SMITHSONIAN INSTITUTIONBUREAU OF AMERICAN ETHNOLOGYBULLETIN 154 RIVER BASIN SURVEYS PAPERS Inter-Agency Archeological Salvage ProgramNUMBERS 1-6 i^ i f vv^ , SMITHSONIAN INSTITUTIONBUREAU OF AMERICAN ETHNOLOGYBULLETIN 154 RIVER BASIN SURVEYS PAPERS Inter-Agency Archeological Salvage ProgramNUMBERS 1-6 UNITED STATESGOVERNMENT PRINTING OFFICEWASHINGTON : 1953 For sale by the Superintendent of Documents, U. S. Government Printing OfficeWashington 25, D. C. ? Price $1.75 LETTER OF TRANSMITTAL Smithsonian Institution,Bureau of American Ethnology,Washington, D. C, April 19, 1951.Sir : I have the honor to submit the accompanying manuscripts, en-titled "Prehistory and the Missouri Valley Development Program:Summary Report on the Missouri River Basin Archeological Surveyin 1948," by Waldo R. Wedel; "Prehistory and the Missouri ValleyDevelopment Program: Summary Report on the Missouri RiverBasin Archeological Survey in 1949," by Waldo R. Wedel; "TheWoodruff Ossuary, a Prehistoric Burial Site in Phillips County,Kansas," by Marvin F. Kivett; "An Archeological Survey of theAddicks Dam Basin, Southeast Texas," by Joe Ben Wheat; "IndianSkeletal Remains from the Doering and Kobs Sites, Addicks Res-ervoir, Texas," by Marshall T. Newman; "Two Rock Shelters NearTucumcari, New Mexico," by Herbert W. Dick; "Geology of theHodges Site, Quay County, New Mexico," by Sheldon Judson; and"The Rembert Mounds, Elbert County, Georgia," by Joseph R. Cald-well, and to recommend that they be published as a bulletin of theBureau of American Ethnology.Very respectfully yours. Frank H. H. Roberts, Jr.,Acting Director.Dr. Alexander Wetmore,Secretary, Smithsonian Institution. PUBLISHEK'S NOTEWith this bulletin the Bureau of American Ethnology inauguratesa new series of Kiver Basin Surveys Papers, designed as an outlet forthe reports resulting from the Inter-Agency Archeological SalvageProgram.These papers will be numbered consecutively, a bulletin being de-voted to them from time to time as they accumulate. Where thereports are of sufficient size to warrant it, they will be issued as acomplete bulletin.A limited edition of bound volumes will be issued but the paperswill be generally distributed in separate form. CONTENTS PAGEForeword, by Frank H. H. Roberts, Jr xiiiNo. 1. Prehistory and the Missouri Valley Development Program: Sum-mary report on the Missouri River Basin Archeological Survey in 1948,by Waldo R. Wedel xvNo. 2. Prehistory and the Missouri Valley Development Program: Sum-mary report on the Missouri River Basin Archeological Survey in 1949,by Waldo R. Wedel 61No. 3. The WoodrufT Ossuary, a prehistoric burial site in Phillips County,Kansas, by Marvin F. Kivett 103No. 4. The Addicks Dam Site:I. An archeological survey of the Addicks Dam Basin, SoutheastTexas, by Joe Ben Wheat 143II. Indian skeletal remains from the Doering and Kobs Sites, Ad-dicks Reservoir, Texas, by Marshall T. Newman 253No. 5. The Hodges Site:I. Two rock shelters near Tucumcari, New Mexico, by Herbert W. Dick. 267II. Geology of the Hodges Site, Quay County, New Mexico, bySheldon Judson 285No. 6. The Rembert Mounds, Elbert County, Georgia, by Joseph R.Caldwell 303Appendix. List of River Basin Surveys reports published in other series. _ 321Index 323ILLUSTRATIONSPLATES{Wedel) 1. Aerial view of Medicine Creek Dam site, looking upstream 602. Aerial view of Medicine Creek Dam site (construction area at top),looking downstream 603. a, Aerial view of excavations at 25FT17, Medicine Creek, Nebr. b,Excavating a prehistoric house site at 25FT17, Medicine Creek,Nebr 604. a, View southeast across excavations at 25FT70, Medicine Creek,Nebr. b, Preliistoric house sites and other features uncovered at25FT70, Medicine Creek, Nebr 605. a, Pottery vessels as uncovered on floor of prehistoric earth-lodgesite at 25FT17, Medicine Creek, Nebr. b, Excavations at Wood-land site 25FT18, at junction of Lime and Medicine Creeks, Nebr__ 606. o. View south across excavations at 25FT18, Medicine Creek, Nebr.b, Fireplaces, post molds, and miscellaneous small pits at base ofWoodland occupation zone, Medicine Creek, Nebr 60VII VIII BUREAU OF AMERICAN ETHNOLOGY [Bull. 154] 7. a, Aerial view of Angostura Dam (center foreground) and part ofreservoir area beyond, Angostura Reservoir, S. Dak. b, Bulldozerof Utah Construction Co. removing sod at start of archeologicaltests at 39FA10, Angostura Reservoir, S. Dak 608. a, Workman excavating stone-lined fireplace basin at 39FA9, Angos-tura Reservoir, S. Dak. b, Deep hearth or roasting pit containingburnt stones and sooty soil, exposed during road construction; An-gostura Reservoir, S. Dak 609. a, Excavated tipi ring at 39FA13, Angostura Reservoir, S. Dak,b, Refuse beside aboriginal quartzite quarry pits on Flint Hill,39FA49, near head of Angostura Reservoir, S. Dak 6010. a, View west across Area A at the Long site, 39FA65, Angostura Reser-voir, S. Dak. b, Excavations at Long site, 39FA65, Area A, Angos-tura Reservoir, S. Dak 6011. a, Grouped secondary burials as found in Mound 39CH9, formerlysituated in spillway line of Fort Randall Dam, Charles Mix County,S. Dak. b, Buried archeological zone (line of white stones and bone)exposed in wall of ravine cutting into Missouri River terrace at39ST23, about 3 miles below Cheyenne River, Oahe Reservoir area,Stanley County, S. Dak 6012. a, Panorama of Oahe Dam site, about 6 miles above Pierre, S. Dak.,looking northwest up the Missouri River. 6, Buffalo Pasture site,39ST6, showing fortification ditch (arrows) ; Oahe Reservoir site,S. Dak. c. Site 32GT1, on Heart River, showing exploratorytrenches; Heart Butte Reservoir, N. Dak 6013. a, Site 39FA10, Trench B, showing archeological remains at variouslevels; Angostura Dam in background. Fall River County, S. Dak.b, East test trench at site 39FA30, on Horsehead Creek, AngosturaReservoir, Fall River County, S. Dak 10214. a, Site 24PR2 at edge of sandstone cliff (foreground) ; looking south-east toward Powder River and Bitter Creek. Moorhead Reservoir,Montana-Wyoming, b, Site 39PE10 (on knoll in foreground),looking west up Moreau River Valley. Bixby Reservoir site,Perkins County, S. Dak 10215. o. Chimney Butte from Manderson-Rockyford Road, with low terraceand site 39SH1 (arrow) in middle distance. Rockyford Reservoirsite. Shannon County, S. Dak. b, Site 39ST14 (Scotty PhillipsRanch), a fortified Arikara village site in the work area just belowOahe Dam; Stanley County, S. Dak 102(Kivell)16. a, View to southwest toward Woodruff Ossuary, site 14PH4. b, Look-ing northeast across burial pit, site 14PH4, after the removal ofplowed soil 14217. a, Looking southeast across burial pit, site 14PH4, showing sectionwalls. 6, Discoloration which marked feature 1, a pit, at depth of12 inches, site 14PH4 14218. o. Looking south across burial pit at site 14PH4, after removal of alldisturbed soil, b, Feature 16, site 14PH4, a partially articulatedadult skeleton 14219. a. Feature 8, site 14PH4, an area of scattered human bones showingsome articulation, b, Feature 9, site 14PH4, a pit filled with dis-articulated human bones 142 CONTENTS IXPAGE20. a, Feature 21, site 14PH4, looking south. 6, Feature 21, site 14PH4,after removal to the laboratory in a plaster cast 14221. a, View in east area of burial pit, site 14PH4, showing typical dis-articulated condition of bones, h, Sherds from burial pit, site14PH4 14222. a, Bone and antler artifacts from site 14PH4, h, Stages in manufac-ture of shell disk beads, site 14PH4 14223. a, Chipped stone artifacts from site 14PH4. h, Stone artifacts fromsite 14PH4 14224. Artifacts from site 14PH4 14225. Marine and fresh-water shell ornaments, site 14PH4 14226. a, Artifacts from various sites. 6, Sherds from Guide Rock Ossuary(site 25WT3) 14227. Restored vessel of the Harlan Cord Roughened type from site 25FR8-_ 14228. Four views of male skull (14PH4r-1436) from Woodruff Ossuary 142{Wheat)Doering site and profile 252Kobs site and view of excavation showing Burial 6 252Pottery, restored vessels 252Pottery, decorated sherds 252Pottery, decorated sherds 252Projectile points: Perdiz Pointed Stem, Scalhorn Stemmed, EddyStemmed 25235. Projectile points: Alba Barbed, Kobs Triangular, and ProvisionalTypes Nos. 1, 2, 3, and 4 25236. Projectile points: Gary Stemmed points 25237. Projectile points: Provisional Types Nos. 5, 6, 7, 8, 9, 10, 11, 12, 13,14, and Wells Contracting Stem 25238. Projectile points: Provisional Types Nos. 15, 16, 17, 18, 19, 20, 21,22, 23, 24; Pedernales Indented Base, and Clovis Fhited (?) 25239. Projectile points, unplaced chronologically: Provisional Types Nos.25, 26, 27, 28, 29, 30, 31, 32; Copena point, Plainview point, andBaird Beveled point 25240. Knives: Types Nos. lo, lb, Ic, Id, 2, and 3 25241. Knives: Types Nos. 4, 5, 6a, and 6& 25242. Scrapers 25243. Choppers 25244. Drills, gravers, and eccentric flint 25245. Ground stone: Atlatl weight, sandstone abraders and saws, millingstone, and handstone 25246. Miscellaneous minerals and paint pigments 25247. Artifacts of bone, antler, and shell 252{Dick)48. a, General view of the Hodges site. 6, Area A at the beginning ofexcavation 28449. a, Artifacts eroding out of the No. 2/3 fill under the rock fall in theshelter of Area A. h, Stratigraphy of section A-4, Area A 28450. Trench in the remaining fill of Area B 28451. Projectile points, drills, refined knives, and snub-nose scraper 28452. Scrapers and rough flake knives 284 X BUREAU OF AMERICAN ETHNOLOGY [Bull. 154]PAGE53. Manos, hammerstone, and choppers 28454. Fresh-water shells, Unimerus tetralasmus Say 284(Caldwell)55. a, Aerial view of the Rembert Mounds location and the accompanyingvillage site. The outlined area indicates the mounds, and thelocation of the test pits is shown by the black dots in the field below.b, Pottery pipes, stone disks, projectile point, and shell beads 32056. Potsherd types 320FIGURES(Wedel)1. Map of Missouri River Basin (heavy broken line), showing reservoirprojects investigated by the Missouri River Basin Survey and co-operating agencies as of December 31, 1948 4(Kivett)2. Plan of Woodruff Ossuary after excavation 1103. North-south profile across Woodruff Ossuary on El line 118(Wheat)4. Map of general area 1535. Map of site distribution with inset showing relation of Addicks Dam,Barker Dam, and Houston, Tex 1636. Sketch map, Addicks "Mound" (42/66A6-1) 1647. Sketch map, Doering site (42/66A6-2) 1688. Contour map and plan of excavation, Doering site (42/66A6-2) 1699. Profiles, Doering site (42/66A6-2) 17010. Burials, Doering site (42/66A6-2) _ _ 1721 1 . Sketch map, site 42/66A6-4, showing test pits 17512. Stratigraphic section, site 42/66A6-4 17513. Sketch map, Kobs site (42/66A6-3) 17614. Contour map and plan of excavation, Kobs site (42/66A6-3) 17715. Profiles, Kobs site (42/66A6-3) 17916. Burials, Kobs site (42/66A6-3) 18017. Sketch map, Grisbee site (42/66A3-1) 18218. Stratigraphic section, Grisbee site (42/66A3-1) 18319. Pottery : Rim and lip profiles 18620. Pottery: Vessel shapes 18721. Pottery: Base shapes 18822. Graph showing comparative projectile-point stratigraphy of Doeringand Kobs sites 21523. Relative chronology of the Addicks Basin sites 243(Newman)24. Stereograph contour of the skull from Burial 2 25825. Stereograph contour of the skull from Burial 4 26026. Stereograph contour of the skull from Burial 5 26227. Stereograph contour of the skull from Burial 6 26228. Stereograph contour of the skull from Burial 7 263 CONTENTS XI{Dick) PAGE29. Stratigraphy of the trench excavated in Area B 27230. Plan of Area A showing excavated portion and stratigraphy 273(Judson)31. Index map showing location of the Hodges and San Jon sites 29032. Block diagram to show the salient features of the Hodges site 29133. Generalized section to show the relation of the two erosion surfaces,the inner canyons, and the alluvial formations 29334. Bar graph to illustrate the relative duration of the various events inthe "Alluvial Chronology" 29435. Diagram to show the relation of the sand (No. 2/3 sand) containing theculture horizon to the other deposits at the Hodges site 296(Caldwell)36. C. C. Jones' map of the Rembert mounds 30837. Map showing test pits and wash-outs in the vicinity of the largemound remnant 31138. Profile of the large mound, adapted from Thomas 31239. Section through large mound, adapted from Thomas 31240. Profiles of test pits in the large mound and vicinity 313 FOREWORDThe Inter-Agency program for the recovery of archeological andpaleontological remains which would otherwise be lost as a result ofthe numerous projects for flood control, irrigation, hydroelectricinstallations, and navigation improvements in the river basins of theUnited States got under way in 1946 as a cooperative effort on thepart of the Smithsonian Institution, the National Park Service, theBureau of Reclamation, and the Corps of Engineers of the UnitedStates Army. Preliminary steps taken over a period of many monthsby the Smithsonian Institution and the National Park Service re-sulted at that time in allotments of funds from the Bureau of Reclama-tion and the Corps of Engineers to initiate preliminary surveys inthe Missouri Basin and other areas throughout the country. Sub-sequently all of the funds for carrying on the work have been providedby the Department of the Interior through its Bureau of Reclamationand National Park Service, although the Corps of Engineers has con-tinued to cooperate in other ways. The National Park Service hasserved as the liaison between the various agencies and has providedthe Smithsonian Institution with all of the necessary information per-taining to the locations of the proposed dams and reservoirs and theirpriorities. It also has carried the responsibility for budgeting thecosts of the program.Through a Memorandum of Understanding between the Smith-sonian Institution and the National Park Service, the SmithsonianInstitution assumed responsibility for the actual field investigations.To that end the River Basin Surveys was established as a unit of theBureau of American Ethnology and has operated both survey andexcavation parties throughout the country. In carrying on this workmany local and State institutions have given valuable cooperationand have assumed responsibility for various projects.Preliminary surveys are made in the areas to be inundated or other-wise disturbed by the construction projects. All archeological sitesfound during the course of such surveys are recorded and those thatappear worthy of further study or that cannot be duplicated elsewhereare recommended for excavation. Excavations are carried on at proj-ects where construction is already under way or is soon to start. Thesurveys have located a wide variety of sites ranging from those oc-cupied by primitive hunting groups thousands of years ago through XIV BUREAU OF AMERICAN ETHNOLOGY [Bull. 154] subsequent stages of aboriginal cultural development to villages in-habited by Indians at the time the first exploring parties reachedthose areas. The surveys also have located historic sites of the earlysettlement periods in various parts of the United States and they,as well as aboriginal locations, are being investigated. The historicphase of the program is in conjunction with the work of the HistoryDivision of the National Park Service.Since July 19-16, when the first reconnaissance parties went intothe field, 213 reservoir areas located in 28 States have been surveyed.Archeological sites located and recorded have reached a total of 2,350,of which some 500 have been recommended for excavation or additionaltesting. Excavations have been carried on by the River Basin Sur-veys at 36 sites in 15 reservoir areas in 9 States. Preliminary reportsand summary articles pertaining to the results have appeared invarious journals. More detailed reports are now being completedand those contained in the present volume are the first of what shouldbe a long series of papers. Frank H. H. Roberts, Jr.,Director, River Basin Surveys. SMITHSONIAN INSTITUTIONBureau of American EthnologyBulletin 154 River Basin Surveys Papers, No. 1Prehistory and the Missouri Valley Development ProgramSummary Report on theMissouri River Basin Archeological Survey in 1948By WALDO R. WEDEL XV 953842?53 1 CONTENTS PAGEIntroduction 1Personnel 8Laboratory activities 8Field work and explorations HNebraska 12North Dakota 19South Dakota 20Wyoming 29Field work in paleontology 31Field work by cooperating agencies 33Kansas 34Nebraska 38Archeological reconnaissance in the Lower Platte Basin 43North Dakota 46R6sum6 and conclusions 48ILLUSTRATIONSPLATES 1. Aerial view of Medicine Creek dam site, looking upstream 602. Aerial view of Medicine Creek dam site (construction area at top),looking downstream 603. a, Aerial view of excavations at 25FT17, Medicine Creek, Nebr. b,Excavating a prehistoric house site at 25FT17, Medicine Creek, Nebr_4. a, View southeast across excavations at 25FT70, Medicine Creek, Nebr.b, Prehistoric house sites and other features uncovered at 25FT70,Medicine Creek, Nebr 605. a, Pottery vessels as uncovered on floor of prehistoric earth-lodge siteat 25FT17, Medicine Creek, Nebr. b, Excavations at Woodlandsite 25FT18, at junction of Lime and Medicine Creeks, Nebr 606. a, View south across excavations at 25FT18, Medicine Creek, Nebr.b, Fireplaces, post molds, and miscellaneous small pits at base ofWoodland occupation zone, Medicine Creek, Nebr 607. a. Aerial view of Angostura dam (center foreground) and part of reser-voir area beyond, Angostura Reservoir, S. Dak. b, Bulldozer ofUtah Construction Co. removing sod at start of archeologicaltests at 39FA10, Angostura Reservoir, S. Dak 608. a, Workmen excavating stone-lined fireplace basin at 39FA9, Angos-tura Reservoir, S. Dak. b, Deep hearth or roasting pit contain-ing burnt stones and sooty soil, exposed during road construction;Angostura Reservoir, S. Dak 609. a, Excavated tipi ring at 39FA13, Angostura Reservoir, S. Dak.6, Refuse beside aboriginal quartzite quarry pits on Flint Hill,39FA49, near head of Angostura Reservoir, S. Dak 6010. a, View west across Area A, at the Long Site 39FA65, AngosturaReservoir, S. Dak. b, Excavations at Long Site, 39FA65, AreaA, Angostura Reservoir, S. Dak 60xvn XVIII BUREAU OF AMERICAN ETHNOLOGY 11. a, Grouped secondary burials as found in Mound 39CH9, formerlysituated in spillway line of Fort Randall Dam, Charles MixCounty, S. Dak. b, Buried archeological zone (line of whitestones and bone) exposed in wall of ravine cutting into MissouriRiver terrace at 39ST23, about 3 miles below Cheyenne River,Oahe Reservoir area, Stanley County, S. Dak 6012. a, Panorama of Oahe dam site, about 6 miles above Pierre, S. Dak.,looking northwest up the Missouri River. 6, Buffalo Pasture site,39ST6, showing fortification ditch (arrows) ; Oahe Reservoir site,S. Dak. c, Site 32GT1, on Heart River, showing exploratorytrenches; Heart Butte Reservoir, N. Dak 60FIGURE 1. Map of Missouri River Basin (heavy broken line), showing reservoirprojects investigated by the Missouri River Basin Survey and co-operating agencies as of December 31, 1948 4 PREHISTORY AND THE MISSOURI VALLEYDEVELOPMENT PROGRAMSUMMARY REPOET ON THE MISSOURI RIVER BASINARCHEOLOGICAL SURVEY IN 1948 By Waldo R. WedelINTRODUCTIONThe Missouri River Basin Survey of the Smithsonian Institution,organized in 1946, continued during calendar year 1948 its archeologi-cal and paleontological investigations at Federal water-control proj-ects throughout the watershed of the Missouri. The present report,third in a continuing series, briefly reviews the year's activities infield and laboratory. Although primarily concerned with the workof the River Basin Surveys, it includes also summary statements on theresearches of various State agencies cooperating in recovery of scien-tific materials that might otherwise be lost because of the water-controlprogram. As in previous yearly summaries, this one makes no pre-tense at complete and final coverage of accomplishments during theperiod involved. Neither should it be regarded as necessarily repre-senting the opinions and judgments of the staff members and otherson whose findings in field and laboratory it is very largely based. Itis, in short, a statement of progress during calendar year 1948, at theend of 21/^ years of sustained work. To that statement have beenadded certain preliminary generalizations regarding the significanceof the findings in the over-all picture of native human history in theGreat Plains and nearby regions.It is unnecessary to discuss in detail here the general background,organization, and purposes of the Missouri River Basin Survey, sincethese matters have been adequately set forth elsewhere.^ Briefly, theproject is one phase of a Nation-wide archeological and paleontologi-cal salvage program administered by the Smithsonian Institution andactively directed by Dr. F. H. H. Roberts, Jr., Bureau of AmericanEthnology. This program is based on a memorandum of understand- 1 See Smithsonian Misc. Coll., vol. 107, No. 6, Apr. 23, 1947 ; and Amer. Antlq., vol. 12,No. 4, pp. 209-225, AprU 1947. 1 2 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154ing, dated 1945, between the Smithsonian Institution and the Na-tional Park Service, and on a series of interbureau agreements betweenthe National Park Service, the Bureau of Reclamation, and the Corpsof Engineers. Its purpose is to locate, record, and evaluate the arche-ological and paleontological materials that will be affected by eachof more than 100 Federal water-control projects proposed or underconstruction by the Bureau of Reclamation and the Corps of Engi-neers ; to transmit this information to the National Park Service, alongwith such recommendations for further action as seem necessary torecover a representative sample of the scientific data which are aboutto be lost ; and to direct the Federal phases of any subsequent recoverywork undertaken, including limited or comprehensive excavation atkey sites. Financial support is provided by the Bureau of Reclama-tion through the National Park Service.In carrying forward the enormous task set for it, the MissouriRiver Basin Survey has been materially aided by various organiza-tions, agencies, and individuals. Only a few of these can be singledout at this time for special mention. In the Missouri River BasinRecreation Survey office. Region 2, National Park Service, ChiefRecreation Planner Guy D. Edwards and Archeologist J. D. Jenningscontinued their helpful and stimulating interest in all phases of thearcheological recovery program. The Bureau of Reclamation, besidesproviding funds to support the work, has freely supplied maps, engi-neering data, construction programs and schedules, and other ma-terials on request. Furthermore, as is noted elsewhere in this report,the Bureau took the initiative in setting up the highly successfulcooperative excavation program at Medicine Creek, Nebr.?a patternof operation that for efficiency and productiveness might well befollowed at other projects where archeological and paleontologicalmaterials are directly threatened by dam construction. The Corpsof Engineers has continued to provide maps and other project data;and in the Omaha district office, T. E. Huddleston has remained anunofficial but extremely helpful consultant in interpretation of aerialphotographs and in other problems pertaining to pre-excavationarcheological work on the Missouri River in North and South Dakota.Local, district, and regional officials and personnel of these and otheragencies, project engineers, representatives of construction companies,and many private individuals have extended innumerable courtesies toRiver Basin Surveys field parties and to the local Survey office inLincoln. The same is true of State and other non-Federal agencieswith whom contacts were made. Especially to be noted in this regardis the generous cooperation of the University of Nebraska in con-tinuing to provide, in its Laboratory of Anthropology, working spacefor the Survey. The active support and direct interest of the Com-mittee for Recovery of Archeological Remains, representing the pS^^o'S"' ARCHEOLOGICAL survey in 194 8?WEDEL 3arclieological profession and various learned societies, must also beacknowledged.The general procedure by which the investigative work of the RiverBasin Surveys is programmed consists of three phases. The first isreconnaissance, in which the primary objective is to get a workingfamiliarity with proposed or potential reservoir areas and their ar-clieological or paleontological possibilities. The second is intensivesurvey, wherein there is complete coverage of the reservoir area andlimited testing, if possible, of the more promising sites located. Thethird is comprehensive excavation which, depending upon variousfactors, may mean either the complete working out of certain sites orelse the large-scale but not complete clearing of a series of key sites.These three phases, of course, are not mutually exclusive ; and as thebacklog of proposed reservoirs awaiting examination dwindles andadequate funds become available, it may be possible to intensify theinitial survey and then move directly into such excavation as seemswarranted.From its inception in 1946 until the end of 1947, the field activitiesof the Missouri River Basin Survey were restricted mainly to pre-liminary reconnaissance and some follow-up surveys. No funds wereavailable for extended excavations, although a small allotment infiscal year 1948 made possible limited excavations late in 1947 at Boy-sen ,Wyo., and in 1948 at Angostura, S. Dak., and Heart Butte, N. Dak.Early in 1948, at the suggestion of officials of the Bureau of Reclama-tion, a major excavation program was set in motion at Medicine Creek,Nebr. Because this work was of an emergency character, funds orig-inally programmed by the Missouri River Basin Survey for otherprojects were reallocated for supervisory and other costs in connectionwith the operations at Medicine Creek. This unexpected drain onalready inadequate funds was followed in July 1948 by a reducedallotment for fiscal year 1949 and then by a pay increase for classifiedsta ff personnel. Thus, the Missouri River Basin Survey went throughcalendar year 1948 with operating costs substantially increased andfunds available materially reduced in amount. As a result, field workwent forward on a smaller scale than originally planned and a portionof the professional staff could not be sent into the field. Except atMedicine Creek, no comprehensive excavation was possible in 1948.It may be appropriate to note at this point one of the major obstaclesto effective programming of field work on this project. For variousreasons, including seasonal climatic, labor, and other factors, the mostsatisfactory period for arclieological work is between, approximately,April and October. This period begins in the last quarter of an ex-piring fiscal year and runs into or through the first quarter of the next.It represents an interval during much of which there is uncertaintyregarding the amount of funds to be allotted for the new fiscal year 4 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154beginning on July 1, and also as to the date on which those funds willbe available. If enough carry-over funds are held back to insurecompletion of field-work begun not long before the close of one fiscalyear, there is an excellent prospect that the amount of the carry-over will be deducted from the next year's allotment. All of thismeans that annually at the very time when an all-out summer field pro-gram should be in the making, dwindling funds and the uncertaintyregarding their replenishment exert a strong restraining influence onthe planning and carrying out of field work.During calendar year 1948, eight reservoir areas were visited byBiver Basin Surveys field parties in search of archeological remains,and 194 sites were added to the preceding year's total in the Surveysfiles. The University of Nebraska Laboratory of Anthropology in-vestigated five other proposed reservoir areas and located 34 sites.Information on the archeological resources of Davis Creek Reservoirarea was provided by the Nebraska State Historical Society from itsfiles, making unnecessary for the present any field surveys in that lo-cality. By year's end, the Eiver Basin Surveys and cooperating agen-cies had visited, partially or completely surveyed, and appraised asto archeological materials 57 Bureau of Reclamation and 9 Corps ofEngineers projects, and had on record a total of 877 sites of archeologi-cal interest. Projects visited were distributed throughout the Mis-souri Basin as follows : Colorado, 4 ; Kansas, 6 ; Missouri, 1 ; Montana,4 ; Nebraska, 23 ; North Dakota, 11 ; South Dakota, 9 ; Wyoming, 8.Paleontological investigations by the River Basin Surveys during1948 consisted of re-examination and collecting at four reservoir sitespreviously examined ; no new localities were visited. The Universityof Nebraska State Museum, a cooperating institution, re-examined twolocalities previously worked. By the end of 1948, a paleontologicalreconnaissance and collecting had been done by the River Basin Sur-veys at 97 reservoir sites, including 93 Bureau of Reclamation and 4Corps of Engineers projects. These were distributed as follows, byStates: Colorado, 3; Kansas, 7; Montana, 25; Nebraska, 23; NorthDakota, 10; South Dakota, 6; Wyoming, 23.In table 1 below are summarized the reservoir projects investigatedby archeological and paleontological field parties of the MissouriRiver Basin Survey and by cooperating agencies up to December 31,1948. In the pages that follow are briefly described the field opera-tions and some of the findings during calendar year 1948 ; additionaldetails concerning some of these explorations may be found in Ameri-can Antiquity, volume 14, No. 4, part 1, April 1949. Reviews of scien-tific salvage operations in preceding years will be found in thesummary reports of the Missouri Basin Survey for 1946 and 1947,published by the Smithsonian Institution. 