X 2.) BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 51 Larcom-Haggard site, 14C01, Cowley County. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 52 .'\rlvaii--.is Cii\ LmliiiU} LUib she, 14C03, Cowley Count}. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 53 u u Pi BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 54 w u BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 55 u u ^^ BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 56 u u mamm SK*K2-Ji??^^ <5 m ^"-,4s?er . ^ u m BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 72 H Ji 0k ^^xt cm. Projectile points, drills, and end scrapers from 14SC1, Scott County, BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 73 u u BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 74 u e ^^ f 9\oung iMirial site 14SC2, Scott Count}', BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 75 >^ BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 76 ^ ; JPctroglyphs at I nsci ipi ion Rock, 14KVV1, EUswonh County. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 77 Petroglyphs rom various localities in Kansas. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 78 Pc'lroirlyph sites in Rice Count}' BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 79 Anthropomorphic and other pctroglyphs at Spriggs Rocks, Rice County. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 80 Potsherds from the Roniger site, 14CS1, in Chase County. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 81 Potsherds from Pratt site, 14PT1. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 82 BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 83 f^^'WSrm^^FS'^s? Kmj'?'''sr^ BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 84 MliiiMiMiMii a Miscellaneous specimens from various sites in Kansas. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 85 Salina burial pit, 14SA1, Saline County. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 86 Saliiia buii.il jiit, 14S.\1, Saline County, BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 87 a b Pottery vessels from Salina village and burial site, 14SA1. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 88 Pollcry vessels from Salina village site, 14SA1. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 89 < BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 90 Pottery vessels from Salina burial pit, 14SA1. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 91 Pottery from Salina village aird burial site, 14SAI. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 92 D ..H;^ c Objects of stone, pottery (j), antler (A), and shell from Salina village and burial site, 14SA1, BUREAU OF AMERICAN ETHNOLOGY BULLETIN 174 PLATE 93 ^e^c y^Af^ f"-^^ ^i5e^^*