1 Amherst2 Anchor3 Angosluro4 AnTalope5 Ape>6 Bodwoter7 Boldhill8 Beover City9 Belloood10 Bii11 Bmby12 Blue Horse13 Bonny14 BoysenISBoK Butte16 Brenner I 7 Brewster 1 8 Bndger19 Broncho20 Buffolo Creek21 Bull Creek22 Coiro23 Connonboll24 Conyon Ferry25 Cedor Bluff26 Cherry Creek27 Clork Conyon28 Crosby23 Culb?rte?n3C CusMng31 Oovls Creek32 Deerfleld33 Oe> Lacs34 0evlli Lake35 Dickinson36 Du Noir37 Edgemont38 Enders39 Ericson40 Fort Rondoll41 Garrison42GlendoN 43 Glen Elder 44 liv. Bas. Sur.?ap. No. 1] ARCHEOLOGICAL SURVEY IN 194 8?^WEDELO'WOi-i >OC0>Ort o ??->?M t~ O O O "O CO ....P-iP^Ah .... Ph ...<)Ph a ..Ph .Ph ..'SO?gPh" c mOtJ a t^ ^wOwP5MMffl cawUfq P^S ? o ca3 f^ b b? a5K5o 2 3 Mg 2oo>'o o C 3 i 5>?.^,P3>tf^i^J?f ? pjs ?"0^ aS^T-09 ? u Pg 3 p.S)QoS O (D J? ?. .- t^ ,-ax3 S= <5 ?.5 l?-2=s?pu S'f??Cm t>a P5PP S-9 fe?'bT3 3a>.x??So"2^--3C'OSjSoS~ooOL4k.k.k<33c<3o3c: a>A_ l> 3aaflCO.o3o3(u'a;.2^.2.=!SboCiii.ii 3c ac3a)ja.Sii:<1<< .Ph .Ph .PhPmPh^PhPliPhPhP^Ph .PhPl^OW?P^QSSpHa is ? fls >>" -is o Sno ZCOM (3AS"a* ca . 3 g ?3 o?CDCCE-lpqO ^.Ph o caracomK; PiII ;j ? Ucop; ? O o rt5?t? a o oOOPk 2 ftSF4Ph 'OT3 o o? o o o 3? >,o.?>? t^t*l^J^t>.t^t^00 0000 00 00 00 0000O0 00^O3OiO?OSOiO5O5O>O5?-(?-ti-lT-lf-l?-li-l 8 BUREAU OF AMERICAN ETHNOLOGY TBull. 154PERSONNEL.No changes were made in the professional staff of the Missouri RiverBasin Survey during 1948, but owing to shortage of funds and in-creased operating costs, steps were being taken at year's end for a staffreduction. On active duty throughout the year were six archeologists : Paul L. Cooper, temporarily in charge of the field office during thewriter's absences in Washington ; Robert B. Gumming, Jr., laboratorysupervisor ; Wesley L. Bliss ; Marvin F. Kivett ; J. J. Bauxar ; and JackT. Hughes. As in the preceding year, Dr. T. E. Wliite, on leave ofabsence from the Museum of Comparative Zoology, Harvard Univer-sity, carried on the paleontological phases of the Missouri River BasinSurvey salvage work. He was in Lincoln, or on field assignment out ofLincoln, from May 18 until November 5.Full-time office and laboratory personnel included Mrs. Ina MayReagan, clerk-stenographer ; Dean Clark and A. E. Nixon, laboratoryassistants; J. M. Shippee, expert laborer; and George Metcalf, fieldand laboratory assistant. One full-time and one part-time temporaryoffice assistant were added to the clerical force during the year. Draft-ing of maps and diagrams, some of the routine processing of specimens,and the maintenance of certain office and laboratory files were carriedon with part-time student and other assistance.Student assistants were again employed wherever possible as mem-bers of the field parties. Robert L. Hall and Warren Wittry, Uni-versity of Wisconsin, joined Mr. Cooper at Heart Butte on June 22, andremained with the party until early September. Gordon F. McKenzie,University of Nebraska, was with this unit from August 1 to Septem-ber 3. Dr. White was assisted in Wyoming and Montana by ErnestLundelius, University of Texas, from June 1 to September 3, and byJohn C. Donohoe, Montana State College, from June 14 to August 1.IMiss Dorothy E. Eraser participated as special consultant in fieldinvestigations in the Fort Bennett, S. Dak., locality with the Cooperparty during the month of August.LABORATORY ACTIVITIESThe field headquarters and laboratory for the IMissouri River BasinSurvey were located throughout the year in the Laboratory of Anthro-pology at the University of Nebraska. Until the end of September,space was provided for these facilities in the basement of Love Memo-rial Library. Thereafter, more spacious and better-lighted quartersdesigned primarily for anthropological research and training weremade available in the basement of the newly completed Burnett Hall.As in the preceding year, storage space at the Lincoln Municipal Air-port was utilized for housing the Survey's 11 vehicles, field equipment pS-^o-'fr' ARCHEOLOGICAL SURVEY IN 1948?WEDEL 9and supplies, and certain specimens. As previously, also, additionallaboratory and study space was generously provided for Survey per-sonnel by the Nebraska State Historical Society.The basic project files were maintained and expanded throughoutthe year. At year's end, the basic site file included 635 site folders,each containing a site survey sheet prepared in the field, a catalog ofartifacts collected and photographs made, and other data relevant tothe particular location. This represents an increase of 192 over thetotal at the end of 1947. All of the material in the basic site file hasalso been duplicated in reserve files.The map file, which includes topographic, geologic, land survey,ownership, and other maps needed for reference and planning pur-poses, was augmented by 101 additional sheets during the year. Itnow includes approximately 667 maps. Many of these were furnishedby the construction agencies and by other State and Federal offices, asrequested from time to time ; others were purchased as needed. Thereare also 295 aerial photographs, most of them on a scale of 1 : 12000.They cover much of the main-stem area along the Missouri Riverbetween Pierre, S. Dak., and the North Dakota State line, as well assmaller areas on Medicine Creek, about Angostura, and elsewhere.Drafting and map-making have been pushed steadily, but almostwholly with part-time student assistance. Archeological site maps,reservoir site location charts, archeological and geological profiles,and house plans to the number of 24 had been prepared by the endof December.The addition of a full-time laboratory assistant in the darkroomin June has vastly facilitated the preparation of specimen photo-graphs, the processing of field and laboratory negatives, and thephotographic duplication of office and laboratory records. Darkroomwork during calendar 1948 includes the processing of 1,087 negativesof field and laboratory photographs; preparation of 1,562 black andwhite prints for reference files and illustrative purposes ; preparationof 121 2x2 black-and-white slides ; cataloging and filing of 373 colortransparencies, most of them in 35-mm. size; and the preparation of262 photographic enlargements in 8 x 10-inch size for publicity andreference uses.An ethnohistorical reference file begun late in 1947 was continuedthrough 1948. At year's end this included 5,202 cards carrying classi-fied information on the early documentary sources for the MissouriBasin Indians. Emphasis was on the native tribes residing formerlyalong the main stem and the references to them by European andAmerican explorers up to the beginning of the nineteenth century.With the beginning of comprehensive excavation, the handling andstorage of specimens has become a major activity. By the end of1948, nearly 100,000 specimens had been cleaned, cataloged, and stored. 10 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154Slightly more than half of these were accumulated during the cal-endar year. The restoration of outstanding or representative piecesand their preparation for eventual study and exhibit has gone forwardon a small scale. It should be noted that the great bulk of specimensnow on hand are primarily of study interest; they include muchbroken and fragmentary material which will be discarded when nolonger needed for analysis and report writing. Under present labora-tory procedures, all materials sent in from the field by archeologicalparties are processed and cataloged, so as to facilitate their study.Only a relatively small fraction of the collections now on hand willeventually find their way into National, State, and local institutions.As in the previous year, the Missouri River Basin Survey furnishedto the National Park Service preliminary mimeogi-aphed appraisals ofthe archeological and paleontological resources of reservoirs investi-gated. For those reservoir areas where intensive survey was subse-quently made, supplementary appraisals have been furnished. Theseappraisals consist of site location maps, summary descriptions ofmaterials seen, and an evaluation of the probable scientific importanceof the remains located. Where further investigation is believed tobe needed, an estimate of the nature and extent of such additionalwork is also included. These reports are distributed by the NationalPark Service to the construction agencies, to district and regionaloffices of the various Federal agencies concerned, and on a selectivebasis to cooperating State and other institutions and organizationswhere their particular fields of interest are concerned.During calendar year 1948, preliminary appraisals on archeologicalresources of 11 reservoir sites were prepared and distributed. Theseinclude Baldhill and Garrison in North Dakota ; Beaver City, Bonny,Buffalo Creek, Culbertson, Norton, Pioneer, Red Willow, and RockCreek in the Republican River Basin of Colorado, Kansas, and Ne-braska ; and Wilson Reservoir site in the Smoky Hill Basin of north-ern Kansas. Supplementary appraisals were distributed for Boysen,Glendo, and Oregon Basin, in Wyoming; and for Canyon Ferry andTiber, in Montana. In addition, a memorandum on Cheyenne RiverBasin archeology, including Angostura, Edgemont, Johnson Siding,Keyhole, and Pactola Reservoir sites in southwestern South Dakotaand northeastern Wyoming, was furnished the National Park Service.In preparation was a preliminary appraisal of the Davis CreekReservoir site, Nebraska, based on archeological work by the NebraskaState Historical Society and Works Project Administration in 1939;and several longer reports of more technical nature. These latter in-cluded a report on certain discoveries in the fall of 1946 at HarlanCounty Reservoir site; and one on Birdshead Cave in the BoysenReservoir site, Wyoming. Late in the year, members of the field staff pLp-N^-'fr' ARCHEOLOGICAL SURVEY IN 194 8?WEDEL 11were beginning to turn their attention to the preparation of technicalsummaries on the 1948 field work at Medicine Creek, Angostura, andlower Oahe Reservoir sites. These summaries were published in theApril 1949 issue of American Antiquity, which is devoted entirelyto papers dealing with the Missouri Basin work of the River BasinSurveys and cooperating State agencies.In addition to the various archeological reports noted above, a71-page appraisal of the paleontological resources of certain riverbasins and reservoir sites in the Missouri watershed was also dis-tributed. This covers 70 reservoir sites, potential, proposed, andunder construction, in Nebraska, North and South Dakota, Montana,and Wyoming. Not distributed were reports prepared by thepaleontologist on Boysen, Canyon Ferry, and Cedar Bluflf Reservoirsites, and a preliminary statement on the physiographic history ofHorsehead Creek in the vicinity of Site 39FA65, in AngosturaReservoir, South Dakota.In November, the Smithsonian Institution published a summaryreport prepared by the field director on the activities of the MissouriRiver Basin Survey during calendar year 1947.As in preceding years, the Lincoln office of the River Basin Surveysfurnished information, as requested, to the daily press, to technicaland semipopular periodicals, and to various other agencies and in-dividuals. In the field, archeologists at various times addressed localcivic and other groups interested in the prehistory of their particularlocalities. Especially gratifying in this connection was the sus-tained public interest shown in the excavations at Medicine Creek,where illustrated talks by the archeologist in charge became a semi-weekly part of his duties.Several staff members participated in various ways in the sessionsof the Sixth Plains Archeological Conference, held at the Universityof Nebraska, in Lincoln, November 25-27, 1948.FIELD WORK AND EXPLORATIONSDuring calendar year 1948, field work was carried on by five unitsrepresenting the River Basin Surveys. Four of these were concernedwith archeological researches, the fifth with paleontology. Arche-ological projects included the following : comprehensive archeologicalexcavations for 4% months at Medicine Creek Reservoir site, underthe direction of M. F. Kivett; 31^ months of reconnaissance and in-tensive survey at five reservoir sites in the Cheyenne River Basin innortheastern Wyoming and southwestern South Dakota, underJack T. Hughes ; 6 weeks of test excavations at Heart Butte, N. Dak.,and a like period in the lower Oahe Reservoir site. South Dakota, 12 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154under Paul L. Cooper; and 2 weeks of mound excavation in thevicinity of Fort Randall Dam, South Dakota, also by Mr. Cooper.The paleontological unit, under Dr. T. E. ^^Hiite, divided 4 months offield activities among Boysen (Wyo.), Canyon Ferry (Mont.), Angos-tura (S. Dak.), and Cedar Bluff, (Kans.) Reservoir sites. Furtherdetails regarding these various projects are given below.Another staff archeologist, J. J. Banxar, was detailed at requestof the Regional Director, Region 2, to the National Park Service fromApril 5 to June 26 for archeological researches at Homestead Na-tional Monument, Gage County, Nebr. W. L. Bliss spent approxi-mately one week in August assisting in archeological excavations bythe University of Nebraska State Museum in Medicine CreekReservoir, Nebraska. NEBRASKAMedicine Creek Reservoir site.?Of outstanding interest in theyear's program were the comprehensive archeological excavationsconducted during the summer at Medicine Creek Reservoir site insouthern Nebraska. A preliminary reconnaissance here by M. F.Kivett and J. M. Shippee during the summer of 1946 had disclosedthe location of 15 archeological sites within the future pool area. In1947, following the disastrous flood of June 21-22 on Medicine Creek,Bureau of Reclamation plans for construction of the dam and re-servoir were speeded up. Accordingly, the River Basin Surveysassigned Mr. Kivett, with George Metcalf as assistant, to approxi-mately 2 months of further survey and limited excavations at andnear the dam site, at points that seemed likely to be affected soon byconstruction activities. This work, a continuation of investigationsbegun some time previously by the Nebraska State Historical Society,disclosed the presence of at least 20 additional localities of archeologi-cal interest here and strongly indicated the advisability of more ex-tended excavation. Findings by the field parties working here in 1947have been briefly summarized in a previous report.In early spring of 1948, a conference was arranged in Denver, Colo.,in which representatives of the Bureau of Reclamation, the NationalPark Service, and the Missouri River Basin Survey participated. TheBureau of Reclamation generously offered to furnish power ma-chinery and hand labor for the salvage of materials to be affectedby reservoir construction, provided the River Basin Surveys wouldassume responsibility for the technical supervision of the work. Theoffer was promptly accepted, and the arrangements were embodiedin a memorandum from the Regional Director, Region 7, Bureau ofReclamation, dated February 6, 1948. Under this arrangement,archeological excavations were carried on from March 29 until August pS.?fo^.'lT'^' ARCHEOLOGICAL SURVEY IN 1948?WEDEL 1320, 1948. The Bureau of Reclamation made available the services of15 to 20 laborers, a road patrol, and, for part of the time, a bulldozer,besides providing surveyors as their services were requested from timeto time. From the Lincoln office of the River Basin Surveys, M. F.Kivett was detailed to supervise the operations and to insure thatthe work would be carried on in accord with recognized archeologicalprocedures and standards. George Metcalf again was assigned asfield assistant.The Medicine Creek project has been outstandingly successful inevery particular. Uniformly cordial and pleasant relationshipsexisted at all times between River Basin Surveys personnel and thevarious representatives of the project, district, and regional officesof the Bureau of Reclamation. No better cooperation could havebeen asked than that extended by the latter agency in this matter.^The division of responsibility under which the work was conductedby the Federal agencies permitted the archeologists to devote theirfull time to the scientific and technical aspects of the operation. Theapplicability of power machinery under close and constant controlto the excavation of village and camp sites, not previously attemptedon a comparable scale in the Missouri Basin, was convincinglydemonstrated.Medicine Creek Dam, now well along in construction, is located 81/2miles north of Cambridge, Nebr., in the southeast corner of FrontierCounty. It is to be an earth-fill structure with a crest length of ap-proximately 5,665 feet and a maximum height of about 115 feet abovestream bed. Provision will be made for a maximum high-water levelof 2,408.9 feet. At normal pool (elevation 2,366 feet), the reservoirAvill be 81/2 miles long, with a shoreline of about 31 miles. It will con-trol a drainage area of some 656 square miles. Purposes of the dam,a unit of the Frenchman-Cambridge project, are flood control andirrigation.Between March 29 and August 20, comprehensive excavations weremade in five sites, with more limited investigations in three others.With the mechanical aids available, entire sites were stripped of theirsod and overlying soil cover, so that the complete village layout couldbe determined and mapped (pis. 1, 2). The actual clearing of housefloors and storage pits, and the closer investigation of remains uncov-ered by the preliminary stripping, was done, of course, with handtools. This technique, which made possible the accurate determina-tion of native settlement patterns and the discovery of numerous smallfeatures not readily determinable by the usual hand-tool methods, isone of the great advantages of mechanized archeology. Remains ' Special acknowledgment should be made of the helpfulness of C. L. Mutch, projectengineer ; H. B. Robinson, District Manager, McCook County ; and A. A. Batson, regionaldirector.953842?53 2 14 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154 uncovered include 28 prehistoric lodge sites, 75 storage pits, 22 refusedeposits, several burials, and more than 31,000 cataloged objects ofpottery, stone, bone, shell, and other materials.Work began on March 29 at site 25FT17, atop a high ridge (pi. 1)that will be incorporated into the west half of the dam structure.Here the major part of one site was excavated, including 8 subrectan-gular house floors (pi. 3 a, 5), 21 storage pits beneath house floors,4 outside storage pits, and 5 midden or refuse areas. From April 26to May 5, the work was extended to 25FT1G, a short distance to thewest and close to the western terminus of the future dam. Here theNebraska State Historical Society had previously excavated two housefloors in 1947. A third lodge floor was cleared in 1948 by the RiverBasin Surveys, as were eight storage pits and three refuse areas.Practically all of these materials are assignable to the Upper Repub-lican culture, although several pits and a few sherds suggest that anolder Woodland occupation preceded the Upper Republican at25FT17.From May 6 until June 14, excavations were carried on at two UpperRepublican sites located in the borrow area on the left (east) bank ofMedicine Creek, a short distance above the dam site. At site 25FT13,8 rectangular house floors were cleared and 17 storage pits were opened,12 of them lying within the house units. Five refuse deposits, eachlying near the entrance of one of the houses, were also excavated.Despite the stripping of large sections of the nearby village area, someof which had undergone considerable erosion, no other archeologicalfeatures were noted. Extensive tests for burials on the bluff over-looking the village site from the east disclosed the remains of but asingle extended adult male Indian; a few accompanying sherds,projectile points, and other objects suggest that this individual mayhave been an inhabitant of this or another nearby village.Site 25FT14, a small Upper Republican village situated some 250yards northwest of the preceding, was also worked partially duringthis period. A single house floor was cleared, and extensive testtrenches were opened. Unfortunately, the larger part of the site hadbeen destroyed by local collectors prior to the present salvage program.On the right bank of Medicine Creek, approximately one-fourthmile above the dam site, was one of the larger village sites excavated inthe reservoir area. This, 25FT70, lay only partially within thefuture pool area, and therefore was not completely cleared. BetweenJune 15 and July 16, nine rectangular house floors were opened here(pi. 4, a, &) ; several of them were marked prior to excavation by slightelevations in the unbroken sod. Among the numerous specimens re-covered from this site, particular interest attaches to the quantities ofcharred corn, mostly shelled, and the lesser amounts of beans, sun- pLp N^fr* ARCHEOLOGICAL SURVEY IN 1948?WEDEL 15flower seeds, and squash seeds taken from storage pits. Several sizablesections of house timbers in an excellent state of preservation will beuseful in determining something of the type of vegetation character-istic of the locality in aboriginal days and may also supply informationtoward an absolute dating for the site and for a part of the UpperRepublican culture, which it mainly represents.One small low-lying area within the limits of 25FT70 yieldedcalcite-tempered cord-roughened pottery, small pits, scattered postmolds, and other artifacts perceptibly different from those in the sur-rounding Upper Republican houses. These, along with the burials oftwo children, are attributed to an earlier Woodland group.At the edge of the bluff near 25FTT0, stepped trenches were cutdown the hillside to check the seeming occurrence of cultural debrisin the lower portion of the terrace fill. These cuts, made in someinstances to a depth of 6 feet, revealed evidence of extensive washingand scattering of cultural refuse from the bluff-top site. Below thesurface refuse mantle and apparently in situ in the fill at the valleyedge, were found scattered bones of a Pleistocene camel and of otherunidentified animal forms. There was no evidence of anything thatmight be construed as Early Man. This area, worked between July17 and 20, was designated Site 25FT29.Several hundred yards upstream and also on the right bank ofMedicine Creek, one subrectangular Upper Republican house site wasopened at Site 25FT36. A second house floor lying above future poollevel was located but not worked. Two midden areas slated for flood-ing were also worked out.Site 25FT18, a Woodland site situated on a low terrace on the leftbank of Lime Creek near its confluence with the Medicine, was com-pletely excavated between August 3 and 20 (pi. 5, &/ 6). Unlike theother and probably later village sites investigated, this one had noearth-lodge floors. Cultural materials and camp refuse occurred in adark, stained layer varying in thickness from 6 to 30 inches andburied beneath 6 to 15 inches of lighter-colored wind-blown deposits.The entire cultural layer and all overlying sterile deposits, involvingapproximately 12,000 cubic feet of soil, were removed. Several fire-places, numerous scattered post molds, and a few small pits or basinlikestructures were cleared, all lying at the bottom and just below the darkculture-bearing stratum from which came most of the artifacts andspecimens (pi. 6, h). The disarticulated remains of two adult burialswere also recovered from the lower portion of the culture stratum.Artifacts from site 25FT18 include calcite-tempered cord-roughenedpotsherds, tubular bone beads, bone awls, small corner-notched projec-tile points, end scrapers, disk beads and rectangular to oval pendantsof shell, and other materials. 16 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154The River Basin Surveys excavations at Medicine Creek in 1948were concerned entirely with remains of pottery-making peoples,some of them, at least, depending partly on maize horticulture. Twocultural horizons appear to be represented?the Upper Republicanand the Woodland. No clear-cut instances of stratification werefound, and there is no direct evidence from the Medicine Creek Reser-voir area itself as to the temporal relationships between these twocomplexes locally. From the broad similarity of the remains here toother culturally related materials found in stratigraphic relationshipelsewhere, however, there seems no good reason to doubt that theWoodland remains represented sparingly at 25FT70, possibly at25FT17, and more fully at 25FT18 were laid down some time beforethe more abundant Upper Republican remains. Kivett has tenta-tively designated the Woodland variant here represented as the KeithFocus.Pending completion of the detailed analyses now under way on the1948 Medicine Creek collections, only a few of the salient findings ortrends of evidence can be noted here. Current views are that theUpper Republican in this locality probably flourished somewherebetween the late thirteenth and late fifteenth centuries. As alreadyindicated, there are timber specimens from some of the Upper Republi-can sites that may yield to dendrochronological analysis and fromwhich the local complex may be datable in terms of the Christiancalendar. Even if that hope fails, it should be possible to determinethe relative building dates of individual houses within a single villageunit, or between groups of houses in different parts of the severalvillages. There are already evident rather definite variations in de-tail in sherd types from the various sites, as well as in sherd samplesfrom different house units of the same village. These variations seemto correlate with other minor but consistent differences in the materialculture inventory, and possibly are due to something more than merefamily or individual tastes.The present excavations have confirmed and greatly amplifiedearlier views regarding the nature of the prehistoric Upper Republi-can occupancy. Subsistence was basically by corn-bean-squash-sun-flower horticulture, with the bison scapula hoe as the characteristicgardening tool. Quantities of mammal and bird bones indicate thathunting was also important, the nearby bison having been heavilydrawn upon. Bone fishhooks and, in some sites, great quantities offreshwater-mussel shells, indicate other sources of food. Under-ground storage pits were a regular adjunct of the dwelling complex,occurring inside as well as outside the houses. They are both smallerand less numerous in relation to the number of house units than thestorage pits of the later Pawnee and other historic horticultural vil- pip.N*o!"r|'"^' ARCHEOLOGICAL SURVEY IN 1948?WEDEL 17lage tribes of the eastern plains, which suggests to me a somewhat lessintensive or less productive food-raising complex. The loose, scat-tered nature of the villages, characterized by rectangular earth-cov-ered dwellings arranged in groups of two to four or more units andseparated by a few rods from another similar cluster, all of them un-fortified (pi. 1), is in striking contrast to the great, compact, oftenditched and palisaded, earth-lodge villages seen among the Pawneeand their neighbors in eastern Nebraska during the sixteenth, seven-teenth, and eighteenth centuries.Judged by the findings at 25FT18 and 25FT70, the Woodland oc-cupancy of the locality was appreciably different from the UpperRepublican. No evidence of domestic crops came to light, and thebone hoe was absent, as it has been so far at other Woodland sitesexcavated or tested in the Nebraska-Kansas region. The bones ofbirds and animals were moderately abundant throughout the culturestrata involved. There is some indication that deer and smaller ani-mals were hunted to a much greater extent than the bison, which, inlater prehistoric and historic times, apparently became the principalmeat source. Pottery was present in much smaller amounts than inthe Upper Republican horizon, and the number of vessels owned perfamily was certainly very considerably lower. Small, well-made,stemmed projectile points suggest use of bow and arrow. There wasno recognizable evidence of fishing. In contrast to the substantialearth lodges of later times, small structures of perishable materialsseem to have been used for habitation. These, presumably, wereerected in or over some of the small basinlike features revealed in theexcavations, and around or near the ash-filled fireplace depressions.Unfortunately, no post-mold configurations recognizable as possiblehouse patterns could be worked out.One of the few disappointments of the season at Medicine Creekwas in the fact that no Upper Republican burial grounds could belocated and worked out. The single extended skeleton on the hilleast of 25FT13 is very likely from the Upper Republican period, butof course is inadequate to show convincingly the physical type andsomatological relationships of the natives who left the great majorityof village sites so far located in and about the future reservoir area.Unless interment was in single isolated or remote gi-aves, or was nota regular method for disposing of the dead among these villagers, itmay be presumed that among the still-unfound antiquities of thelocality there must be a number of burial grounds.For the Woodland horizon, the available evidence is not much moresatisfactory. The two burials at 25FT70 were those of children, andthere is but a single measurable skull from the grave at 25FT18.Since these sites seem to have been generally rather small, associated 18 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154burial grounds of any size are perhaps not to be expected. One istempted to wonder, incidentally, whether the placing of a grave,with single or compound burial, in the camp area by the Woodlandpeoples may have been a factor in their abandonment of the sites.Concerning the external relationships of these early Medicine Creekpeoples and their contacts with contemporary groups, little informa-tion came to light. At none of the sites worked was there any recog-nizable evidence of puebloan pottery or other Southwestern tradematerials. Absent, too, were obsidian, steatite, and other exotic min-erals from the regions to the west, as well as finished products or rawmaterials certainly attributable to peoples of other culture to thesoutheast or east. The general impression is thus one of closely self-centered small communities too busy with their own local affairs toengage extensively in long-distance trading or other ventures.As has been noted, where the sites examined this summer lay inunbroken ground or had not been subjected to long-continued anddeep erosion, they were overlain by varying depths of light wind-blown soil. This overlying mantle, capped with sod, generallyobscured the village horizon, so that potsherds, chipped stone andrejectage, and refuse animal bone could be detected only along theeroded margins. With removal of the overburden, the old occupa-tional level showed up as a dark-stained zone varying in thickness andin the amount of cultural admixture. The house sites had been dug tovarying depths into this old village level ; some houses had been butslightly below the contemporary surface, others had been in pitsup to 15 or 18 inches deep. The implication would seem to be thatat the time the Upper Kepublican peoples inhabited the locality, theterrace and bluff surfaces were at a somewhat lower level and thatmoisture was adequate to maintain a relatively stable vegetationcover. Subsequent to abandonment of the village sites, there has beenincreased soil movement, with no indication that a stable humuslevel was again reached until the present sod developed and a newsoil-forming period set in.This apparent correlation of prehistoric corn-growing peoples withformer soil surfaces now buried by aeolian deposits has been noted atother places in the Republican drainage and elsewhere in the Ne-braska-Kansas region. Some of the sites at Medicine Creek have beeninspected very briefly by soils experts, but more detailed studies willbe necessary before the true significance of this seeming correlationcan be set forth. The prospects seem excellent, however, for eventualdating, through the associated archeological materials, of some of theclimatic fluctuations that have characterized the region for manycenturies past, but within the time of men who attempted to maketheir living through the cultivation of maize and other crops. pip NTri""^' ARCHEOLOGICAL SURVEY IN 1948?WEDEL 19NORTH DAKOTAThe only archeological field work programmed for North Dakotaby the River Basin Surveys during 1948 was limited excavation atthe Heart Butte Reservoir site. Previous investigations here includea brief reconnaissance by Paul L. Cooper and J. J. Bauxar on August15 and 16, 1946, at which time a small part of the proposed reservoirwas examined and local residents were consulted. During the 1947season, a party from the University of North Dakota and the NorthDakota Historical Society, under leadership of Dr. Gordon W. Hewes,spent approximately 1 week in the locality. The 1948 operations hereby the River Basin Surveys were designed to make a final test of thearcheological possibilities of the locality. The party was in chargeof Paul L. Cooper. From June 12 to June 20, he was assisted by ThadC. Hecker, of the North Dakota Historical Society, and the workconsisted mainly of further survey, chiefly on the south side of theriver. From June 21 to August 1, the party consisted of Cooper andtwo student helpers, Warren Wittry and Robert Hall. One locallaborer was employed from July 6 to July 30.Heart Butte Reservoir site.?Heart Butte Dam is under construc-tion on Heart River, in northern Grant County approximately 15miles south of Glen Ullin. It is an earth-fill structure, with a heightof 123 feet and a crest length of 1,850 feet. At full pool (elevation2,084.5 feet) , the reservoir will be about 14 miles long, with a maxi-mum width nowhere mUch exceeding 1 mile. The Heart here fol-lows a general easterly course through a steep-walled, flat-flooredvalley lined with well-developed terraces. Timber is sparse, con-sisting of some juniper on the slopes and deciduous trees?mostlyCottonwood, ash, and chokecherry?on the bottoms. Most of theground, except where under cultivation, is heavily grassed.Following a check of previously unvisited portions of the area,most of the activity between June 21 and August 1, 1948, consistedof excavation at site 32GT1, a pottery-bearing site (pi. 12, c) on thenorth bank of the Heart River about 3 miles above the dam site. Theexcavations totaled some 380 feet of trenches cut into undisturbedportions of the site. Cultural materials were found chieflly at depthsvarying from 6 to 24 inches beneath the present surface. Tliere werelarge quantities of bison bone, most of it badly split and broken;bones of the domestic dog also occurred. Artifacts were not plentiful.Pottery showed simple stamping, with cord-impressed designs, andappeared to be related to late Mandan or Hidatsa wares. House re-mains or other structures were not observed, but small fireplaces oc-curred in some numbers. These were simple affairs?small burnedareas, sometimes basin-shaped, and capped by ashes. 20 BUREAU OF AMERICAN ETHNOLOGY [BuU. 154During the last week in the area, tests were made at a rock shelter,32GT5. This produced some evidences of late, brief, and casual oc-cupancy ; tests on the slope below were mostly negative. Other sitesexamined include two bison kills or hunting camps, five small sitesyielding scattered bits of pottery and other debris, and two smallflint-littered areas that may have been workshops.On the basis of the several examinations made between 1946 and1948, it may be concluded that the known antiquities threatened withdestruction by Heart Butte Dam are not likely to constitute a seriousloss to archeology. Temporary hunting camps, some of them prob-ably belonging to the village tribes resident on the Missouri 50 or 60miles to the east, seem to be the chief remains. The bison kill, 32GT6,on the north bank of the river some 4 miles above the dam site, maybe of some antiquity and would possibly repay partial excavation.The considerable overburden would make this a costly and difficulttask, however, and it seems not improbable that the information to bederived therefrom could be duplicated from other similar sites lyingoutside of areas that will be affected by reservoir construction. Un-less remains not now known are turned up in course of construction.Heart Butte Reservoir site can probably be written off as no greatloss to archeology or paleontology.SOUTH DAKOTAArclieological investigations by the River Basin Surveys were car-ried on at three Federal water-control projects in South Dakotaduring 1948. One party under Jack T. Hughes, with J. M. Shippeeas field assistant, and with local labor from time to time, was in theAngostura Reservoir near Hot Springs, S. Dak., from June 2 untilSeptember 15. On September 27, Hughes and Shippee made a briefreconnaissance of Pactola and Johnson Siding Reservoir sites, onRapid Creek a few miles west of Rapid City, locating one small siteat the latter locality. Another party, led by Paul L. Cooper andincluding Robert Hall, Warren Wittry, and, during the month ofAugust, Gordon F. McKenzie and Miss Dorothy E. Fraser, workedfrom August 1 until September 10 along both banks of the Missouribetween Pierre and the Cheyenne River. This section, with anabundance of native village sites, will be flooded by the proposedOahe Dam to be built a few miles upstream from Pierre. During thelast 2 weeks in November, Cooper and Shippee excavated a burialmound situated on the spillway line of Fort Randall Dam, now underconstruction a few miles north of the Nebraska-South Dakota line.Excepting this latter work, which was done in conjunction with theCorps of Engineers, most of the South Dakota field operations con-sisted of survey and test excavations. RIv.Bas.Sur. ARCHEOLOGICAL SURVEY IN 1948?^WEDEL 21Pap. No. IJAngostura Reservoir site.?Angostura Dam is now under construc-tion (pi. 7, a) on the Cheyenne River, in Fall River County about 9miles south of Hot Springs. Here a concrete dam and earth-fill dikesrising some 130 feet above stream bed and having a crest length ofabout 1,775 feet, will impound a pool 11 miles long by li/^ miles wide.At maximum pool (elevation 3,200 feet) an area of about 4,995 acreswill be under water; at minimum pool (elevation 3,162.5 feet) thepool area will be about 2,690 acres. Purpose of the project is irrigation.The future reservoir will lie in a narrow sheltered belt curvingaround the south and east flank of the Black Hills. To the west andnorth, the terrain is hilly and rough, with coniferous forests on thehigher portions. To the east and south are gently rolling grass-cov-ered hills and plains. Within the reservoir area, the Cheyenne hasonly two tributaries of note. Sheps Canyon enters from the left a fewhundred yards above the dam site ; it contains springs but the channelis badly clogged. Less than 3 miles above the dam, Horsehead Creekjoins the Cheyenne from the right ; it has a valley with broad bottoms,well-developed terraces, and an intermittent flow of water. Most ofthe reservoir area is a wide valley, with a strongly developed series ofterraces along the river. Cottonwood formerly stood along the streambanks, with stands of yellow pine in the ridges ; the greater part wasgrass-covered.Including the findings during a brief reconnaissance in 1946, a totalof 71 archeological sites is now on record for the locality in and nearAngostura Reservoir. The gi-eat majority of these are sites that willbe destroyed by the construction of the reservoir and associated works.None show evidence of any single long-time occupancy ; frequent andrepeated use of the locality by various peoples over a long period oftime seems rather to be indicated.Three sites from which small sherd samples were collected suggestrelationships on at least two time levels with pottery-making culturesto the east. One of these sites, lying on Horsehead Creek, shows cord-roughened ware with simple unthickened vessel lips bearing diagonalimpressions. These sherds were found in thin, shallow, layered refusedeposits, along with quantities of broken bison and other large bones.A hunting camp may be indicated, possibly one used periodically bysmall groups of Indians residing normally farther to the east or south.Some relationships may be suspected with a late prehistoric pottery-making group, such as the Upper Republican peoples of the Loup-Platte-Republican region.At two other sites pottery of a different sort was found. This warewas thin and well made ; the surfaces commonly bore simple stampingor, in several instances, incised decoration. It is reminiscent of potterypreviously found at sites in Ericson, Mullen, and Harlan County Reser- 22 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154 voir areas in Nebraska, in northeastern Nebraska, and on SundanceCreek in Crook County, Wyo. It is associated at Angostura withsmall triangular arrow points, planoconvex scrapers, chipped knives,etc. ; no habitation sites have yet been identified, although there is sometendency for the material to occur in spots suggesting midden orformer lodge sites. No white trade goods were found in association,though the horizon represented is generally supposed to be very lateprehistoric or protohistoric.Among the far more numerous nonpottery sites, there seem to beseveral variant complexes. Owing to the extremely limited excava-tions to date, however, they cannot as yet be satisfactorily defined.They include numerous camp sites situated on the ridges, terraces, andbluffs along the river. Some are covered by a few inches to several feetof wind-blown sand or fiuviatile silts. Chipped stone, flakes, rejectage,bone fragments, and similar refuse occurs in varying amounts. Somesites have rock-bordered fireplaces ; at others there are rock-lined firebasins (pi. 8, a) ; still others show rock-filled pits. In a number ofinstances, grinding implements have been found about these hearths.Projectile points, like the hearths and some other remains, showappreciable variation from site to site, and may have real diagnosticvalue for the establishment of cultural units.Tipi rings were recorded at only five locations within the reservoirarea, but are said to be very common elsewhere in the locality. Aselsewhere, so here there were few artifacts in association. At one site ( 39FA13) , a tipi ring directly overlay a stone-filled hearth from whichit was separated vertically by several inches of soils (pi. 9, a). It isbelieved that most of these rings are probably among the more recentremains of the region.Of more than passing interest are several sites in the lower part ofthe reservoir area where roasting pits occur (pi. 8, h). These areabout 21/2 feet in diameter by 3 feet deep, narrowing somewhat in theupper portions. Charcoal and fire-cracked stones occur in the bottom,above which is dark, sooty soil. The pit walls are usually hardenedand fire-reddened. Thin layers of clean sand in the lower part of thefill suggest periodic reuse. Associated with these pits are smallside-notched points, planoconvex scrapers, mealing slabs and hand-stones, and other items. In one or two instances, the pits are locallyreported to have been used for burial purposes, with bone awls, largeand small tubular bone beads, and other objects accompanying thedead. No pottery was found on these sites, although in southernNebraska and western Kansas somewhat similar pits occur in sitesof the pottery-using semihorticultural Dismal River peoples (PlainsApache) of the late seventeenth and early eighteenth centuries.Noteworthy, too, is a camp site (39FA65) lying some 3 miles upHorsehead Creek, on its right bank. This site, portions of which will pS N^fl"'^' ARCHEOLOGICAL SURVEY IN 1948?WEDEL 23certainly be inundated by the reservoir pool, lies at the edge of asmall "badlands" area (pi. 10, a). From the eroded margin of thesite came a large Plainview type projectile point and several basalfragments of lanceolate points with narrowed straight to concavebases, fine flaking, and ground proximal edges. Lanceolate point frag-ments were also recovered in excavations (pi. 10, 5), which furtherdisclosed small circular fireplaces without stones, and considerableamounts of minute, paper-thin flakes from the manufacture of chippedartifacts. Strangely enough, in the work done here to date, virtuallyno animal bones were encountered. Other artifacts include chalce-dony plate knives, medium to large bifacial blades, end and sidescrapers, a gougelike fragment somewhat reminiscent of the ClearFork, Tex., gouges, and a subrectangular mano. Some of the bladessuggest affinities with the finely made implements recovered by Robertsat the Agate Basin bison kill in Wyoming a few miles to the west, andwith specimens found by the University of Nebraska State Museumon Hat Creek and White River, in Nebraska, a few miles to the south.It is believed that the site may have considerable antiquity, at anyrate as compared with most of the others recorded in the district, butmuch more work and more intensive geological studies will be needed.Much of the chipped stone and rejectage found at sites in Angosturaevidently was gathered by the natives at aboriginal quarries in theregion. None of these occurs within the future pool area, but one ofthe largest lies about 6 miles airline northwest of the head of thefuture reservoir. Shown on the United States Geological Survey'sEdgemont quadrangle sheet as "Flint Hill," this is at or near the headof Hell Canyon. Scores of large craterlike pits and piles of roughworkshop debris (pi. 9, &) litter nearly a half section of high table-land. Gray, purple, brown, red, and yellow quartzite were obtainedhere from the Cretaceous Dakota formation. Numerous artifacts ingreat variety of size and form have been collected in and about thesepits during many years by their owner, Mr. Neal Conboy, who gen-erously showed members of the Survey party over the ancient diggings.On lower benches and slopes nearby are to be seen many tipi rings;others to the number of many hundreds are said to be scattered forsome miles northward from the quarries. Additional aboriginaldiggings occur in the area, notably at Battle Mountain near HotSprings. The general appearance of those observed was very similarto the better-known and more extensive Spanish Diggings some 70miles to the southwest in Wyoming.Within historic times?that is, after about the first quarter of thenineteenth century?the Black Hills were dominated by the TetonDakota. Before that, during the latter eighteenth and early nine-teenth centuries, they were part of the Cheyenne range, with suchother tribes as the Kiowa, Comanche, and Arapaho also utilizing the 24 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154game, fuel, and other resources of the region. It is possible that at astill earlier period Shoshoneans from the Wyoming basin occasionallywandered into the district. There is neither documentary nor arche-ological evidence that corn-growing peoples ever established them-selves here in any strength or over any long period, although it ispossible that some planting of a perfunctory sort may have beenpracticed occasionally, as by the Cheyenne, even as late as the nine-teenth century. Wliat is evident, however, from the still incompletearcheological record is that throughout a long period of time, migra-tory nomads seasonally or otherwise availed themselves of the naturaladvantages the region offered over the barren, sun-baked, and wind-swept surrounding plains. Moreover, since pottery-making peoplesfrom the east and south evidently did venture into the locality fromtime to time, the prospects seem good for injecting time perspectiveinto the local archeological picture. Additional study is needed todetermine the cultural allocation of the pottery horizons representedhere, and these will have to be fitted stratigraphically or otherwiseinto their proper position in the apparent succession of nonpotteryhorizons from the west.Angostura is one of several reservoirs proposed for the upper Chey-enne Eiver Basin and the Black Hills region. Lying between the areaof semihorticultural, pottery-making peoples to the east, and the non-pottery-making hunters and gatherers to the west, it offers an un-usual opportunity to study the interrelationships and to work out thechronological relationships of representatives of these two widelydivergent subsistence economies. In the midst of an otherwise dryuninviting region, the Black Hills offer, within a comparatively lim-ited geographical area, the advantages of mountain, valley, and plainsenvironments. Here could be found water, wood, shelter, useful min-erals, an abundance of large and small game, and a rather surprisingvariety of native flora. Small wonder that the historic hunters, astheir predecessors long before, returned season after season to theHills.Fort Randall Reservoir site.?Archeological work at Fort RandallReservoir site, a Corps of Engineers project on the Missouri just abovethe Nebraska line, consisted of the excavation of a burial mound lyingon the left bank just below the dam site. This mound, 39CH9, layin the path of construction activities for the spillway, and so wasslated for complete removal. Through the cooperation of the Omahadistrict engineer and the area engineer at Pickstown, S. Dak., the con-struction agency provided a bulldozer, and the River Basin Surveysfurnished a limited amount of hand labor and the technical super-vision. Mr. John Trantina, geologist for the engineers at Pickstown,kindly surveyed the mound before its removal. Paul Cooper was in pf" ^^o^lT'^' ARCHEOLOGICAL SURVEY IN 194 8?^WEDEL 25charge of the excavation, and was assisted by J. M. Shippee. Thiswork was carried on between November 9 and 24.The mound lay at the edge of the blujffs some 200 feet above the Mis-souri Kiver bottoms, its summit marked by the Corps of Engineersmonument L-15 (elevation 1,454 feet). Before excavation, it roseto a height of about 4 feet, with a diameter of approximately 60 feet.Its surface was cut up by various small excavations, wherein occas-ional bits of bone could be seen. Otherwise, the only remains ofaboriginal activity in the immediate vicinity were a few flint chipsfrom the surface of a cultivated field lying nearby to the north.Upon excavation, it was found that the mound covered a subrec-tangular pit, with which timbers had been associated. There were noburials in this pit. Several secondary burials occurred in the moundfill and at its base. These included at one point a sort of platform oflong bones laid side by side, on top of which were placed four skulls(pi. 11, a) . It is presumed that all the interments had been secondary,although two were apparently partially articulated. No certain as-sociation could be determined between any of the burials and the arti-facts, several of which came from the mound fill and out of animalburrows.Artifacts included two small vertically elongate pottery vessels,with subconical base; both were apparently grit-tempered, and onebears a fabric marking on the surface. There were also two chippedprojectile points, one large and corner-notched, the other small withside notches. Fragments of washerlike shell objects, and a few diskand tubular shell beads came to light. There were also numerous per-forated canine teeth, the holes through the roots distinguished by cut-ting rather than by drilling, and with mesial and lateral surfacesflattened.There are several specific trait similarities between this material andthat obtained in 1947 from burial mounds (39CH4) on Wlieeler Bot-tom some 13 miles upstream. At neither of these sites, however, doesthe burial complex as a whole correspond to any other now knownburial mound of the Dakota region. The associated village complex,too, remains at present unlmown. Studies now under way on thecultural and skeletal materials may throw further light on the pre-historic mound builders whose vestiges occur but sparingly west to theupper Missouri Valley.Oahe Reservoir site.?The Oahe Reservoir is one of five major water-control projects planned by the Corps of Engineers for the MissouriRiver in South and North Dakota. The proposed dam site (pi. 12, a)is in Hughes and Stanley Counties, South Dakota, about 6 miles north-west of Pierre, S. Dak., and 1,123 miles above the mouth of the Mis-souri. Here, across the lower half of Wood Island, will be erected 26 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154an earth-fill structure, with a maximum height of 242 feet, and a crestlength of 9,800 feet. At full pool (elevation 1,620 feet), an area ofsome 298,000 acres will be flooded, and the waters will back up theMissouri to the vicinity of Bismarck, N. Dak. Purposes of the proj-ect, on which actual construction has not yet begun, are given as floodand silt control, power, irrigation, and navigation.In the portion of the Missouri Valley that will be flooded by OaheKeservoir, the river winds through a flat-floored, alluvium-filled trenchfrom 1 to 4 miles wide, bordered by bluffs from 200 to 400 feet high.Above the flood plains, which are often as much as 1 or 2 miles wide,benchlike terrace remnants rise at various elevations from 20 to 100feet or more. Frequently, the river swings against the base of thebluffs on one side, leaving on the opposite bank a broad terrace or "bottom" with a gentle ascent to the bordering uplands. Thus, as aresult of the meandering habit of the stream, the banks are char-acteristically a series of bottoms or flats alternating with rugged, hillysections. The larger tributaries, all entering from the west and in-cluding from south to north, the Cheyenne, Moreau, Grand, andCannonball, have all cut their valleys 100 to 400 feet below the up-lands, have well-developed flood plains a mile or more wide, and arebordered by well-defined terraces of varying age. Groves ofdeciduous trees, including especially cottonwood, ash, elm, and oak,stud the valley bottoms and line the stream banks, often growing alsoin the narrow ravines that descend from the bluffs.Eemains of aboriginal village sites are extraordinarily plentiful onthis section of the Missouri. They include some of the largest, bestpreserved, and most impressive sites in the Missouri watershed andGreat Plains region. In historic times, after circa 1800, variousbands of the Dakota Sioux ranged across this territory, but withoutestablishing any fixed towns. Along the main stem, the South Da-kota section of the future reservoir was the habitat of the village-dwelling Caddoan-speaking Arikara, who dwelt at various times ina series of fortified and unfortified towns from approximately Pierrenorthward. Farther upstream, above the Grand, were the towns ofthe Mandan, historically in the vicinity of the Heart River and above,but previously probably located in part farther downstream. Earliersettlements of this people possibly underlie some of the Arikara town-sites below the Grand. The Cheyenne on their historic (lateeighteenth century) movement westward from the Red River drainageto the Black Hills, are thought to have tarried for a time on theMissouri. It is quite likely, too, that still other semisedentary groups,of whom there is little or no known record at present, at one timemade this region their home. pip?kf*fr' ARCHEOLOGICAL SURVEY IN 194 8?WEDEL 27In contrast to the valley below Pierre and above Knife River, aconsiderable amount of archeological work has been carried on atvarious localities in the future Oahe Reservoir area. Before the war,surveys were conducted from time to time by the University of SouthDakota Museum in South Dakota, by the North Dakota HistoricalSociety in North Dakota, and by Logan Museum of Beloit College inboth States (1929-31). In 1932, the Bureau of American Ethnologyworked in the vicinity of Mobridge, S. Dak. ; in 1938, Columbia Uni-versity and the North Dakota Historical Society operated a joint ex-pedition in the vicinity of Bismarck; and in 1939, a ColumbiaUniversity-University of South Dakota-WPA expedition excavated ina number of sites betw^een Pierre and the Cheyenne River. In 1947,a joint expedition representing the University of North Dakota andthe North Dakota Historical Society excavated a prehistoric site belowFort Yates, N. Dak. So far, only summary reports, fortunately ofhigh quality, are available on these latter operations. Several reportson the archeology of the Missouri in North Dakota have also appeared,but the reconstructions of prehistory so far published for the mainstem below the Heart River rest largely on survey work rather thanon the badly needed excavations that alone will give the larger pic-ture. Thus, despite the apparently extensive nature of work to datein this general locality, by comparison with what remains to be donein the way of detailed examination of key sites, only a beginning hasbeen made.In the 6 weeks allotted to reconnaissance in the future Oahe Reser-voir, it was manifestly impossible to cover the entire area. Prioritywas given, therefore, to that section lying between the proposeddam site near Pierre and the mouth of Cheyenne River some 40 milesupstream. Coverage included mainly the west bank, with some recon-naissance at several sites reported along the east bank ; not all parts ofthe entire section were walked out. Some test trenching was carriedout in the vicinity of Fort Bennett. Prior to the River Basin Sur-veys work, a list of some 30 sites known for this locality had beenprovided by the South Dakota Archeological Commission. Mostof these were revisited by the 1948 Surveys party. Additional siteswere also located, so that there is now on record a total of 46 recordedsites on the west bank of the Missouri between Fort Pierre and theCheyenne River and about 15 sites on the east bank.Of the 61 sites visited by, or reported to, the 1948 Surveys party,the majority appear to be remains of earth-lodge villages. Fortifica-tions consisting of ditches (pi. 12, &) , low ridges of earth, and occasion-ally including traces of bastions, were noted at a number of sites. Hutrings, circular or oblong depressions, storage pits, refuse deposits, and 28 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154 other features characterize many of these sites, particularly wherethe surface has not been broken by cultivation. In size, they rangefrom small communities of perhaps a dozen or less houses to greatsettlements like the Fort Sully site, where hundreds of house pitsand quantities of refuse cover an area nearly a half mile long by 200yards wide. At some, there is evidence of occupation by Indians intothe time of the White conquest ; but it is a safe guess that many wereinhabited at an earlier period.As Strong observed in 1936, ". . . the Upper Missouri area is ex-ceeded in the size and number of sites as well as in cultural importanceonly by the pueblo region of the southwest, the lower Mississippi area,and the mound region centering in Ohio. . . . Yet this highly import-ant archeological area is at present represented by a single majormonograph and that only partially based on actual excavation." Interms of our present salvage problem, it may be noted further that theextent of probable flooding of archeological sites in this region is notgenerally appreciated. With a dam 200 feet or more high just abovePierre, most of the terrace sites, at least as far upstream as Fort Yates,N. Dak., will be under water at full pool. Among the better knownsites that face destruction will be: Buffalo Pasture (39ST6), on a 70-foot bench just above the upstream toe of Oahe Dam (pi. 12, &) ; LowerCheyenne Village (39ST1) on a 60-foot bench at the mouth ofCheyenne River; Fort Sully village site (39 SL4), on Telegraph Flat150 feet above the present river; and the Rygh (39CA4) and Leaven-worth (39C09) sites on opposite banks some 13 miles north of Mo-bridge, S. Dak. Even those sites not under water the year aroundwill be subject to slumping and eventual slippage into the reservoirwhen the Pierre shale which underlies many of them becomes water-logged. This is already taking place at the Lower Cheyenne Villagesite, much of which has been destroyed in the last half century byslumping ; and it is quite likely that additional sites will be similarlyaffected.Of immediate concern is the area surrounding the proposed damsite. Here no less than four archeological sites face destruction whendam building gets under way. Two of these, 39ST15 and 39ST16, lieon and beside the access railroad and classification yard, approximately2 miles below the dam site on the right bank. A third, 39ST14(Scotty Phillips Ranch site), lies about half a mile below the down-stream toe of the proposed dam, in the work area. It is also situated onthe right bank, and is a fortified site measuring approximately 175 by275 yards, with perhaps 18 or 20 house pits within the enclosure. Ithas apparently never been excavated or tested, but offers an exception-ally good opportunity for extensive and relatively inexpensive excava-tion by the controlled use of power machinery and hand labor. Across pip.N^o^.'lT'^* ARCHEOLOGICAL SURVEY IN 1948?^WEDEL 29the river, directly in line with the proposed intake structure and lyingbeneath the dam fill, is a small cluster of house pits, designated39HU22. The fourth site lies approximately 1,100 yards above theupstream toe of the proposed dam, on the right bank of the stream andwell within the future pool area. This is the fortified Buffalo Pas-ture site, 39ST6, where limited excavations have been made within theditched portion. Excepting this latter, all of the sites here noted willundoubtedly be destroyed or very extensively damaged as soon asconstruction begins. It is imperative, therefore, that systematic sam-pling be done at the earliest possible moment at all of the sites, andthat at least one, preferably 39ST14, be comprehensively investigated.Because of the great size of this reservoir, it seems certain thatmany hundreds of sites will be forever obliterated when it fills. Thatsimilar remains can be found outside the area to be flooded is improb-able. It is particularly important, therefore, that salvage operationshere be pushed as rapidly as possible. Further reconnaissance in theas yet unsurveyed remainder of the pool area should be completed aspromptly as possible, so that key sites can be chosen. Excavation isa time-consuming task at best ; and if the initiation of a program ofsystematic excavation be too long deferred after dam constructionbegins, archeologists will be confronted at the last moment with animpossibly huge piece of work.WYOMINGArcheological field work in Wyoming during 1948 was limited topreliminary reconnaissance at two proposed Bureau of Reclamationprojects in the northeastern part of the State. These were : EdgemontReservoir site, in Weston County, and Keyhole Reservoir site, in CrookCounty. The field work was done by J. T. Hughes and J. M. Shippeefrom September 16 to 20 at Edgemont, and from September 21 to 25at Keyhole. Both localities are in the headwater drainage of theCheyenne River, on the flanks of the Black Hills uplift.Edgemont Reservoir site.?The locale involved here lies on BeaverCreek, a tributary of the Cheyenne, some 35 miles northwest of Angos-tura and about 18 miles south of Newcastle, Wyo. The Beaver is asmall meandering stream in a barren plains country. Its valley islined with numerous terrace remnants and abandoned channels. TheBlack Hills are a few miles to the northeast. No figures on dam speci-fications are at hand, but the pool area is to be approximately 5 mileslong by 1 mile in maximum width. Purpose of the project is irrigation,and flood and silt control.The 4-day reconnaissance of the proposed pool area disclosed a totalof 28 sites. These occur almost continuously on and in the terracesalong both banks of the stream. Generally they are fairly superficial,953842?53 3 30 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154but overburden in some cases ranges up to several feet in depth. Com-monly the sites are marked by clusters of fire-cracked stones and quan-tities of flakes. Hearths of surface, basin, and pit types are appar-ently represented, and there appears to be a considerable variety instone artifact types. At one place, a cut bank shows near the surfacea roasting pit, which is underlain at successive lower levels by basin-like fireplaces. Here, and elsewhere in the locality, there would seemto be good prospects for correlating a succession of human occupationswith a sequence of geological events reflected in the formation of minorstream terraces.No pottery-bearing sites were recorded during the present survey,although farmers living upstream from the proposed reservoir areareport the occasional finding of pottery vessels along the Beaver.Most of the sites seen were littered with greater or lesser quantities ofchipped stone and rejectage, projectile points, blades, scrapers, knives,blanks, chert cores, etc. A few sites yielded manos. Obsidian flakeswere found at Site 48WE29. That a long range in time is probablyinvolved may be inferred from the fact that specimens recovered rangefrom stone points of early types to glass trade beads of the historicperiod.Keyhole Reservoir site.?The Keyhole Reservoir site is on the upperBelle Fourche River, at the west edge of the Black Hills some 65 milesnorthwest of the Edgemont Reservoir site. A dam 105 feet high and,with dike, about 1,100 feet long, will be located some 11 miles airlinenortheast of Moorcroft, Wyo. At full pool, the reservoir will be about10 miles long and will have a maximum width of 6 miles. The reser-voir is intended for flood and silt control, and for storage of irrigationwaters for the Belle Fourche project in western South Dakota.In the western or upper part of the reservoir site, relief is compara-tively low. The river meanders through a wide, shallow valley, tree-less except along the immediate banks of the watercourses and flankedby rolling short-grass upland prairies. The eastern portions, on theother hand, have rather abrupt valley edges and some canyon topog-raphy, and are surrounded by pine-covered hills. Vegetation charac-teristically varies from sagebrush and short grass in the west to a fairlyheavy yellow pine cover on the east.In the 5 days available for reconnaissance here, it was not possibleto search the entire pool area, but a check of the most likely portionsdisclosed 29 sites of archeological interest. A thorough reconnaissancewould probably reveal many more. With two or three exceptions,all of those found will be largely or entirely destroyed when thereservoir fills.Judged by the results of the reconnaissance, sites seem to be espe-cially plentiful in the lower portions of the future pool area, around pip-N^o/lT"^' ARCHEOLOGICAL SURVEY IN 194 8?WEDEL 31the confluence of Deer, Mule, and Cottonwood Creeks with the BelleFourche. They are situated in a variety of topographic positions.Rock hearths seem to be rather less plentiful than at Edgemont orAngostura. No pottery-bearing sites or evidence of white trade con-tacts were seen. A number of the occupational areas are of some sizeand quite productive ; where protected as they are in some instances byoverburden, they should well repay excavation. Camp-site debrisconsists of chipped stone, a few projectile points, blades, scraping andcutting tools, and abundant flakes and spalls. Fragments of boneand shell occur at some sites, and some obsidian was found. Note-worthy is the fact that a number of heavy lanceolate, "fish-tailed,"and notched projectile point forms were recovered, closely resemblingforms which have been found elsewhere under conditions suggestingconsiderable antiquity. There is thus a likeliliood that hunting hori-zons earlier than most of those represented at other reservoir sitesaround the Black Hills here await further exploration.In the opinion of the archeologist making this reconnaissance,Keyhole oflfers more promise than does Edgemont and would probablyprove relatively as productive and important as Angostura.FIELD WORK IN PALEONTOLOGYPaleontological field investigations in the Missouri River watershedAvere carried on by the River Basin Surveys from June to October,1948. Three reservoir sites, all of which had been partially investi-gated during 1947, were revisited in further search for vertebratefossils. No new reservoir localities were examined. The work wasunder the supervision of Dr. T. E. White, who was assisted throughoutmuch of the season by two student helpers?Ernest L. Lundelius,University of Texas, and John C. Donohoe, Montana State College.From August 2 to 5, White and Lundelius participated in a field con-ference of the Society for Vertebrate Paleontology and in a tour ofcertain paleontological localities in Wyoming, under sponsorship ofthe University of Wyoming. During the latter part of August, theyspent approximately 2 weeks making physiographic st'idies on Horse-head Creek, in Angostura Reservoir, South Dakota, in connectionwith the archeological investigations at Site 39FA65, as elsewheredescribed in this report.From June 4 to July 12, Dr. White's party operated in the BoysenReservoir area on Big Horn River north of Shoshoni, Fremont County,Wyo. This work was materially expedited through information sup-plied by Mr. Harry A. Tourtelot, of the United States GeologicalSurvey. Although fossil material was fragmentary and scarce, asufficient variety of specimens was collected to establish definitelythe age of geologic formations heretofore only tentatively correlated. 32 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154From 6 localities, most of them on the lower course of CottonwoodCreek, in the Lost Cabin faunal zone of the Lower Eocene, remainsof 23 species of fossil mammals were collected. These were mainly ofsmall forms ; they include insectivores, primates, rodents, carnivores,condylarths, perissodactyls, and artiodactyls. One of the insectivoresrepresented is a species new to science. Also included in the materialis the most nearly complete skull yet found of the primitive insectivore,Didelphodus, and the skull and jaws of a small carnivore, Didymicfus,previously represented only by upper and lower dentitions. Theseand other specimens promise to contribute important information toknowledge of the morphology of Eocene mammals. Reptilian re-mains collected will clarify a number of details concerning the cranialmorphology of one genus. A technical report on the paleontologicalfindings in this area is being prepared for publication.From July 14 to August 19, work was carried on in the Oligoceneand Miocene deposits of Canyon Ferry Reservoir site, on the MissouriRiver north of Townsend, Broadwater County, Mont. Here theparty was aided through information supplied by J. Leroy Kay of theCarnegie Museum, Pittsburgh, Pa. Nearly 125 specimens representing14 genera were obtained from 3 localities in the Oligocene and 2 locali-ties in the Miocene. The remains of rodents, insectivores, and smallartiodactyls were most abundant. The small Oligocene mammalsof this montane area, when compared to those of the same age onthe plains, interestingly illustrate the principles of geographicalvariation as well as do the living species.The period from August 21 to September 3 was spent in AngosturaReservoir, on the Cheyenne River south of Hot Springs, S. Dak.,assisting with geological studies the archeological work being carriedon by Hughes for the River Basin Surveys at site 39FA65. Boringswere made with a hand auger to determine the depth and characterof the valley fill on Horsehead Creek in the immediate vicinity of thesite, and geologic profiles were compiled. On the basis of these pre-liminary studies, it appears that the physiographic history of Horse-head Creek will have to be worked out in conjunction with that of theCheyenne River, and that the key to a geological dating of the terracein which 39FA65 is situated probably lies in correlation of theterrace with the receding falls of the Cheyenne. The falls now liea short distance above the canyon in which the dam is located, butin the canyon the maximum height of the waterfall member corre-sponds to the height of the oldest terrace on Horsehead Creek, inwhich site 39FA65 is located. It seems possible, therefore, that thegorge of the Cheyenne at the dam site was cut, for the most part,after the occupation of site 39FA65. A report has been prepared onthese preliminary findings for inclusion with a detailed report on pLp.^fo^.'lT'"' ARCHEOLOGICAL SURVEY IN 1948?WEDEL 33the archeology of the site; and it is hoped that additional observa-tions will be possible before the archeological and geological featuresinvolved are covered by the reservoir waters.Following completion of the physiographic studies at Angosturaand drafting of a report on them, Dr. White again went into the field.From September 23 to October 1, the Upper Cretaceous Carlile Shalein the Cedar Bluff Keservoir site on the Smoky Hill River southeast ofWakeeney, Trego County, Kans., was prospected for fossils. A num-ber of fossil fish were found, but crystallization of gypsum and weath-ering of marcasite in the matrix had reduced the remains to the pointwhere they were not worth collecting. It is believed that no furtherpaleontological salvage work is needed at this reservoir site.The fossils collected during 1948 include no material of outstandingexhibit value. Despite their often fragmentary nature, however, theyhave considerable usefulness and importance for study purposes.Since much is still to be learned about the paleontologic horizons repre-sented at Boysen and Canyon Ferry, it is highly desirable that furtherwork be done prior to filling of the pool areas. Owing to the smallsize, marked rarity, and scattered occurrence of most of the fossils onthese time levels, it has been found most economical and profitable torevisit the localities briefly in successive seasons and to collect thematerials newly weathered from the clay beds. The grain-by-grainremoval of matrix through normal agencies of erosion operating overconsiderable areas of fossiliferous deposits uncovers more materialthan would sustained or large-scale excavation, either by hand meth-ods or with power equipment. In this respect, the fossil-collectingthroughout many of the reservoir sites in Wyoming and Montana pre-sent problems quite unlike those confronting archeologists, or thosefaced by paleontologists preoccupied with large faunal forms and latergeological horizons.FIELD WOEK BY COOPERATING AGENCIESAs in the preceding year, several State-supported agencies partic-ipated actively in the 1948 investigations at Federal water-controlprojects. There were three of these cooperating agencies in Nebraska,two in North Dakota, and one in Kansas. Five were engaged whollyin archeological investigations; the sixth conducted both archeologicaland paleontological work. The operations of these agencies were en-tirely on a voluntary basis, and involved no distribution of Federalfunds. Where the cooperative work was done through a formal agree-ment between the River Basin Surveys and a particular agency, thelatter has provided reports from time to time on its activities, andsummaries of these reports have been included in the periodic and 34 BUREAU OF AMERICAN ETHNOLOGY [Bull. 354 otlier statements of progress furnished the National Park Service bythe Lincohi office of the River Basin Surveys.This section of the present report is based hirgely on data suppliedfrom time to time by the agencies concerned, either through statementsof progress, through preliminary reports prepared for publication, orthrough discussions with field personnel. Some of the sites notedwere visited by River Basin Surveys personnel while the field investi-gations by State agencies were under way. The courtesy of theseagencies and their representatives in making available advance infor-mation on their findings is hereby gratefully acknowledged.KANSASArcheological salvage operations in Kansas during 1948 were con-ducted by the University of Kansas Museum of Natural History. Aparty under the supervision of Carlyle S. Smith worked from June 19to August 20 in the Kanopolis Reservoir area on Smoky Hill River andalso outside the reservoir area along the Little Arkansas River in RiceCounty. The Kanopolis locality had been examined between August4 and 10, 1946, by M. F. Kivett and J. M. Shippee for the MissouriRiver Basin Survey. Through the cooperation of interested individ-uals, including especially William O. Leuty, Corps of Engineers, andG. L. Whiteford, of Salina, a total of 18 sites was recorded in the 1946survey. The advisability of a more intensive investigation and theprobable need for systematic excavation at certain sites was pointedout in the preliminary report of that work. With virtual completionof the dam early in 1948 and the imminent flooding of many of thesearcheological locations, it became imperative that any salvage activitybe undertaken immediatley. The Missouri River Basin Survey wascommitted to the limit of its resources elsewhere, and so the Universityof Kansas undertook what proved to be the final salvage effort on thisreservoir site. Most of the sites in the pool area were submergedbeneath the rising reservoir waters by the time the 1948 field workended.Kanopolis Dam, constructed by the Corps of Engineers, is on theSmoky Hill River about 12 miles southeast of the town of Kanopolis,in Ellsv/orth County. It has a height of 131 feet above stream bed,and a crest length of some 15,400 feet. At full pool (elevation 1,508feet) , slightly more than 13,000 acres of land will be under water, andthe lake will extend upstream about 20 miles by river to a point justwest of Kanopolis and within 3 or 4 miles southeast of Ellsworth.In this section of its course, the Smoky Hill pursues a directionsouth of east. It flows in a flat-floored valley lined with rather abruptbluffs capped here and there by outcropping ledges of Dakota sand- Pap.No^."l]"'' ARCHEOLOGICAL SURVEY IN 194 8 WEDEL 35 stone. LoAver portions of the valleys of several formerly permanenttributaries and canyons have now been flooded by the reservoir.These include Elm and Clear Creeks on the south, Bluff and Thomp-son Creeks on the north, and several short canyons near the dam.Native vegetation consisted of hardwood timber along the streambanks?chiefly cottonwood along the river, with elm and hackberryon the tributary creeks and in the canyons. The uplands, where notunder cultivation, are covered with grass.Archeological remains in the locality are of several kinds. Theyinclude petroglyphs, burial cairns, and occupational sites. Thoughthese cannot in all cases be allocated to recognized cultural horizonsor time periods, it is clear that a span of some centuries and thepresence of several different native groups can be inferred.In the immediate vicinity of the reservoir, there are at least threeknown petroglyph localities. The principal one, 14EW1, is a promi-nent sandstone cliff known locally as Indian Hill on the north sideof the valley between Horse Thief and Bed Rock Canyons, about ?'^2miles northwest of the dam. Here human, animal, geometric, andother figures cover some 50 feet of the sandstone face. Horses andmounted figures, boatlike characters, and other items suggest that partof the gallery may be post-White contact ; other portions may be con-siderably older. The underside of several large blocks that haverolled down the slope bear figures that were probably once includedin the cliff front. Smith's party found at the base of the bluff adiamond-shaped flint knife, several end scrapers, and chert chips,perhaps left by the Indians who cut the figures into the sandstone face.There has been considerable disfigurement of the locality by picnick-ers ; it lies above full-pool level and thus will not be flooded, but thelikelihood is strong that further vandalism will eventually destroymost of the characters in what is probably one of the most extensiveand interesting petroglyph sites in Kansas. The other known petro-glyph localities hereabouts are 14EW7, known locally as Three CaveFarm, on the north side of the valley about 3 miles east of Kanopolis ; and 14EW14, on Elm Creek about 2 miles north of its confluence withthe Smoky Hill. Both these are relatively small and contain only afew characters.Burial cairns at one time apparently occurred in some numbers onthe bluffs along Smoky Hill River. Most of them, however, havebeen destroyed b}^ removal of the stones for use in construction offences. The University of Kansas party opened two of these struc-tures at site 14EW13, at the extreme end of a ridge between Red RockCanyon and Elm Creek overlooking the Smoky Hill Valley. Bothcontained small rectanguloid cists covered with sandstone slabs, but 36 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154the burials that they presumably at one time held had disintegrated orbeen removed long ago. A flint chip and scraper fragment were theonly included cultural materials.A few miles outside the reservoir to the south, on top of a grassybutte overlooking a branch of Thompson Creek, four other cairns ina group of seven were opened. One, approximately 12 feet in diame-ter and 15 inches high, consisted of a rectanguloid cist measuring 30by 66 inches surrounded by small, irregularly shaped stones andcovered with slabs whose weight had crushed a flexed skeleton within.Most of the skull had disintegrated, and there were no artifacts pres-ent. The other three cairns contained no recognizable cists or burials,and were devoid of cultural materials. This site, located on the Hud-son farm, has been designated 14EW21:.The occupational sites seen in 1946 and re-examined more carefullyin 1948 suggest that at least three, and possibly four, pottery-makingpeoples were present at one time or another within the present reser-voir limits. No large or impressive sites were noted at any time;areally limited sites with rather thin deposits suggest short-lived oc-cupancies or perhaps transient residence by hunting parties frommore permanent and larger communities situated elsewhere.A single stratified site, 14EW6, was briefly worked in 1948. Thislay on Thompson Creek, near its confluence with the Smoky Hill.From the two uppermost of four culture-bearing strata came pot-sherds of Geneseo Plain and Geneseo Simple Stamped types, a tri-angular projectile point, end scrapers, a mano fragment, and nu-merous fragments of animal bone refuse. These occurred in thin, darksoil strata, along with charcoal, ash lenses, and basin-shaped hearthsunderlain by fire-reddened earth. Judged by the cultural materials,these two levels belong to the Great Bend aspect, thought to be aWichita manifestation flourishing in central Kansas between circa1475 and 1675. A third and lower stained soil level also contained afew hearths, and yielded two small plainware sherds apparently ofprehistoric pottery and identifiable as of either Upper Kepublican orWoodland type. The fourth and lowermost level had a few hearthsand broken animal bones, but no artifacts from which cultural affili-ations could be determined. These four culture-bearing strata evi-dently represent intermittent camp sites ; the sterile intervening layersof sedimentary materials presumably were laid down by prehistoricflood waters from the converging creek and river.The prehistoric Upper Kepublican culture is represented in col-lections from several small sites throughout the reservoir area. Oneof these localities, 14EW19, is on a terrace on Thompson Creek abovethe stratified site 14EW6. Others were noted in the vicinity of ClearCreek and on Elm Creek, where bits of burnt clay daub suggest the RiT.Bas.Sur. ARCHEOLOGICAL SURVEY IN 1948?WEDEL 37Pap. No. 1]remains of earth lodges. None of these, however, could be excavatedin the time and with the very limited manpower available.Woodland materials were also noted at a number of small sites in thelocality. One of these, 14EW12, at the mouth of Elm Creek, showed athin deposit of Woodland pottery in one portion, whereas Upper Ke-publican materials occurred in other parts. With respect to temper-ing, the Woodland pottery suggests two variants, one characterized byangular calcite fragments, the other by rounded sand grains orcrushed quartz. Small-stemmed or corner-notched projectile pointswith barbed blades also occurred here, as well as planoconvex endscrapers, a chipped celt, and numerous side scrapers. At 14EW13,on the ridge overlooking 14EW12 from the east, and near the twodisturbed cairns briefly noted above, were additional calcite-, shale-,and grit-tempered pottery fragments of apparent Woodland origin.Associated projectile points were small, mostly under three-quartersof an inch long, with sharp barbs and expanding stems; one had ser-rate blade edges. Other artifacts included planoconvex scrapers, aflat sandstone grinding slab, a perforated shell hoe, a small biperforateshell pendant, a T-shaped drill, and several side scrapers.Some 10 miles south of Ellsworth, on a branch of Ash Creek, the Uni-versity of Kansas party excavated an as yet unidentified subsurfacefeature with which were associated rocker-marked, dentate-stamped,zone-decorated, and other potsherds of evident Hopewell allinities.Corner-notched and expanded-base projectile points, chipped disks,flake knives, planoconvex scrapers, splinter awls, chipped celt frag-ments, sandstone abraders, and other items were also associated. Al-though no Hopewellian sites are on record for the Kanopolis Keser-voir area itself, one or two sherds found on the surface by local collec-tors suggest that this horizon may have been represented by peopleswho at times utilized or lived in the locality.In summary, it appears from the salvage operations of 1946 and1948 at Kanopolis Eeservoir that the locality has been intermittentlyoccupied on successive time levels by various native peoples. Thesedemonstrably included representatives of: the Great Bend aspect,dated circa 1475-1675 ; the Upper Republican culture, preceding 1450 ; and one or more Woodland variants, which perhaps antedated A. D.1200. It is possible that there were also Hopewellian peoples here,either coeval with the Woodland or somewhat later, but in any casepreceding the Upper Republican. No trace has been found of anyprepottery cultures here, nor is there any archeological evidence, un-less it be in the petroglyphs showing horsemen, of the Cheyenne,Arapaho, and other late hunting peoples who are known to haveroamed the area in the nineteenth century. 38 BUREAU OF AMERICAN ETHNOLOGY [BuU. 154NEBRASKAAs in the preceding year, three State agencies engaged in coopera-tive scientific salvage work in Federal water-control projects inNebraska during 1948. The Nebraska State Historical Society car-ried on archeological excavations at Medicine Creek Reservoir fromearly June until August 19, under the leadership of A. T. Hill,Director of the Museum. This work was a continuation of researchesbegun in the preceding summer. A summer field session of the Uni-versity of Nebraska Laboratory of Anthropology, under Dr. J. L.Champe, excavated sites in Harlan County Reservoir during Juneand July; and subsequently this agency carried on limited recon-naissance at several proposed reservoir sites in the Lower Platte Basinin central Nebraska. The University of Nebraska State Museum hada paleontological party at work in the Medicine Creek Reservoir, andanother at Harlan County Reservoir. From mid-July until Au-gust 25, archeological work was carried on by the Museum at a deeplyburied prepottery site on Medicine Creek a short distance below themouth of Lime Creek. The Museum researches were under the gen-eral supervision of Dr. C. B. Schultz and W. D. Frankforter ; Mr. andMrs. Preston Holder were in direct charge of the archeological studiesat Medicine Creek.A brief resume of investigations and findings by these agencies ateach reservoir project follows.Harlan County Reservoir site.?The Harlan County Dam, nowunder construction by the Corps of Engineers, is on the RepublicanRiver 3 miles east of Republican City, Nebr., 13 miles west of Frank-lin, Nebr., and approximately 235 miles above the confluence of theRepublican with the Smoky Hill River. Project plans call for anearth-fill structure with a height of about 106 feet above stream bedand a crest length of 11,950 feet. At full pool, elevation 1,973.5 feet,an area of about 53,500 acres will be under water, and the reservoirwaters will extend northwest up the Republican Valley for approxi-mately 14 miles, to a point above Orleans, Nebr. Another arm ofthe reservoir will reach southwest up the valley of Prairie Dog Creekto a point near the Kansas-Nebraska State line just north of Woodruff,Kans. Purposes of the dam are flood control and irrigation.In the vicinity of the reservoir, the Republican flows through aflat-floored valley from 1 to 2 miles wide, bordered by extensive bot-toms and enclosed by loess bluffs. Terrace remnants occur at varyingheights above the channel of the main stream as well as along itslarger tributaries. Gently rolling loess hills lie to the north of theriver valley, while the region on the soutli tends to be somewhat moredissected and broken. Cottonwood, willow, box elder, elm, and ashfringe the watercourses and the base of the bluffs, especially along the pLp.N'o^.fr' ARCHEOLOGICAL SURVEY IN 194 8?WEDEL 39south side of the valley, and small game and fur bearers still existin some numbers. The bottoms and much of the uplands are in culti-vation, but extensive areas are still used for hay and pasture.It has been known for some years that aboriginal village and burialsites exist in considerable numbers along the Kepublican and its smallerperennial tributaries. Systematic investigations on a modest scalehave been made from time to time at various localities along its course.At the beginning of the nineteenth century, the Pawnee had at leasttwo villages of earth lodges near the point where the stream leavesNebraska to enter Kansas?one near Guide Rock, in Webster County,Nebr., and another near Republic, Kans. Evidences of several earlierpeoples also are present ; some of these were clearly horticultural, atleast in part, but others seem to have relied largely or entirely onhunting and gathering for their subsistence economy. The accumu-lating evidence suggests that these different peoples probably did notoriginate in the same locality or from the same basic culture complex,but that they were of diverse origins and antecedents. For most ofthe complexes indicated, however, the available information has beenrather meager, resting on the scattered sampling of many sites ratherthan on the intensive and comprehensive examination of a few keysites.The Harlan County Reservoir site was first examined by a RiverBasin Surveys field party, consisting of M. F. Kivett and J. M. Ship-pee, during the summer of 1946. Their operations, carried on fromAugust 20-25, consisted of surface collecting, some small-scale test-pitting, and interviews with local and other persons familiar with thelocality and its antiquities. Earlier investigations by the NebraskaState Historical Society and the University of Nebraska Laboratoryof Anthropology also supplied helpful leads in this survey. InNovember, several of the more promising locations were rechecked,and one burial site, 14PH4, was excavated.From these investigations, it was determined that of 23 knownsites in the vicinitj'- of the reservoir, including 16 native occupationaland 7 burial areas, 14 apparently lay below the 1,973-foot contourlevel and would be inundated at flood stage. Six others lay belowthe 2,000-foot level where it seemed likely that erosion by wave actionor otherwise might affect them adversely; and three occurred underconditions that suggested they would very likely be damaged by con-struction work. Moreover, it was apparent that not less than fourarcheological complexes, probably involving a time span of as muchas a thousand years, were represented within the limits of the futurepool area. Included among these are Woodland variants, UpperRepublican and Dismal River sites, and at least one apparently as-signable to an unnamed but probably late pottery horizon also rep-resented elsewhere in eastern Nebraska and northern Kansas. 40 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154Since the nature of certain of these complexes, as well as their inter-relationships, remain obscure, it seemed highly desirable that sys-tematic excavations be carried out before the sites were destroyed bythe rising reservoir waters. Because of inadequate funds and priorcommitments elsewhere, the River Basin Surveys were unable to under-take the needed investigations here. The University of NebraskaLaboratory of Anthropology was accordingly invited to make suchinvestigations as seemed desirable and feasible, especially in the lowerportion of the future pool area. That agency accepted, and took overthe work of exploring certain especially promising sites on the banksof Prairie Dog Creek.The principal site worked, 25HN37, is located on the left bank ofPrairie Dog Creek about 3 miles above its confluence with the Repub-lican, and some 6 miles southeast of Alma, Nebr. Occupational re-mains here are scattered over an area at least 1,000 feet long and 250feet wide, along the rim of a terrace which forms a bluff perhaps 30feet high on the north bank of the creek. With the aid of machinery,a trench nearly 800 feet long was opened along the edge of the siteterrace overlooking the creek. A fire-pit uncovered near one end ofthe trench led to widening of the cut over what was subsequentlyidentified as a house site. Similar features were worked out in as-sociation with each of three fireplaces located in tests made by Kivettand Shippee in November 194G. Not far from the house sites, aroasting pit was also opened.The house sites, as defined, each consisted of five post holes arrangedsymmetrically around an ash-filled fireplace, to form a pentagon from12 to 15 feet in diameter. Within this pentagon, charcoal and debriswere mixed throughout the earth to a depth of 5 or 6 inches below thelevel of the top of the fireplace. No evidence of an outer ring of post-holes could be found. In three cases, a pair of smaller postholes wasfound 12 or 13 feet from the pentagon on the east side, possibly rep-resenting the outer end of a doorway. There were no traces of wat-tling clay, such as often occurs in the more familiar earth lodge sitesof the Central Plains. Details of construction for the habitationshere represented are uncertain, but it seems clear that the lodges musthave differed rather widely from the earth lodges of the Pawnee andother semisedentary village tribes. There were no cache pits withinthe house sites.The roasting pit lay not far from the three houses. In cross section,it was bell-shaped, with walls and bottom burned a bright red. Itmeasured 25 inches in depth, with a diameter at the orifice of 40 inchesand at the floor of 52 inches. Ash, clay, earth, and charcoal strata,along with animal bone and some artifacts, filled the pit.Artifacts were not abundant in the excavations, but the limitedsample so obtained and the rather extensive series of materials col- Ri^^'-Bas.Sur. ARCHEOLOGICAL SURVEY IN 1948?WEDEL 41lected from the cultivated site surface leave little doubt regarding thecultural affiliations of the site. Potsherds are thin, hard, and fine-textured, usually dark in color, with fine sand or mica tempering;exterior surfaces are plain or simple-stamped; and rims are mod-erately high, slightly flared, and have rounded or flattened and evertedlips. No recognizable puebloan sherds or other trade artifacts are re-ported. Other artifacts include numerous end scapers; scraper-graver combination tools ; drills of several forms ; triangular projectilepoints with or without side notches ; sandstone abraders ; bison scapulahoes and choppers; fleshers without serrate blades; bison-rib shaftwrenches ; broken awls ; long tubular bone beads. Bison, deer, beaver,turtle, and other bone refuse occurs. No evidence of white tradecontact was found in situ, though some of the surface iron and brassmay have belonged to the Indian occupants.A report on the 1948 findings at 25HN37, prepared by Champe, hasbeen published in the April 1949, issue of American Antiquity.The University of Nebraska State Museum was represented atHarlan County Dam by two students. In the early part of the sum-mer, while construction work was suspended because of a labor strike,these men spent their time prospecting for fossils in the vicinity ofthe reservoir. Several promising Pliocene quarries were located, alloutside the future pool area. Upon resumption of construction ac-tivities, the Museum representatives devoted their time to watchingthe excavations for paleontological remains and to tests at some ofthe newly discovered localities nearby. During the winter, a newfossil quarry at the south end of the dam axis was reported to theMuseum by Corps of Engineers personnel, but adverse weather pre-vented investigations during the remainder of the calendar year.Medicine Greek Reservoir site.?The 1948 investigations of theNebraska State Historical Society here involved the excavation ofhouse units and refuse areas at six village sites, and test excavationsat a seventh. The sites selected lay on the right (west) bank of thecreek, outside the then federally owned lands and mostly upstreamfrom those under study by the River Basin Surveys. Two of thesites, 25FT19 and 25FT20, lay just south of Lime Creek; the otherswere all within a distance of approximately 2 miles to the north. Atnone were the remains present completely worked out.Sites investigated, with the principal features opened in each, in-clude 25FT19, two house floors and one midden ; 25FT20, one housefloor ; 25FT22, two house floors and two middens ; 25FT28, one housefloor and one midden; 25FT30, five house floors and one midden;25FT39, two house floors and two middens ; 25FT23, tests only. Allthe house floors uncovered were subrectangular in form and laya few inches to approximately 2 feet underground ; characteristically,they showed four primary or central postholes. These structural 42 BUREAU OF AMERICAN ETHNOLOGY tBull. 154 evidences, the pottery remains, and the work in chipped and groundstone, in bone, shell, and other materials all conform closely to thelarger series obtained in the River Basin Surveys excavations nearby.By extending the foundation of controlled data on which laboratoryanalysis and interpretation are based, these remains constitute animportant supplement to the data gathered at various times byother agencies. Federal and non-Federal, in the Medicine Creeklocality. There are, to be sure, some variations from site to site;but the materials collected by the Historical Society in 1948 canundoubtedly be safely ascribed to the Upper Republican horizon. Itis not possible at the moment to assess the site variations in terms ofpossible time or other significant differences.Archeological work at Medicine Creek by the University ofNebraska State Museum consisted of excavations at a deeply buriedprepottery site designated 25FT50. This is located in the right (west)bank of Medicine Creek a few hundred yards downstream from themouth of Lime Creek, on which two other early prepottery sites,25FT41 and 25FT42, had previously been worked. At site 25FT50,the cultural materials occurred in a zone about 2i/2 to 3 feet thick inthe lower part of a terrace fill provisionally identified by Museumpaleontologists as Republican River Terrace 2. There were twolevels of concentrated debris and stained soil separated by a lightercolored intermediate zone in which much less cultural material wasfound. Most of the artifacts and about half the hearths uncoveredwere in the lower level, designated Occupational Level 1. Scatteredabout through the occupational refuse were numerous animal bones,including bison, antelope, deer, coyote, rabbit, and smaller forms, aswell as occasional reptiles, amphibians, and birds. Most of the largerbones had been cut, broken, or otherwise modified by human industry ; there were no large mammal skulls. Artifacts included leaf-shapedprojectile points with concave base; trapezoidal scrapers, some withgougelike bits ; ovoid and lanceolate blades ; drills ; abrading or grind-ing stones; a flattened stone spheroid with equatorial groove; eye-letted bone needles; crude bone awls; a bipointed bone object; andmiscellaneous cut and worked bone fragments. Some of the speci-mens show a general resemblance to artifacts from the later potteryhorizons of the locality; but others, such as the projectile points,gougelike scrapers, and perhaps the grooved stone are reminiscentof earlier horizons. Also the depth of overburden argues stronglyfor an age considerably greater than that of the Woodland and UpperRepublican pottery sites found in the immediate vicinity but underdifferent physiographic conditions.A preliminary statement by Mr. and Mrs. Holder regarding thefindings at 25FT50 has been published in the April 1949, issue ofAmerican Antiquity. pip.N^o^.'rr"^' ARCHEOLOGICAL SURVEY IN 194 8?WEDEL 43Paleontological researches by the Museum at Medicine Creek in-cluded work at two Pliocene fossil deposits, 25FT40 and 25FT47,which w^ere endangered by construction of an access road. From25FT40 were gathered the remains of some 25 species of insectivores,rodents, carnivores, perissodactyls, artiodactyls, fish, birds, and othervertebrates, some of them reportedly representing new species. Theseforms are said to "represent the latest Pliocene assemblage so fardiscovered in the Great Plains region and therefore will be of utmostimportance in establishing the Pliocene-Pleistocene boundary . . .Several specimens are the largest recorded from the Pliocene of theGreat Plains and probably represent the latest survivors of theirtimes."At 25FT47, the faunal list is much shorter; the quarry has beententatively identified as Ash Hollow (Middle Pliocene) in age.ARCHEOLOGICAL RECONNAISSANCE IN THE LOWER PLATTE BASINThe Lower Platte Basin includes that portion of the Platte Riverwatershed lying east of Lake McConaughy, near Ogallala in KeithCounty, Nebr. The Platte itself courses generally eastward near thesouthern boundary of the basin. In eastern Nebraska, it is joinedfrom the north by two major tributaries?the Loup near Columbus,and the Elkhorn near Ashland. These two tributaries drain muchof the Sandhill and Loess Plains areas of central and eastern Nebraska.Water development plans of the Bureau of Reclamation for thelower Platte Basin include construction of about 16 major reservoirs.Most of these will be distributed in a belt some 50 miles wide extendingfrom northern Gosper County northeastward to Antelope County;others lie outside this belt to the north on the upper reaches of theLoup system, or else to the eastward about the confluence of theLoup and Platte Rivers.Preliminary examinations of several of the proposed reservoirsites in this region were made in May 1947, by a party from theRiver Basin Surveys. Projects visited at that time include Amherst,Buffalo Creek (renamed Bison), Cairo, Ericson, Mullen, and Rock-ville. Also traversed at the time was Brewster, where, however, noactual survey was attempted. Preliminary reports of the findings atthe several projects here listed have been issued, although in no casewas complete coverage of the future reservoir area possible.In 1948, the University of Nebraska Laboratory of Anthropologyundertook reconnaissance at several additional projects proposed forthe Lower Platte Basin. This work was carried on over a period of 3weeks, from August 15 to September 2, following termination of theLaboratory's excavation program at Harlan County Reservoir. Twouniversity students in anthropology, J. H. and D. Gunnerson, made 44 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154the actual field investigations, under the general supervision of Dr. J.L. Champe. Units visited include Gushing, in Howard and GreeleyGounties ; Loretto, in Boone Gounty ; Plum Greek, in Gosper Gounty ; and Rosedale (formerly Glearwater) in Antelope Gounty. Gushing,Loretto, and Rosedale are located north of the Platte River; PlumGreek lies just south of that stream in south-central Nebraska.The observations of the Laboratory field party at the four reservoirsites visited would seem to indicate that the water-control projectswill directly affect a number of aboriginal sites representing varioustime levels and several cultural complexes. These include remainsapparently assignable to the certainly prehistoric Woodland andUpper Republican horizons ; and also the traces of one or more laterprehistoric, or possibly protohistoric, peoples whose identity has notyet been established. Further surveys at all four reservoir sites, andprobably some excavation, are believed to be warranted by the newevidence now at hand.Gushing Reservoir site.?The site of the proposed Gushing Reser-voir is on Spring Greek in Howard and Greeley Gounties, Nebr.Spring Greek enters the Loup River from the north just below thejunction of its North and South Forks. The proposed dam site isa short distance northwest of the town of Gushing, from which pointthe reservoir will extend upstream to the vicinity of Wolbach, inGreeley Gounty. The maximum water surface area will be about1,700 acres.Seven archeological sites, all lying below the 1,850-foot contour,were recorded in the future pool area by the University of NebraskaLaboratory of Anthropology field party. Five of these yieldedpottery remains as well as other occupational debris; on the othertwo, only bone fragments, stoneworking refuse, chips, and simi-lar materials were found. Of the pottery sites, three yielded sherdsof Upper Republican type and one had sherds suggestive of Wood-land wares. From one site came a rather curious assemblage ofpottery fragments including several grit-tempered sherds apparentlyin the Upper Republican-Nebraska culture tradition ; others of differ-ent type with broad shallow gi-ooves and trailed horizontal lines oc-casionally cut by diagonals ; and one fragment with horizontal single-cord impressions on the rim exterior and cord-impressed diagonalson the outer edge of the vessel lip. From the limited observationsmade, it is not clear whether a stratification of archeological com-plexes is here indicated, or whether some other interpretation will becalled for.The occurrence of grass-impressed wattling clay on several sitessuggests the former presence of earth-covered habitations of fairlysubstantial character. These, in turn, would imply a fairly sedentary pipN^o^'ff"^" ARCHEOLOGICAL SURVEY IN 1948?^WEDEL 45mode of life on the part of the erstwhile inhabitants, and possibly asemihorticultural subsistence economy.Loretto Reservoir site.?The proposed Loretto Reservoir is to belocated on Beaver Creek in Boone County, Nebr. Beaver Creek joinsthe Loup River from the north near Genoa, some 25 or 30 miles south-east of Loretto. The dam site is not far above Loretto, whence the poolwill extend about 8 miles northwest along the Beaver. At maximumproposed pool elevation of 1,849 feet, an area of about 2,360 acres willbe under water.Seven sites of archeological interest, all situated below the 1,900-foot contour, have been recorded for the Loretto Reservoir site. Sincea heavy growth of weeds covered much of the ground surface, it isquite probable that additional sites remain undiscovered. Potteryremains were found at four sites; and, while the samples collectedwere not large, they suggest the former presence of at least threedifferent pottery-making peoples. Small Woodlandlike pottery frag-ments occurred at two sites; another yielded a sherd of apparentUpper Republican type ; and still another had "hole-tempered" sherds,one of which bore deep trailed parallel lines as surface decoration.These are pottery types not generally expected to occur on the sametime level; and they suggest that the locality may have been occupiedor visited at different times by representatives of successive widespreadaboriginal cultures.The three nonpottery sites were characterized by bone fragments,chips, and other village-site refuse ; there is no way of determining atthe moment the relationships of these remains to those of the pottery-using natives of the locality.Plum CreeJc Reservoir site.?The site of the proposed Plum CreekReservoir is on Plum Creek in northern Gosper County, Nebr., a fewmiles south of the Platte River. No data on size and operation of theproposed project are available at this time, other than a map pre-pared by the Tri-County Project, P. W. A., in 1941. This shows twodifferent reservoir sites; both are in Gosper County, but the upperextends a short distance into Frontier County. Both locations wereexamined by the Laboratory party, but torrential rains and theresulting impassible roads cut short the survey of the upper site.Ten archeological sites were recorded in this locality during the1948 survey. With a single exception, these were marked only byoccasional bone refuse, and by flint chips, cores, and other nonceramicremains. The one exception is a hilltop site, 25GO20, where potsherds,chips, cores, burned and unburned bone scrap, and other materialswere picked up. Much material is reported to have been found atthis spot in past years by local collectors ; and of the sites seen by the1948 survey party, this one seemed the most promising and prolific.953842?53 4 46 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154Judged from the potsherds seen, the site appears to be attributable tothe Upper Republican horizon.Information from local collectors indicates that pottery-bearingvillage sites assignable to Upper Republican peoples occur on TurkeyCreek, tributary of the Republican River, in southern Gosper County,but that only a single site of this complex?25GO20, noted above?hasbeen recognized to date in the Plum Creek Reservoir area.Rosedale Reservoir site.?Location of this proposed project, form-erly called the Clearwater Reservoir, is on Clearwater Creek in Ante-lope County, Nebr. Clearwater Creek is a permanent stream empty-ing into the Elkhorn River from the southwest a few miles below theproposed reservoir and east of the town of Clearwater. The dam siteis to be approximately 2 miles south of the town of Clearwater and9 or 10 miles west of Neligh. The reservoir pool will extend about 6miles up the creek, to or slightly beyond the Holt County line. Atmaximum level, 2,850 acres will be under water. The surrounding ter-rain is gently rolling to hilly, with extensive sandy areas ; the creekvalley has low terraces rising 8 to 10 feet above the flood plain and afringe of hardwood timber along the stream banl^s.Despite the comparatively small area involved here, archeologicalsites appear to be fairly plentiful and of varied character. Elevensites were recorded in the 1948 survey by the Laboratory of Anthro-pology ; it is very probable that others remain undiscovered. Six ofthe sites yielded no pottery whatsoever ; their age and cultural affilia-tions cannot even be guessed at with the meager evidence now at hand.Potsherds strongly suggestive of Upper Republican wares werefound on two sites ; in one instance, they occurred at a depth of 24-30inches in a road cut. In another site they seemed to be coming out ofa shallow buried stratum, along with other debris and sherds of pos-sible Dismal River affiliations. The suggested association, and in anycase the exact stratigraphic relationships here, deserve further investi-gation. Three other sites yielded sherds with shell and/or grit tem-pering, simple stamped exterior surfaces, and parallel horizontaltrailed lines which had evidently encircled the rims of the vessels.This material is somewhat reminiscent of ceramic remains found pre-viously at Ericson, Harlan County, and other reservoir sites, as well aselsewhere in north-central Nebraska. In general, this material givesthe impression of being relatively late in time, but since none of thesites in which it occurs has yet been comprehensively excavated anddescribed in print, the complex represented remains pretty largelyunknown and is still unclassified.NORTH DAKOTAIn North Dakota, cooperative archeological salvage work wascarried on by a field party from the University of North Dakota, work- pip.?fo^.'ri"'^' ARCHEOLOGICAL SURVEY IN 194 8?WEDEL 47ing in conjunction with the North Dakota Historical Society. Underthe active supervision of Dr. Gordon W. Hewes, a party of six was inthe field at the Baldhill Reservoir from June 25 to August 1, 1948.During this period, two burial mounds were opened and several villagesites nearby were briefly test-pitted. Unfavorable weather and a con-sequent delayed crop season, which prevented access to several promis-ing bottomland village sites, hampered the operations. This work, itshould be noted, followed a reconnaissance for the River Basin Surveysduring 1947 by a party of four, under M. F. Kivett, at which time threemound groups and seven occupational sites were recorded. Kivett'sparty also partially excavated a burial mound in Griggs County tosalvage exposed skeletal material, and recommended further work inthe region.Baldhill Reservoir site.?Baldliill i)am, a Corps of Engineers proj-ect, is under construction on the Sheyenne River, 16 miles upstreamfrom Valley City, N. Dak., and about 271 miles above the mouth of thestream. It is to be an earth-fill structure having a height of approxi-mately 61 feet above stream bed and a length of 1,650 feet. Purposesof the project are flood control, water supply for industrial andmunicipal use, and pollution abatement.For the 1948 investigations by the University of North Dakota-North Dakota Historical Societj^ party, two mounds in Barnes Countywere selected. These were designated as site 32BA1 by the River BasinSurveys. They lay about 90 yards apart on the left bluff of theSheyenne River nearly one mile below its junction with Baldhill Creek,and some 60 or 70 feet above the river. Both mounds were approxi-mately circular in outline, perhaps 100 feet in diameter, and between 6and 7 feet in height. They appeared to have been built wholly oftopsoil gathered up nearby, without any preliminary clearing awayof the topsoil on the spot over which they were erected. Beneath oneof the mounds, the original soil surface was marked by a fine layer ofcarbonized grass stems and ash, suggesting that the grass had beenburned off before the tumulus was built. Glacial boulders, of whichmany occurred on the field surface about the mounds, had been freelyincorporated in the mound fill. No humus layers or other significantstratigraphic features were noted in the mound sections ; both appearto have been constructed within a relatively short span of time ratherthan by gradual long-term accretion.Upon excavation, both mounds were found to have a central under-lying burial pit whose fill contained fragmentary and disarticulatedhuman skeletal remains of all ages and both sexes. The central pitin Mound B measured 11 by 5 feet, with a depth of 3 feet or morebeneath the original ground surface. These pits had evidently beenroofed over with transversely laid oak timbers which, in the case of 48 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154Mound B, rested on rows of glacial boulders set along the sides of thepit. It is presumed that the space below the timbers was left openwhen the mound was built, and became filled with soil and rocks whenthe timbers finally decayed. Mound A contained a second shallowerpit without roof, just east of the principal chamber. In this werefound four partially articulated and moderately well-preserved adultskeletons, apparently buried together side by side and alternatelyoriented.Artifacts were not plenf iful in either mound, nor were they of suchnature as to be very helpful in determining the cultural affiliationsand chronological level of the associated skeletal materials. FromMound A came a medium heavy stemmed projectile point; a heavybone punchlike object; a human molar with ground-off root; anend-perforated freshwater mussel-shell ornament ; a painted bird-bonetube fragment ; four partly worked or ground carnivore teeth ; a com-plete human upper dental arch and palate carefully cut and grounddown along with its included teeth ; and miscellaneous scrapers, flakes,etc. Mound B yielded a small cylindrical copper bead, a clay bead( ?), a crudely chipped-stone digging (?) tool, and one complete andone fragmentary bison-rib beaming or smoothing tool. The disin-tegrated skull, ribs, and vertebrae of a bison, possibly stained orpainted, lay just above the fill near one end of the central grave pitin Mound B.A preliminary report on the 1948 findings at Baldhill has been pub-lished by Hewes in the April 1949 issue of American Antiquity. Aspointed out there, it is hoped that further study of the physical an-thropology of the intact and restorable human remains from these twomounds, plus dendrochronological analysis of the log fragments fromthe burial chamber will throw further light on the difficult problemof dating and identifying culturally the eastern Dakota burial mounds.In certain particulars the Baldhill mounds opened by Hewes arereminiscent of findings by Montgomery, Strong, Meleen, and Cooperat other burial mounds in eastern North and South Dakota ; but, un-fortunately, the complex as such cannot yet be specifically and directlyrelated to that at any other reported mound sites of this general regionor of the Minnesota woodlands to the east.RESUME AND CONCLUSIONSIn the foregoing pages, we have briefly reviewed the archeologicaland paleontological field work carried on in the Missouri River Basinduring calendar year 1948 by the River Basin Surveys and coop-erating agencies, in conjunction with the Federal water-control pro-gram. Archeological investigations by River Basin Surveys partiesincluded extended excavations at Medicine Creek, Nebr., and at Fort pip.?fo^;iT'^' ARCHEOLOGICAL SURVEY IN 1948?WEDEL 49Randall, S, Dak. ; surveys and test excavations at Angostura, S. Dak.,at Heart Butte, N. Dak., and in the lower Oahe Reservoir site, S. Dak. ; and reconnaissance at Oahe, Pactola, and Jolinson Siding, in SouthDakota, and at Edgemont and Keyhole, in Wyoming. Paleontologi-cal work by the River Basin Surveys included re-examination and col-lecting at Boysen Reservoir site, Wyo. ; at Canyon Ferry, Mont.;and at Cedar Bluff, Kans.Cooperating agencies, and the localities investigated by each, in-cluded the following : University of Kansas Museum of Natural His-tory, at Kanopolis, Kans.; University of Nebraska Laboratory ofAnthropology, at the Harlan County, Cushing, Loretto, Plum Creek,and Rosedale Reservoir sites, Nebr. ; University of Nebraska StateMuseum, at Harlan County (paleontology) and Medicine Creek (ar-cheology and paleontology) , Nebr. ; Nebraska State Historical Society,at Medicine Creek, Nebr. ; University of North Dakota-North DakotaHistorical Society, at Baldhill, N. Dak.These investigations represent a widely scattered sampling?spa-tially, temporally, and culturally?of the aboriginal remains of theMissouri Basin region. From Kansas to North Dakota, in a varietyof ecological settings, the traces of native pre-White and early con-tact peoples came under observation and detailed study. In some in-stances, important new data were added to previous records of theprehistory of the localities involved, and the place of these findingsin the over-all picture of native life in the Great Plains was madeappreciably clearer. Elsewhere, the findings are represented by datainadequate as yet to permit exact placement of the remains in the pres-ent scheme of things. Despite the uncertainties and differences ofopinion attached to some of the discoveries and their interpretation,it is clear that our knowledge of human prehistory in the Basin hasadvanced materially over the past year as a result of these researches.It is also clear that many of the problems now puzzling the archeol-ogist here cannot be solved without the assistance of qualified studentsfrom other disciplines. With the archeological manifestations oftenare linked various phenomena having to do with past climatic vari-ations, depositional and erosional problems, stream changes, and othermatters of much importance to the student of human prehistory butfor the most part beyond his particular abilities of interpretation.Included in the 1948 archeological field work are several sites whoseage almost certainly is to be measured in terms of millennia ratherthan of centuries. These are the Allen site, 25FT50, in MedicineCreek Reservoir, Nebraska, and the Long site, 39FA65, in AngosturaReservoir, South Dakota. At Medicine Creek, site 25FT50 is one of agroup of three that have been under study since 1947. Sites 25FT50and 25 FT41, the only ones so far comprehensively investigated, have 50 BUREAU OF AMERICAN ETHNOLOGY I Bull. 154been assigned by University of Nebraska paleontologists to the basalportions of Kepiiblican River Terrace 2, which is provisionally equatedwith the beginning of the Mankato substage of the Wisconsin glacia-tion. If this correlation is correct, the sites in question would ante-date Eden Valley Yuma and also the Lindenmeier Folsom horizon, asthese have been dated by other geologists. Presumably, they wouldalso equate in time with the native horse, the Columbian mammoth,and Bison antiquus, as the stratigraphic range of these mammals hasbeen reported for Nebraska. Despite the fact that the Lime Creeksites, especially 25FT41 and 25FT50, have produced a considerableamount of refuse animal bone, there is as yet no published proof thatany of this material represents extinct forms, or that the bison re-mains found are other than those of the modern plains species. Thatthe sites involved are in or near the base of one of the older terracefills of the local drainage system is clear; but until the faunal andgeological evidence bearing on the sites has been fully detailed andverified, the estimates of antiquity suggested by the paleontologistsconcerned must be viewed with reserve.So far as the archeological complex, or complexes, at Lime Creekare concerned, there is little at the present writing to substantiatethe placing of the material chronologically anywhere near or previousto the Lindenmeier Folsom horizon, where the association with extinctbison, B. taylori^ seems well documented and is generally accepted.With exception of certain projectile points, most of the stone andbone artifacts so far reported from Lime Creek can be duplicated inany large collection of materials from Upper Republican and otherarcheological complexes in southern Nebraska, even though the latterhave no historical connection and are without question from a muchlater time level. As a complex, it seems to me that the materials fromthe Allen site are rather more reminiscent of the remains from SignalButte 1 than of any defined Folsom complex ; whether further exca-vation would heighten or lessen this general similarity I have, ofcourse, no way of knowing. More detailed and extended analysesand comparisons than have yet been made are needed before this prob-lem and the true relationships of the Lime Creek materials can besettled.At Angostura, site 39FA65 is an occupational zone of considerableextent but with as yet very limited material culture inventory. Sim-ple fireplaces are present ; as in the Lime Creek sites, no stone seemsto have been used in connection with these features. Occupation ap-pears to have been intermittent and of short duration, apparentlyalong the edge of a creek valley whose contours were unlike those ofthe present valley. Several "early" point types have been found onthe eroded edge of the site ; in similar position, as well as in the cul- Pap.N^f.'ri"'' ARCHEOLOGICAL SURVEY IN 194S?WEDEL 51 I ture stratum, have been found several fragments of lanceolate points i characterized by narrow straight to concave bases, fine flaking, and I ground basal edges. These have some similarity to certain specimens i. from the Allen site, and also to a series of well-made blades collectedby Roberts in 1942 from a bison kill in the Agate Basin between Lusk , and Newcastle in eastern Wyoming. It is presumed that the inhabi-tants of site 39FA65 subsisted mainly on large game, but to datebone fragments have been strangely absent. It is not clear whetherthis is attributable to local soil conditions or to other factors. Fur-ther work at this promising and important site is planned for 1949. I Presumably later in time than the occupation of the Long site at ! Angostura were a series of poorly defined and meagerly representedcomplexes found along the upper reaches of the Cheyenne and itstributaries in southwestern South Dakota. In the vicinity of the An-gostura Dam there are several sites exhibiting physical stratificationin the form of dark soil horizons of varying thickness, separated bynoncultural deposits. Archeological remains occur in the darker soilformations, chiefly as fireplaces of various kinds with which are asso-ciated very limited quantities of other cultural materials. There issome stratigraphic evidence that simple hearths built on a layer, orwithin a circle, of stone may be a rather early form; that this typewas superseded by a shallow basin partly filled with stones ; and thatthis in turn was followed by pits whose diameter and depth wereapproximately equal, whose sides and bottoms were hardened andreddened by prolonged heating, and which usually contain fire-crackedstones and sooty soil. The last type is reminiscent of the somewhatsimilarly shaped but usually larger roasting pit of the protohistoricDismal River horizon in the upper Kansas-Republican drainage.There are some slight indications that the projectile points and otherartifacts found sparingly in association with each of the hearth typesmay be distinctive, but the available samples are too small to beconclusive. At some sites, manos, metates, and fragments are ratherplentiful; elsewhere they are scarcer or wholly absent. Seasonalshifts of emphasis in food-getting, rather than distinct culture com-plexes, may be indicated by such differences. It is noteworthy, too,that few of these sites show any great quantity of animal bone refusesuch as one would expect if the occupants had relied largely onhunting for their subsistence. This may mean that most of theslaughtering was done outside the living area, and only the edibleportions of the game brought into camp; but somehow this is nota very satisfying explanation.The findings at Medicine Creek and at Angostura, insofar as theyinvolve the remains of nonpottery-making peoples, are of interest forseveral reasons. In the first place, they help to fill in some of the 52 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154long and little-known intervals of time during which the Great Plainswere certainly inhabited, if sparsely, before the arrival of pottery-making semihorticultural peoples in the western plains. They indi-cate, or strongly suggest, recurrent short-term occupancies in manylocalities throughout a wide area that today impresses one as ratheruninviting and inhospitable. At the same time, the very meagermaterial culture assemblage from nearly all sites so far found prob-ably indicates a plane of living but little above a bare subsistencelevel and far below that characteristic of the historic horse nomadsor the semihorticultural Indians of late prehistoric, protohistoric,and historic times.No less important to Plains prehistory than the foregoing are theadditional data gathered during 1948 concerning the pottery-makingIndian inhabitants of the region. These, in general, are probablyfrom a later period than the materials considered above ; but it mustbe emphasized that not all of the potteryless horizons at Angostura,for example, can be proved to be earlier than some of the ceramiccomplexes in and east of the Black Hills. It is possible that Wood-land and Upper Republicanlike penetrations here will be found tointerdigitate with some of the later lithic horizons of the upperCheyenne drainage. At the moment, direct stratigraphic evidencebearing on this question is not at hand.Nothing in the 1948 findings in the Missouri Basin controverts thelong-held view that, among pottery-bearing complexes, those desig-nated as Woodland are the earliest in the region. It is perhaps morenearly correct to speak of a Woodland series, since there is growingevidence of more than one trait assemblage within what has beencalled the Woodland horizon. Kivett has proposed the term "Keithfocus" for a group of sites in southern Nebraska and western Kansaswhich are partially characterized by calcite-tempered, cord-rough-ened pottery; small stemmed or corner-notched projectile points,sometimes with serrate blades; small chipped celts; tubular bonebeads, either plain or with encircling incisions ; and secondary burialsin communal ossuaries which also contain great numbers of shell diskbeads, triangular corner-perforated shell pendants, and certain othertraits. The two Woodland habitation sites?25FT18 and a smallarea in 25FT70?excavated at Medicine Creek Reservoir area in 1948,may be assigned to the Keith focus, and it seems probable that thesame or a similar complex is represented at Kanopolis Reservoir inKansas. At both Medicine Creek and Kanopolis, the Woodland sitesare small, the artifact yield is low, and evidence of maize-growingis nonexistent. In general, one gets the impression that small popu-lation groups and a rather simple hunting and gathering subsistenceeconomy, albeit one with a well-established pottery tradition, are tobe inferred. pip.N^o^'lT"^" ARCHEOLOGICAL SURVEY IN 1948?^WEDEL 53Probably also assignable to a Woodland horizon are the moundsexcavated in the Fort Randall Reservoir area?39CH4 and 39CH9.As already noted, some of the individual traits at these burial sitesoccur in widely separated localities from northern Kansas to Minne-sota. Presumably the structures, whose associated village complexremains unknown, represent a western extension or variant of one ofthe Minnesota Woodland cultures. The mounds opened by Hewesat Baldhill lack many of the material traits found scatteringly inother burial structures around Devils Lake and elsewhere in easternNorth and South Dakota, but in all probability likewise have a basicrelationship to some Woodland manifestation of the Minnesota region.Neither the Fort Randall mounds nor those at Baldhill show a veryclose similarity as complexes to anything called Woodland in thecentral Plains.Woodland potsherds from several occupational sites in FortRandall Reservoir, though few in number, suggest at least two othervariants with counterparts to the south in Nebraska. Sherds dec-orated with single-cord impressions are reminiscent of pottery fromsites on Eagle Creek, in Holt County, and from Loseke Creek, inColfax County. Others have punched rim bosses and exterior cord-roughening. To what extent these and other slight but seeminglyconsistent pottery variations are paralleled by differences in othermaterial traits remains to be determined. Such a definition of theseveral variants now lumped together as Woodland, and determina-tion of their relative temporal position, would be a long step aheadin the understanding of plains prehistory.It has been noted above that evidence of maize-growing by theWoodland peoples on Central Plains sites is very scanty or else isaltogether wanting. For the northern plains, in the Dakotas on bothsides of the Missouri River, there is virtually no published informa-tion from which the nature of the local Woodland subsistence econ-omies may be judged. It may be surmised that the westerlymanifestations represent largely nonhorticultural peoples, whereasthe easterly ones in which burial mounds occur may prove to besemihorticultural ; but this is sheer conjecture and needs verificationor correction.In the Central Plains region of Kansas and Nebraska, the post-Woodland archeological complexes for the most part represent peopleswhose subsistence economies were based to greater or lesser degreeon maize-bean-squash horticulture. Long before introduction of thehorse by European invaders, small unfortified villages composed ofrectangular semisubterranean earth-covered lodges were scatteredalong the smaller stream valleys from the Smoky Hill drainagenorthward through Nebraska and apparently into the Dakotas. The 54 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154 extensive excavations at Medicine Creek in a series of Upper Re-publican village sites afford a good insight into the nature of thisearly semihorticultural occupation of the Great Plains. It is note-worthy that neither the comprehensive work by the River BasinSurveys at Medicine Creek nor the widely scattered earlier investiga-tions by other agencies in and about the Kansas River Basin havedisclosed any satisfactory evidence of trade or other contacts betweenthese prehistoric Plains communities and their contemporaries in thePueblo region of New Mexico.Materials probably assignabh; to the Upper Republican horizonwere found during 1948 at Kanopolis, at Harlan County, at severalof the small proposed reservoir sites surveyed in the Lower PlatteBasin by the University of Nebraska Laboratory of Anthropology,and on one or two sites at Angostura Reservoir. With exception ofthe last-named occurrence, all of the localities listed are within thepreviously known range of the Upper Republican culture. Severalvariants apparently exist, and these may have temporal significance.It seems likely that the large mass of data obtained at MedicineCreek, under carefully controlled conditions and in a relatively limitedlocality, will make possible a beginning toward separation into moreor less distinct site or horizon complexes. It will be interesting tosee, when more intensive work has been done, whether the sites tenta-tively assigned to the Upper Republican at Angostura will continueto be so classified ; and also whether they show any evidences of horti-culture or, alternatively, are to be interpreted as seasonal huntingcamps for peoples normally resident farther to the east or south.The relationships of the central plains Upper Republican materialsto certain sites along the Missouri in the Fort Randall Reservoir area,where cord-roughened body sherds and incised rims of collared formbring to mind the pottery remains at Medicine Creek, are still to beworked out.On a later time level than the Upper Republican culture are severalsites tested or more intensively worked at Kanopolis, Harlan County,Angostura, and other proposed or potential reservoir areas. AtKanopolis, near the junction of Thompson Creek and Smoky HillRiver, two levels of a stratified site (14EW6) yielded pottery andother remains attributable to the protohistoric Great Bend aspect.This complex, which appears to represent a sixteenth and earlyseventeenth century Wichita (?) occupation, occurs at several largevillage sites, marked by middens, cache pits, and other evidences of afairly sedentary semihorticultural mode of life, on the southern tribu-taries of the Smoky Hill in McPherson County, Kans., and on tribu-taries of the Arkansas River in central and southern Kansas. Site14EW6, lying somewhat north and west of the previously known P^^-^^pS-j^j"""- ARCHEOLOGICAL SURVEY IN 1948?WEDEL 55 range of the complex, may be presmned to mark a periodic campsite, perhaps utilized by hunting or trading parties.In Harlan County additional information on the nature of the Dis-mal Eiver culture complex has been accmnulated. Dismal River siteshave been found so far mainly in the Sandhills and High Plains ofwestern Nebraska, between the 99th and 104th meridians and south-ward from the Niobrara to the Smoky Hill Valley in Kansas. AtHarlan County, as in previous work on Stinking Water Creek inChase County, Nebr., house patterns differing rather widely from thoseof the earlier Upper Republican peoples, and suggesting a less sub-stantial type of dwelling, were uncovered. To date, in this locality,no satisfactory evidence of horticulture, of contact with Whites, or oftrade relations with southwestern peoples, has been found. Most ofthe artifacts correspond nicely to those found previously at other Dis-mal River sites in southern Nebraska and in western Kansas, and therecan be no question as to the cultural relationships of the materials here.Ethnohistorical considerations leave little room for doubt that the Dis-mal River remains, which elsewhere have been found in such contextas to indicate a late seventeenth and early eighteenth century dating,are the remains of Plains Apache who were displaced southward aboutthe middle of the eighteenth century by arrival of the Comanche fromthe west.Because the year 1948 witnessed no extended excavations in any ofthe numerous village sites along the Missouri River in the Dakotas,there seems little point to adding further conjectures and guesses tothe literature on the area. The surveys carried on have merely con-iSrmed the extraordinary abundance and richness of aboriginal re-mains here; and preliminary sherd analyses have verified the viewsthat regional and temporal differences occur. Unfortunately, it isnot yet possible to translate the site-to-site or locality-to-locality differ-ences now apparent into a clear story of cultural growth. It wouldnot be surprising if something similar to the Woodland-Upper Repub-lican-Lower Loup-Pawnee sequence in the Nebraska region were event-ually demonstrated for the main stem in the Mandan and Arikaraareas, but much more thoroughgoing studies and more comprehensiveexcavations than have been made in the past will be necessary to testthis possibility. And, linked with this much-needed main-stem work,additional data must be gotten from the westerly tributaries of theMissouri before it becomes possible to tie in satisfactorily the pottery-bearing manifestations scattered westward to or beyond the BlackHills and the valleys of the Little Missouri and the Yellowstone.From such widely scattered, often sketchy, and very uneven data asare now available concerning the archeological remains of the MissouriBasin, no detailed synthesis of culture growth and human development 56 BUREAU OP AMERICAN ETHKOLOGY [Bull. 154can yet be made. The broad outlines of human history in the region,however, are becoming increasingly clear. It seems evident thatthroughout a span of time very likely to be measured in terms ofmillennia, simple hunting and gathering peoples dominated much ormost of the area. This occupation began at a time when large gameanimals of species that have since become extinct still roamed thewestern plains ; and there is geological, paleontological, and otherevidence that climatic and physiographic changes of appreciable mag-nitude have taken place during the period that has elapsed since. Itmay be assumed that the population groups during this stage weresmall and scattered, and that most of the energies of the peoples weredevoted to the getting of food, shelter, and clothing. Since mostknown sites are situated in or on stream terraces and in otherwise un-protected places, little is left of the material culture of the originaloccupants save items made of relatively durable materials. Theextent and nature of work in skin, basketry, wood, and other perishablemedia is mostly unl^nown, as is the physical appearance of the peoplethemselves. No positive evidence of the domestic dog has been foundin association with the Folsom, the so-called Yuma, or the later pre-historic hunting and gathering complexes of the Missouri Basinregion.In the semiarid sagebrush plains of central Wyoming, there isreason to believe that a late prehistoric occupation by a small-gamehunting and gathering people, who were possibly Shoshonean, tookplace. Further work at such reservoir localities as Edgemont, Key-hole, Moorehead, and Angostura ought to show how far eastward thisincursion of Great Basin peoples and economies extended. Elsewherein the short-grass plains of Montana, Wyoming, and adjacent areas, abasic economy probably consisting of big-game hunting combined withgathering lasted into the historic period, when the mode of life wasconsiderably altered and greatly enriched by acquisition of the horsein the late seventeenth and early eighteenth centuries.In the eastern portion of the Missouri Basin, along the Missouriitself and in the valleys of its major tributaries as far upstream as theYellowstone, primarily hunting and gathering economies gave way inprehistoric times to peoples who practiced food growing as well as foodgathering. It is not yet certain at what time period or cultural levelthe use of domestic plants here began. A few kernels of maize havebeen reported from a Woodland site in central Nebraska ; but the fewWoodland sites where comprehensive excavations have been carriedon have yielded no evidence of horticulture, except possibly in the caseof the Sterns Creek complex in eastern Nebraska. It is possible thatthoroughgoing investigations at other sites attributed to the Wood-land period, particularly in eastern Nebraska and in the Dakotas, will pft.N^o!*fr" ARCHEOLOGICAL SURVEY m 1948?WEDEL 57produce evidence of tillage or of domestic crops. Settlements of theWoodland peoples were small and unfortified; habitations were ofperishable materials and apparently of light construction ; and exceptfor the presence of limited amounts of pottery in the western Wood-land sites, there is little or nothing to suggest that the mode of lifefollowed differed very markedly from that of the late preceramic hunt-ers of the plains. At best, it would seem that incipient horticulturemay be suspected for the Woodland peoples whose remains have sofar been studied in the Missouri Basin west of the main stem.Following the widespread and as yet ill-defined Woodland occu-pancy of the trans-Missouri plains, with its doubtful or incipient hor-ticulture, came another of markedly dissimilar character. This one,provisionally dated to the thirteenth, fourteenth, and fifteenth cen-turies, is represented by numerous village sites assigned to the UpperRepublican and related complexes. Maize, beans, squash, and sun-flowers were cultivated ; hunting and the gathering of wild plant foodswere secondary subsistence sources; and fishing, including in somelocalities the extensive collecting of freshwater mussels, was also prac-ticed. The people resided in small, relatively permanent, unfortifiedcommunities of rectangular earth-covered lodges; and the clusteredarrangement of habitations in at least some instances suggests thatkin groups may have been the basis of society. No satisfactory evi-dence of community ceremonial centers has yet been adduced for thesesettlements. Pottery was much more abundant and better made thanthat of the preceding Woodland peoples ; and there were fairly well-developed and varied industries in stone, bone, horn, and shell. Thisprehistoric small-town stage is abundantly represented in the drain-age basins of the Kansas-Smoky Hill-Republican and the Platte-Loupsystems. There are some suggestions of a counterpart along theMissouri in the Dakotas; but its existence there, and its contempo-raneity, if it exists, with the Central plains materials, remain to bedemonstrated.In central Kansas and in east-central Nebraska, the small-townUpper Republican communities were apparently superseded in thelate fifteenth and sixteenth centuries by larger aggregations of semi-sedentary peoples who practised an intensive maize-bean-squash horti-culture, together with some hunting and gathering, and dwelt in vil-lages of circular grass houses or, north of the Kansas River drainage,of circular earth lodges. Considerable crop surpluses are suggestedby great numbers of large storage pits in the village sites; and it maybe presumed that these horticulturists traded some of their surplusto contemporary hunting peoples ranging the plains to the west.Somewhat later in time were the archeological manifestations termedDismal River, and several vaguely defined and unnamed pottery-mak- 58 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154ing complexes found from the Kansas Kiver to the Niobrara andprobably beyond. Regionally distinctive complexes, of which thereis evidence in the preceding small-town stage, are increasingly ap-parent ; and there is also indication of more extensive trade contactsthan can be demonstrated for the prehistoric period. These variousarcheological complexes, moreover, were late enough in point of timeso that in several instances more or less satisfactory correlations canbe made with historic tribes of the region, such as the Wichita,Pawnee, and Plains Apache.In the Central Plains, compactly built and fortified towns seem tohave developed after the close of the small-town period and after thegathering of people into larger communities was well under way. Asimilar tendency toward defensive measures is indicated on the upperMissouri. Among the longer established and more strongly horticul-tural village tribes of the eastern Missouri Basin, such as the Pawnee,Arikara, and Mandan, this period immediately before, during, andafter the first contacts between the Indians and Europeans seems tohave been the high-water mark of native cultural achievement in theMissouri Valley. Following it, as historical and archeological dataclearly show, came the swift rise to ascendancy of the horse-usingbison hunters of the western plains and the parallel decline in influ-ence of the older village cultures.In the broad view, it is obvious that hunters and primitive horticul-turists flourished for a long time in the Missouri River Basin, beforeas well as after the coming of the white man. It is also clear that thearcheological complexes representing these aboriginal occupancies,though relatively simple, varied considerably from locality to localityand from one time period to another. Here and there the variouscomplexes have been aligned in what are very probably correct chrono-logical and developmental sequences. There still exist, however,enormous gaps in our information. In the western portions of theBasin, including Montana, Wyoming, and adjacent areas, numeroussites have been located but very few have been systematically andcomprehensively excavated. Along the Missouri River in theDakotas, hundreds of village sites representing several centuries ofresidence by various tribes, likewise remain very inadequately known.Throughout the whole Basin, there are gTOwing indications that man'ssuccessive occupancies may correlate with climatic fluctuations whoserecords can be studied in stream terraces, buried soil horizons, windand water deposits, and other physiographic and geologic phenomena.There are excellent prospects, I think, that as archeologists expandand bring into sharper focus the story of native man's residence in theBasin they will be able to contribute information of importance tostudents of the various earth sciences. This, of course, is not the pri- K.N^o^.'lT'^' ARCHEOLOGICAL SURVEY IN 194 8?WEDEL 59mary goal of archeological research, any more than is the mere accu-mulation of great numbers of specimens; but it promises to be animportant byproduct of our investigations.As the Federal water-control program in the Missouri Basin speedsup and expands, the need for an equally accelerated scientific recoveryprogram is also accentuated. As dam after dam reaches completion,increasing numbers of archeological sites will disappear beneath therising reservoir waters. In many instances, as particularly along theMissouri River in the Dakotas, the remains that will be destroyedcannot be duplicated outside the reservoir areas. Unless these uniqueand irreplaceable segments of native American history are to be irre-trievably lost, funds for their study and partial salvage must be madeavailable promptly and in adequate amount. The time remainingfor this task is fast running out. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 3 Aerial view of excavations at 25FT17, Medicine Creek, Nebr Three prehistoric housesites and associated refuse areas shown here are now covered by the Aledicme L-reekDam (Neg 25FT17-124.) ^), Excavating a prehistoric house site at 23i< 1 1/, MedicineCreek Nebr. Note central fireplace basin surrounded by four large primary andnumerous smaller secondarv post molds, with vestibule entryway opening to south.(Neg. 25FT17-111.) BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 4 a. View southeast across excavations at ISFIVO, Medicine Creelv , Nebr. Square excava-tions at center have exposed four house sites; dam under construction at top and leftcenter. (Neg. 25FTOO-45.) b, Prehistoric house sites and other features uncoveredat 25FT70, Medicine Creek, Nebr. (Neg. 25FT70-11.) BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATES a, Pottery vessels as uncovered on floor of prehistoric earth-lodge site at 25FT17, MedicineCreek, Nebr. Upright charred tinabcrs at rear mark approximate wall of house pit.(Neg. 25FT17-147.) b. Excavations at Woodland site 25FT18, at junction of Limeand Medicine Creeks, Nebr. (Neg. 25FT18-48.) BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 6 a "^ " m bView south across excavations at iSF I'lS, Medicine Creek, Nebr. Control strips orblocks have been left standing at 10-foot intervals to show nature of culture zone andoverlying deposits. (Neg. 25FT18-22.) h. Fireplaces, post molds, and miscellaneoussmall'pits at base of Woodland occupation zone, \Iedicine Creek, Nebr. (Neg. 25FT18-16.) BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 7 ^A "^^ :^Wi Aerial view of Angcsiu... L'?;.. (center foreground; .luJ iu:i ;.: ;L;,e;...^: ...l.. ijc\(jnd,Angostura Reservoir, S. Dak. Circles indicate some of the archeological sites withinthe future pool area. (Neg. 39FA0O-20.) b, Bulldozer of Utah Construction Com-pany removing sod at start of archeological tests at 39FA10, Angostura Reservoir,S. Dak. Part of construction area visible in background. (Neg. 39FA10-2.) BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 8 ^?*- ''il^'V'^; a, Worknicn cxca\annt,' sii inc-iiiied tircplnce basin at .^yFAy, Aiit,'(jsiiira Reservoir, S. Dak.(Neg. 39FA9-15.) b, Deep hearth or roasting pit containing burnt stones and sootysoil, exposed during road construction; Angostura Reservoir, S. Dale. (Neg. 39FA68-2.) BUREAU OF AMERICAN ETHNOLOGY BUULETIN 154 PLATE 9 Excavated tipi ring at 39FA13, Angostura Reservoir, S. Dak. Stones marking tipiring are underlain in center foreground bv an older rock-filled hearth consisting in partof broken manos and grinding ''slabs. (Neg. 39A13-3.) b, Refuse beside aboriginalquart7ite quarry pits on Flint Hill, 39FA49, near head of Angostura Reservoir, S.Dak. (Neg. 39FA49-12.) BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 10 ."rsf^^m^t View west across Area A at the Long site, 39FA65, iVngostura Reservoir, S. Dak. Testcuts are visible to right of center; Horsehead Creek in background. (Neg. 39FA6S-10.)b, Excavations at Long site, 39FA65, Area A, Angostura Reservoir, S. Dak. Beside andjust above the charcoal-laden zone was found the basal fragment of an obliquelyflaked point, here shown in situ (small arrow). (Neg. 39FA6S-1L) BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 11 a, Grouped secondary burials as found in Mound 39CH9, formerlv situated in spillway lineof Fort Randall Dam, Charles Mix County, S. Dak. (Neg. 39CH9-32.) b. Buriedarcheological zone (line of white stones and bone) exposed in wall of ravine cuttinginto Missouri River terrace at 39ST23, about 3 miles below Cheyenne River, OaheReservoir area, Stanley County, S. Dak. (Neg. 39ST23-1.) H 2 &, -Ci : u; " ? ;dh ? ? :;? c u " t: tnt/,. o u : O ?* . r. QO:^ n! 4J ?: - SMITHSONIAN INSTITUTIONBureau of American EthnologyBulletin 154 River Basin Surveys Papers, No. 2Prehistory and the Missouri Valley Development ProgramSummary Report on theMissouri River Basin Archeological Survey in 1949By WALDO R. WEDEL 61 953842?53- CONTENTS PAGEIntroduction 65Personnel 67Laboratory activities 68Field work and explorations 70North Dakota 71South Dakota 74Wyoming and Montana 84Field work by cooperating agencies 87Kansas 88Montana 90Nebraska 91North Dakota 96Conclusions 97ILLUSTKATIONSPLATES13. a, Site 39FA10, Trench B, showing archeological remains at variouslevels; Angostura Dam in background. Fall River County, S. Dak.b, East test trench at site 39FA30, on Horsehead Creek, AngosturaReservoir, Fall River County, S. Dak 10214. a, Site 24PR2 at edge of sandstone cliff (foreground) ; looking southeasttoward Powder River and Bitter Creek. Moorhead Reservoir,Montana-Wyoming, b, Site 39PE10 (on knoll in foreground),looking west up Moreau River Valley. Bixby Reservoir site,Perkins County, S. Dak 10215. a, Chimney Butte from Manderson-Rockyford Road, with low terraceand site 39SH1 (arrow) in middle distance. Rockyford Reservoirsite. Shannon County, S. Dak. b, Site 39ST14 (Scotty PhiUipsRanch), a fortified Arikara village site in the work area just belowOahe Dam; Stanley County, S. Dak 10263 PREHISTORY AND THE MISSOURI VALLEYDEVELOPMENT PROGRAMSUMMARY REPORT ON THE MISSOURI RIVER BASINARCHEOLOGICAL SURVEY IN 1949 By Waldo R. WedelINTRODUCTIONContinuing its studies of the archeological and paleontological ma-terials that will be adversely affected by the expanding Federal water-control program in the Missouri River watershed, the Missouri RiverBasin Survey carried on its field and laboratory activities throughoutcalendar year 1949. For various reasons the year was an unusuallytrying one, even frustrating in some respects ; but within the limit ofavailable funds and in the face of a rapidly changing personnel pic-ture, a measure of progress nevertheless can be recorded.The Missouri River Basin Survey, now in its fourth year of opera-tion, was initiated in 1946. It is one regional phase of the nation-wide River Basin Surveys, directed by Dr. Frank H. H. Roberts, Jr.,Bureau of American Ethnology, for the Smithsonian Institution, andis supported by funds transferred by the Bureau of Reclamationthrough the National Park Service. The nature of the interbureauagreements which form the basis for all these investigations, and alsothe general background, organization, and objectives of the MissouriRiver Basin Survey, have been detailed in other papers.^ In briefestoutline, the purpose of the Surveys program is to locate, systematicallyrecord, and appraise the archeological and paleontological materialsthat will be lost as a result of the Federal water-control projectsplanned or under construction by the Bureau of Reclamation, De-partment of the Interior, and the Corps of Engineers, Departmentof the Army ; to pass on this information to the National Park Serv-ice ; to make recommendations, where needed, as to the steps requiredto insure recovery of a minimum representative sample of the scientific ^ See Smithsonian Misc. Coll., vol. 107, No. 6, April 23, 194T ; Amer. Antiq.. vol. 12, No.4, pp. 209-225, April 1947 ; and the annual reports of the Bureau of American Ethnologyfor 1945-46, 1946-47, 1947-48, 1948-49. 65 66 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154information that would otherwise be lost; and to direct the Federalphases of such subsequent detailed recovery work as may be pro-vided for, including limited or comprehensive excavations at key sites.The present report, fourth in a continuing series, briefly reviewsthe field and laboratory operations in archeology and paleontologyby the Missouri River Basin Survey during calendar year 1949. Likethe annual summaries which have preceded it this one is in no sensea complete and final report of accomplishments. Rather it repre-sents a statement of progress made during the period indicated ; andany interpretations advanced are subject to revision in the light ofmore penetrating analyses by the staff members and others on whoseindustry the present report is largely based. Also included are briefsummaries concerning the work of several States agencies cooperatingactively in the recovery of materials which might otherwise be lost.As in previous years, the Missouri River Basin Survey is indebtedto various organizations, agencies, and individuals for assistance ofmany kinds. Among the Federal agencies with which the Surveywas in particularly close contact, mention should be made especiallyof various officials in the National Park Service, notably in the ChiefHistorian's Office in Washington, and in the Missouri River BasinRecreation Survey Office, Region 2, Omaha ; the Bureau of Reclama-tion, including its regional and various field offices; and the districtand field offices of the Corps of Engineers. Among State and othernon-Federal agencies, the continued readiness of officials of the Uni-versity of Nebraska and its Laboratory of Anthropology to furnishspace at a nominal fee for the Survey offices and laboratory was anespecially appreciated courtesy. Survey field parties, as well as theLincoln office, were the recipients of innumerable courtesies and serv-ices from numerous project engineers, representatives and employeesof construction companies, and private individuals, to all of whom,though they go nameless here, sincere thanks are due. Last but byno means least, the continued efforts and unflagging interest of theCommittee for Recovery of Archeological Remains, which representsthe archeological profession of the Nation and various learned socie-ties, must be gratefully acknowledged.I have indicated that the year 1949 was a particularly trying one.There were several reasons for this. First and foremost was thequestion of funds. Reduced allotments by the National Park Servicefor fiscal year 1949, which included the first half of calendar year1949, and mandatory salary increases for which no supplementaryfunds were granted, meant that the working funds actually availablewere nearly 30 percent below those for the preceding year, from whichthere had been virtually no carry-over. The prospect of little or nocarry-over at the end of fiscal year 1949 meant that field plans forthe summer of 1949, which included proposals for excavation of se- p^p- No^-2^"'*- ARCHEOLOGICAL SURVEY IN 1949?WEDEL 67lected sites in Fort Randall, Garrison, Oahe, Tiber, Mullen, Yellow-tail, Harlan County, and Baldhill Reservoir areas, plus preliminarysurveys elsewhere, could not be set in motion because of uncertaintyconcerning congi-essional action on the appropriations bill on whichthe plans depended. By the time appropriations for fiscal year 1950(July 1, 1949 to June 30, 1950) were clear, the 1949 summer fieldseason was virtually over and it was impossible to organize any large-scale excavation projects. The Missouri River Basin Survey wasable to conduct limited field investigations in 1949 only because oftransfer of special funds, earmarked for specific purposes, by theBureau of Reclamation and the National Park Service. One of thefew bright spots in the 1949 picture was the fact that the increasedfiscal year 1950 funds, which finally became available in September1949, will perhaps ease the way toward planning and implementingof a large-scale excavation and survey program in calendar year 1950.Also complicating the Survey operations was a rather heavy per-sonnel turn-over, particularly among professional staff members.Owing in part to the straitened financial condition of the organiza-tion, i. e., reduced allotments in the face of increased operating ex-penses, two archeologists were separated from the project early in1949. Two others resigned, one in May and one in September. Thus,to the immobilizing of professionals who should have been in thefield was added the loss of others who would have been preparingreports on the work that had been accomplished previously. Asidefrom the resulting inevitable increase in backlog of unreported dataand materials, the effects on morale of remaining project personnelof all this uncertainty regarding the future of the program can per-haps be imagined. There is scant consolation, of course, in the reali-zation that other phases of the River Basin Surveys program werestruggling under the same, or similar, handicaps and uncertainties.PERSONNELSeveral changes took place in the professional staff of the MissouriRiver Basin Survey during 1949. Owing to a reduction in the fundsavailable for fiscal year 1949 as compared with the allotment for thepreceding year and to increased operating expenses for the project,it became necessary to reduce the staff soon after close of the calendaryear 1948. Early in January, Wesley L. Bliss, archeologist, was re-leased, and J. Joe Bauxar, archeologist, was transferred to Oklahoma.In May, Marvin F. Kivett, archeologist, resigned to accept the positionof Museum Director at the Nebraska State Historical Society; andin the same month, through transfer of special funds for specificsurvey purposes, Richard P. Wheeler was assigned to the Lincolnoffice. In September, Jack T. Hughes resigned to return to school ; 68 BUREAU OF AMERICAN ETHNOLOGY I Bull. 154and in November the staff paleontologist, Dr. T. E. Wliite, was tem-porarily transferred, to another project. At year's end, the profes-sional staff included three full-time archeologists : Paul L. Cooper;Robert B. Gumming, Jr., laboratory supervisor; and Mr. Wheeler.Full-time laboratory and office personnel included Dean Clark andA. E. Nixon, laboratory assistants; J. M. Shippee, expert laborer;George Metcalf, field and laboratory assistant; and Mrs. Ina MayReagan, clerk-stenographer. Drafting was carried on throughout theyear with part-time student assistance.Several students were employed as members of field parties. NeilIsto, Newell, S. Dak., accompanied Mr. Wheeler on a survey tripduring June and early July, and subsequently participated in ex-cavations at Angostura under Mr. Hughes. Also at Angostura asstudent helpers were Glenn Kleinsasser, Freeman, S. Dak.; andDonald K. Barnes, University of Missouri.LABORATORY ACTIVITIESThe field headquarters, staff offices, and laboratory of the MissouriFiiver Basin Survey were located throughout the year in the Labora-tory of Anthropology of the University of Nebraska. In anticipa-tion of a marked expansion of Survey activities in the near future,negotiations were under way at year's end for additional laboratory,processing, and storage space at the Lincoln Air Base. As in thepreceding year, storage space for field equipment and garaging forthe Survey's vehicles were maintained at the Base.In the laboratory, the basic project files, including site surveyrecords, artifact and photograph files, and other relevant data, weremaintained and expanded. Reserve files containing duplicate recordsof all material in the basic site files were also maintained.The reference map file, including topographic, geologic, land sur-vey, ownership and other types of maps frequently needed for plan-ning or other purposes, was expanded by 218 additional sheets duringthe year. It now totals approximately 985 maps.Drafting and map-making by the Survey have gone ahead steadily ; as in previous years, this was done largely with part-time studenthelp. Completed during 1949 were 20 profile sections and housefloor plans for sites excavated in 1948 at Medicine Ci'eek Reservoir,Nebraska ; site location maps for 10 reservoir areas, including Bixby,Boysen, Cannonball, Davis Creek, Edgemont, Harlan County, Mul-len, Onion Flat, Oahe, and Pomme de Terre; nine maps of sites inAngostura Reservoir area ; and one general map of the Missouri Basinshowing reservoir sites investigated for archeological and paleonto-logical remains to the end of calendar year 1948. Pap.No^2^"^' ARCHEOLOGICAL SURVEY IN 194 9?WEDEL 69In the darkroom 646 negatives were processed, including manyreturned from field parties and others made in the laboratory. Theselatter included upwards of 50 plates of selected typical artifacts fromMedicine Creek and Angostura Reservoirs, many of which are in-tended for illustrative purposes in forthcoming publications. Black-and-white prints made during the year numbered 1,392, in additioato 72 enlargements for publicity and other related purposes. Trans-parencies processed numbered 146, most of them colored. A con-siderable portion of the time of the darkroom assistant was givenover to the reflex copying of field records; about 2,000 sheets weremade up by this method and added in the appropriate location to theproject files.The relatively limited amount of field work in which Survey per-sonnel was engaged during the year, and the survey nature of muchof what was undertaken, resulted in an appreciably smaller thanusual flow of specimens from the field to the laboratory. A total of15,857 specimens was processed, i. e., cleaned, cataloged, and stored,during the year. The restoration and repair of outstanding or repre-sentative pieces was carried on to a limited extent, including therestoration of 16 pottery vessels from Medicine Creek and of onefrom Oahe Reservoir. The bead-laden skeleton of an adolescent fromthe Woodruff Ossuary (site 14PH4) in Harlan County Reservoirwas mounted for display, and has been placed temporarily on ex-hibit in the museum of the Nebraska State Historical Society in theState capitol in Lincoln.As in previous years, the laboratory sent out for expert examina-tion and report the animal bone, shells, and vegetal material shippedin at various times by field parties. From the identified materialreturned to Lincoln, a comparative collection of Missouri Basinmolluscan shells and animal bone has been started, in the hope offacilitating future tasks of preliminary identification. A similarseries of human skeletal remains for preliminary comparative workis also being started. It is gratifying to note that the accumulatingcollections of documented animal bone from various archeologicalperiods and localities throughout the Missouri River Basin promiseto contribute significant data toward elucidation of several specificproblems, including, for example, such matters as the evolution ofthe domestic dog, changing faunal assemblages and varieties through-out time and space, and the varied utilization of native fauna byIndian groups at various periods of time.With regard to publication and dissemination otherwise of in-formation gathered in course of field and laboratory studies, the Mis-souri River Basin Survey has continued to furnish the National Park 70 BUREAU OF AMERICAN ETIUSTOLOGY [Bull. 154Service with mimeographed appraisals of the archeological and pale-ontological resources of reservoirs investigated. During calendaryear 1949, three preliminary statements on reservoirs not previouslyreported were distributed; and two supplementary statements con-taining revised appraisals of reservoir localities previously reportedon were issued. Preliminary appraisals included statements on DavisCreek Keservoir site, in Nebraska ; on Cannonball Eeservoir, in NorthDakota; and on Philip Reservoir site, in South Dakota. Supple-mentary statements were issued on Dickinson Reservoir, NorthDakota, and on Shadehill Reservoir, South Dakota. Other prelimi-nary reports completed but not distributed at year's end included:Bixby Reservoir, South Dakota ; Moorhead Reservoir, on the Montana-Wyoming line ; Rockyford Reservoir site. South Dakota ; and OnionFlat, Raft Lake, and Soral Creek Reservoirs in Wyoming.Eight papers dealing with archeological salvage operations in theMissouri River Basin during 1948 were published in the April 1949issue of American Antiquity, quarterly journal of the Society forAmerican Archeology. These included three reports by River BasinSurveys personnel, four by representatives of cooperating agencies,and one by the writer as Field Director of the Missouri River BasinSurvey. Papers prepared by staff members of the Lincoln office andby the writer included the following titles : "Recent Investigations inFort Randall and Oahe Reservoirs, South Dakota," by Paul Cooper; "Investigations in Western South Dakota and Northeastern Wyo-ming," by Jack T. Hughes ; "Archeological Investigations in MedicineCreek Reservoir, Nebraska," by Marvin F. Kivett; and "Some Pro-visional Correlations in Missouri Basin Archeology," by Waldo R.Wedel.A summary report prepared by the then Field Director on the ac-tivities of the Missouri River Basin Survey during calendar year1948 was awaiting publication by the Smithsonian Institution as theyear ended.As in past years, the Lincoln office of the River Basin Surveyscontinued to furnish information, as requested, to the daily press, tovarious periodicals, and to other agencies and individuals. Arche-ologists in the field on several occasions addressed local groups in-terested in the excavations and findings in their particular localities.Staff members in the Lincoln office participated in the sessions ofthe Seventh Plains Conference for Archeology, held at the Universityof Nebraska, in Lincoln, on November 25-26, 1949.FIELD WORK AND EXPLORATIONSDuring calendar year 1949, field Avork was carried on in the MissouriRiver Basin by two units of the River Basin Surveys. These were pip. Na 21"'*' ARCHEOLOGICAL SURVEY IN 1949?WEDEL 71concerned exclusively with archeological researches, including surveysand excavations. Even this limited amount of field work, tragicallyinadequate in the face of the growing needs of the salvage program,were possible only because of transfer of funds by the National ParkService and the Bureau of Keclamation over and above the regularallotments received earlier in the fiscal year 1949.The survey work began on May 27, when Richard P. Wheeler,archeologist, left Lincoln on a 46-day reconnaissance of 10 reservoirsites in North and South Dakota, Montana, and Wyoming. Assistedafter June 2 by Neil Isto, student helper, Mr. Wheeler spent 1 to 4days each at the following projects : Bixby, Philip, Rockyford, andShadehill Reservoir sites in South Dakota ; at Cannonball and Dick-inson Reservoir sites in North Dakota; at Onion Flat, Raft Lake,and Soral Creek Reservoir sites in Wyoming ; and at Moorhead Reser-voir site on the Montana-Wyoming State line. The field work ter-minated early in July, and Mr. Wheeler returned to Lincoln on the11th of the month to compile reports on his researches and then toprepare for further field assignment.Under an emergency allotment received in June 1949 from the Bu-reau of Reclamation, the intensive surveys and excavations begunduring the summer of 1948 at Angostura Reservoir, South Dakota,were resumed in early July. Jack T. Hughes, archeologist, was againplaced in charge of this unit. He was aided in the field by GeorgeMetcalf, field assistant, J. M. Shippee, expert laborer, and a workforce consisting of students and local labor. On September 3, Rich-ard P. Wheeler replaced Hughes as field supervisor, and continuedthe excavations until November 8, when he returned to Lincoln forthe winter.Of the 11 projects where field work was conducted during 1949,8 had not been previously surveyed by River Basin Surveys personnelfor archeological materials. These included Bixby, Philip, Rocky-ford, Cannonball, Onion Flat, Raft Lake, Soral Creek, and Moor-head. Shadehill and Dickinson had been visited very briefly in 1946,and preliminary reports on their archeological possibilities were is-sued in 1947. All of the units listed above except Raft Lake weresurveyed in preliminary fashion in 1947 for paleontological remains,and brief reports on this phase of the salvage operations were issuedin August 1948. No paleontological field work was possible in theMissouri River Basin during 1949, owing to inadequate funds.NORTH DAKOTAArcheological work by the River Basin Surveys in North Dakotaduring 1949 consisted of reconnaissance at two Federal water-controlprojects. Cannonball Reservoir site, in Grant County, was surveyed 72 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154during the period June 10-14 ; and on June 15-17, additional investi-gations were made at Dickinson Reservoir, first surveyed by the RiverBasin Surveys in 1946. Both units visited lie on westerly tributariesof the Missouri. The field party consisted of Richard P. Wheeler,staff archeologist, and Neil Isto, student helper.Cannoriball Reservoir site.?The site of the Cannonball Reservoir, anirrigation and flood-control project of the Bureau of Reclamation,is on Cannonball River in Grant County, southern North Dakota, some60 miles west of the Missouri River. An earth-fill dam, now underconstruction about 5 miles southeast of Elgin, will rise 131 feet abovestream bed ; it will have a crest length of 1,690 feet, plus a dike sec-tion 4,310 feet long at the west end. The conservation pool (eleva-tion 2,222 feet, m.s.l.) will have a surface area of 6,760 acres; at super-storage (elevation 2,269 feet, m.s.l.), the area flooded will be about8,750 acres, and the pool will extend more than 25 miles upstreamfrom the dam.In the future pool area, the Cannonball meanders through a broadvalley bordered by bluffs and lined with well-developed erosional ter-races. The valley, as well as the rolling uplands to north and south,are covered with short grass. Trees occur sparingly, chiefly as small,scattered groves of cottonwood and box elder along the stream banksand in the adjacent coulees, with occasional juniper and aspen onthe steeper valley slopes.Approximately 4 days were devoted to the preliminary surveyof the Cannonball Reservoir site in 1949. Only the lower 2 milesof the future pool area were thoroughly inspected; above this sec-tion, spot checks were made to a distance of about 19^/^ miles abovethe dam site.Five sites of archeological interest, i. e., areas marked by surfacefeatures or refuse concentrations, were located and recorded, besideswhich a few scattering occurrences of stone artifacts and chips werenoted here and there on the ground surface. Four of the sites con-sisted of boulder circles or tipi rings. Two of these were marked bya single boulder circle each ; a third, 32GT20, in a borrow area nearthe dam site, consisted of 15 circles ranging in diameter from 15 to201/2 feet; and the fourth included two clusters of rings. No arti-facts were found on or near any of these tipi-ring sites.The fifth site yielded projectile point fragments, knife fragments,a stone drill, and miscellaneous chips, flakes, and spalls. No potterywas found, and limited tests revealed no artifacts or refuse belowthe sod. The site is believed to represent a temporary camp, but thereare insufficient data to permit an estimate as to its age or culturalaffiliations.The results of the 1949 reconnaissance at Cannonball Reservoir donot suggest a great abundance of archeological remains or materials Pap. No^'21"''' ARCHEOLOGICAL SURVEY IN 194 9?WEDEL 73of any marked antiquity. Since only about one-fifth of the futurepool area was examined however, it seems evident that further studieswill be required before a final appraisal of the archeological resourceshere can be made. Since initial water storage is scheduled for 1951or 1952, survey of the unexplored sections of the reservoir area shouldbe completed at an early date.On the basis of the 1947 paleontological survey, there appears tobe no need for further salvage work in that line at Cannonball, barringunexpected finds during construction operations.Dickinson Reservoir site.?The Dickinson Dam and Reservoir, anirrigation, flood control, and municipal water-supply project of theBureau of Reclamation, is nearing completion on Heart River inStark County, southwestern North Dakota. An earth-fill dam 49feet high and having a crest length of 2,270 feet is under construc-tion about 2 miles west of the town of Dickinson. At flood-controlstorage, the reservoir surface will be at 2,424.4 feet, m. s. 1., and willhave an area of 1,650 acres. The reservoir will have a length of about4 miles, occupying the Heart River Channel and the lower portionsof several tributary coulees.Like neighboring streams to the north and south, the Heart Rivermeanders through a moderately wide valley bordered by rolling up-lands. In the reservoir area, there are extensive bottom lands andlow terraces, frequently cut by old stream channels. Timber is sparseand of small size; where the land is not under cultivation, grasspredominates.Portions of the Dickinson Reservoir area were surveyed in pre-liminary fashion by a River Basin Surveys field party consisting ofPaul Cooper and J. J. Bauxar on August 17 and 18, 1946. At thattime, only two localities of possible archeological interest were noted : One on the north bank of the river just above the dam site, where afew modified flakes and refuse were found; the other on the southside below the dam site, where there were a few fragments of bone.Near this latter spot, according to a local informant, an Indian burialaccompanied by a fragment of pipestone was unearthed previously.In 1949, portions of 3 days were spent in further reconnaissance,chiefly in sections not visited during 1946. A few surface artifactswere found, including a side-notched projectile point, a drill point,a knife, and what is possibly an awl fashioned from the fibula of ahorse ; and from a cut bank on the east side of Ash Coulee, 30 inchesbelow the ground surface, were taken six bison bones. No occupa-tional, burial, or other sites worthy of detailed investigation wereseen by, or reported to, the Survey party.It is possible that the heavy brush and grass cover in the reservoirarea has obscured some archeological remains, or that older materialhas been buried by deposition. The generally negative results of the 74 BUREAU OF AMERICAN ETHNOLOGY IBull. 154two surveys made, however, suggest tliat no remains of importancewill be lost by construction of Dickinson Reservoir. Unless construc-tion or other developmental operations uncover remains of which noindication has so far been forthcoming, no further archeological workis recommended for this project.The 1947 paleontological reconnaissance indicates that no seriousloss of fossil remains will take place at Dickinson, unless construc-tion uncovers materials not now apparent.SOUTH DAKOTAMuch the greater portion of the 1949 field work of the MissouriRiver Basin Survey was carried on in various water-control projectsin South Dakota. All of these operations, including surveys at fourproposed and active projects as well as excavations at a fifth nownearing completion, were on tributaries of the Missouri west of themain stem.The surveys, beginning in late May and continuing into earlyJune, were made by Richard P. Wheeler, assisted after June 2 byNeil Isto, student helper from Newell, S. Dak. Leaving Lincoln onMay 27, Mr. Wheeler visited the following projects in South Dakota:Rockyford, on the White River in Shannon County, May 29-30;Philip, on Bad River in Haakon County, May 31-June 1; Bixby,on Moreau River in Perkins County, June 4-5; and Shadehill, onGrand River in Perkins County, June 6-9. From this point, theparty continued northward for further surveys in North Dakotaand elsewhere.In continuation of detailed investigations begun in 1948, intensiveresearches were carried on from July 10 until November 8, 1949, atAngostura Reservoir site, on Cheyenne River, southeast of HotSprings. The field party here was under supervision of Jack T.Hughes, archeologist, until September 3; during the remainder ofthe season, Richard P. Wheeler was in charge.Angostura Reservoir site.?Site of the Angostura Dam and Reser-voir, an irrigation project now nearing completion by the Bureau ofReclamation, is on Cheyenne River about 9 miles south and east ofHot Springs, in Fall River County, S. Dak. The dam, approximately130 feet high, will create a pool a mile or more wide extending south-ward nearly 4 miles, from which point it will narrow rapidly to thehead of the pool 6 or 7 miles farther west. An arm of the reservoirwill extend 3 or 4 miles to the southeast up the valley of HorseheadCreek, and there will be a shorter embayment in the lower valley ofDry Creek below Jackson Narrows.The archeological explorations here in 1948 had disclosed some 71sites in and near the future pool area. As indicated in a previous pip No^2l"'^' ARCHEOLOGICAL SURVEY IN 194 9 WEDEL 75report in this series, that for 1948, none of the sites showed evidenceof long occupancy, but there were good indications that many of thesites had been occupied repeatedly by small groups whose culturalequipment varied from period to period. For the most part, the sitesyielded only stone artifacts suggestive of hunting and gatheringeconomies; at three, small quantities of pottery were found. Therange of time represented is somewhat uncertain, but was believedto cover a span of perhaps several millenia, extending up to shortlybefore the historic contact period. Owing to the relatively low yieldfrom nearly all sites and the consequent need for extensive excavationon a large scale, the 1948 work consisted largely of what we may termintensive testing or sampling.The 1949 investigations represented an extension and intensificationof the work done in 1948. No attempt was made at complete excava-tion of any of the known sites. Instead, some 12 or 15 of the morepromising ones, which also appeared to be representative of the severalhorizons thought to be present in the reservoir area, were tested bymeans of trenches and pits. The artifact yield, in general, continueddisappointingly low ; but the data will doubtless be adequate to allowthe partial delineation of several more or less distinct complexes whichevidently existed at different periods of time.With initial storage of water scheduled to begin in the spring orearly summer of 1949, the first work was necessarily directed to furtherexploration of several nonpottery sites situated on terraces to beflooded just above the dam. These included particularly sites 39FA10,39FA56, and 39FA68, all lying about the junction of small waterwaysdraining Ked Canyon and Sheps Canyon a few hundred yards south-west of the dam. At 39FA10 (pi. 13, a) some 3 feet of sand was foundto overlie a hardpan; within this sand were three culture-bearingzones separated and capped with light sterile deposits, the lowest ofwhich lay directly on hardpan. Within the culture-bearing zones werehearths of various kinds, many flakes and chips, occasional bone frag-ments, and some artifacts, but no pottery. A few bits of graphite,presumably gathered for pigment, show utilization of one of the manymineral resources of the Black Hills. It is suggested that the threecultural zones represent occupation on as many separate occasionsby gi-oups whose cultural affiliations and time remain to be determinedwhen wider comparisons have been possible.At site 39FA56 there was evidence of two occupational levels. Theupper contained pit hearths or roasting pits, with which seemed to beassociated a few fragments of obsidian and occasional other materials;the lower included shallower basin-shaped hearths. Graphite isreported from this site also. 76 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154At site 39FA68, several trenches revealed a rather complicated situa-tion. At one point there is a cultural deposit approaching 3 feet inthickness, overlain by nearly 2 feet of sterile deposit. Elsewhere,trenches from 1 to 3 feet deep disclosed windrows of burnt stones,hearths, charcoal, bone fragments, scrapers, a few blades, and variousother remains. In the deepest test, a small lanceolate base-notchedprojectile point came to light at a depth of 6 feet. Two roasting pitsare also recorded from the site. The levels are somewhat irregular,and the stratigraphy is evidently complicated by a variety of geologicfeatures such as terrace fills, local gullying, etc. Further investiga-tions here in 1950 are urgently needed, if possible with the assistanceof geologists familiar with terrace problems.Another series of sites investigated were four locations on Horse-head Creek, which joins the Cheyenne some 3 miles south of the dam.Extensive test pitting was done at site 39FA23, one of the few sherd-bearing sites in the locality and situated about 3 miles up the creek,on its south bank. The site appears to be actually on an abandonedand partially filled creek chamiel. Here were found small projectilepoints, scrapers, knives, manos, cord-roughened potsherds, and animalbone. Hearths and ash lenses were rather plentiful, but no goodevidence of house sites came to light. Wliat seems to be indicated is aseries of brief occupations marked by thin discontinuous culture strataseparated by alluvial materials presumably deposited by periodicoverflows from the creek. There is no evidence of great age or pro-longed residence, or of marked cultural differences between levels. Thematerial found is somewhat reminiscent of the Upper Republicanhorizon, a pre-Wliite culture widespread in Nebraska, northern Kan-sas, and adjacent regions.Less than a mile upstream, where a small gully empties from thesouth into Horsehead Creek, is site 39FA30. Surveys here had shownseveral charcoal layers exposed in the shaly walls; and on the westside of the gully, where it joins the creek, several thick, cord-roughenedWoodland sherds and a small projectile point had been found insitu. Excavations at this spot disclosed several more hearths anda deeply buried fireplace, as well as another small point, but no moresherds. It is possible erosion had removed most of the pottery com-ponent prior to the 1949 work. The opposite, or east, gully wall ex-cavation (pi. 13, h) disclosed a few fragments of mussel shell, frac-tured pebbles, and some metate fragments dissociated from any ofthe charcoal lenses. The charcoal layers themselves proved unpro-ductive ; and it seems probable that, instead of being old occupationalsurfaces successively inhabited by peoples in a prepottery culturalstage, they actually represent burned-off grass surfaces subsequentlyburied by sand, gravel, and alluvium from the rapidly eroding patchof "badlands" nearby to the south at the head of the gully. pip. No^'21"'^' ARCHEOLOGICAL SURVEY IN 194 9?WEDEL 77An intrusive and poorly preserved flexed burial, associated with ashallow broken metate, was also found in the east slope of the gully.Its age and cultural affiliations remain undetermined.A few hundred yards downstream from the two sites just dis-cussed, the cut bank on the north side of Horsehead Creek showedseveral layers of charcoal containing animal-bone refuse. This lo-cality, thought to represent a possible prepottery site, was designatedsite 39FA61. Exploratory tests nearby yielded more bones, as wellas burned stone fragments, flakes, and similar detritus. Furthersurveys in the vicinity resulted in discovery of a well-marked hearthpartially eroded out of the south bank of the creek to the southeastof the earlier finds. Test pits showed cultural debris from just belowthe surface of the terrace to a depth of approximately 3 feet. Thisincluded miscellaneous chips and flakes, animal bones, shell frag-ments, drill points, scrapers, knives, a side-notched concave-base pro-jectile point and another of simple triangular form, and a number ofcord-marked and tooled potsherds from various depths. The scarcityof pottery-bearing sites in the Angostura locality, and the better tlianaverage yield of artifacts in tlie limited tests made, suggest the needfor further investigations at this spot.Among the more important finds of the 1948 season at Angostura,as reported in the annual summary for that year, were the evidencesof a fairly early prepottery occupation at the Long Site, 39FA65.This lies on the north side of Horsehead Creek about 3 miles aboveits mouth, in a proposed recreational area. Well flaked lanceolatepoint fragments were found here, on the surface and in situ under-ground, along with small circular fireplaces without stones, and quan-tities of paper-thin flakes from the manufacture of chipped artifacts.Despite the low yield of artifacts, it seemed evident that there wererelationships between these materials and those from other early sitesin the Great Plains, where artifacts of the Plainview type were comingto light.Further investigations at the Long Site in 1949 consisted of athorough re-examination of the entire site surface, and excavations atthe more promising exposures. At Area A, where hearths and pro-jectile point fragments were uncovered in 1948, the tip of anotherobliquely flaked point was found in situ. At Area B, on the eastedge and slope of a gully near the western part of the archeologicalzone, one well-defined fireplace and several possible hearths were un-covered, as well as a mano, but no points or other diagnostic artifactscame to light. At Area C, across the gully west of B, intensive testpitting disclosed charcoal fragments, occasional flakes, and bits ofwhat appeared to be oxidized wood, but no concentrations of occupa-tional debris were noted. Drift material in the form of fragmentaryand complete artifacts were picked up on the gully slope below Area953842?53 6 78 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154C, and traces of one or two small hearths occurred at the edge of theexposure. It seems likely that the area worked here was actuallyperipheral to an occupational zone that was very largely destroyedby the cutting of the arroyo.Despite the rather meager and unsatisfactory nature of the findingshere in 1949, there would seem to be a good possibility of the existencebetween Areas A and B of additional occupational debris. The areainvolved covers several acres, however, and further exploration with-out benefit of power machinery would seem to be of dubious value anddecidedly uneconomical. The stripping of overburden to whateverdepth is necessary to expose a large area of the old occupational levelwould perhaps result in a somewhat fuller and more adequate pictureof the situation.Several sites in the middle and upper portions of the reservoir areawere also tested during the 1949 season. These included two locationsat which pottery occurred sparingly, namely, site 39FA45 on DryCreek, and site 39FA48 on the south side of Cheyenne River abouta mile below the mouth of Dry Creek. Meager samples only wererecovered at these stations ; no structural features were noted. Far-ther upstream, in the general vicinity of Jackson Narrows, sites39FA35, 39FA38, and 39FA42 were briefly investigated. At thesecond of these, 39FA38, tests to a depth of 2 or 3 feet disclosed twodark soil zones separated and covered by light wind-blown sand.Chips, animal bones, a mano, a blade fragment, and other items cameto light here, as well as a sherd from the surface. No outstandingfinds were made at any of these sites.In addition to the sites investigated in the future reservoir area,brief trips were made from time to time in search of other remainsin the general vicinity. A number of interesting and hitherto un-reported petroglyph groups were recorded, chiefly along the CheyenneRiver above the future pool area. At one, designated 39FA78, thereare many and varied drawings, including several large anthro-pomorphic figures resembling dancers, each of whom has a zigzagserpentine figure dangling from the mouth in what suggests a repre-sentation of a snake dancer. There is a detritus-laden talus belowthe cliff on which these figures occur, and some of the figures extendto a depth of several feet below the surface of this talus. Charcoallenses and probable hearths occur at depths of approximately 2 and10 feet in the edge of the river terrace lying between the inscribedsurface and the Cheyenne River. At another site, 39FA79, a mile ortwo down river and just above the head of the reservoir area, severalgroups of figures done in varied style occur. Wliat is evidently arecent group includes a thunderbird, feathered lances, and numeroussmall anthropomorphic figures carrying gvms, all done in a pleasingly pip No^^l""^' ARCHEOLOGICAL SURVEY IN 194 9?WEDEL 79precise fine-line technique. Nearby are other groups done in ratherbroad shallow lines, and still others which are delineated by widelightly pecked lines. In a few places the sandstone surface has un-questionably been smoothed by rubbing in preparation for the draw-ings ; a mano found at the base of the ledge may well have been usedfor this very purpose. It would seem likely that the differences intechnique evidenced by these groups probably have chronologicalsignificance; but, other than the clearly late groups including repre-sentations of guns, there is no way of associating the groups at presentwith one or another of the prehistoric occupations with whose remainsthe Cheyenne River Valley abounds.The 1949 investigations at Angostura have amplified the pictureof native utilization of the area in prehistoric times, as judged fromprevious findings, although the broad outlines remain fundamentallythe same as at the end of 1948. In general, the region appears tohave been intermittently occupied by a long succession of nativepeoples throughout a period of perhaps several millennia. None ofthe sites yet investigated shows evidence of prolonged continuousoccupation, or of intensive pursuit of horticulture. The pottery-bear-ing sites have disclosed no satisfactory evidence of semipermanentstructures such as are found at sites of the semihorticultural peoplesfarther east and south. Such evidence may, of course, yet come tolight with more comprehensive excavation. The general impression,however, is one of frequent but short-lived occupations of the localityby culturally impoverished hunting and gathering peoples, or bylightly equipped seasonal foraging parties of peoples who normallyresided elsewhere. One would suppose from this that the rather variedresources of the Black Hills, including an abundance of native gameanimals and fur bearers, of wild fruits and berries, and of useful stonesand minerals, were perhaps as important to prehistoric men as theywere to the bison-hunting nomadic tribes?Cheyenne, Sioux, andothers?of the historic period.The strategic importance of the Angostura locality in working outthe prehistory of the northern plains west of the Missouri River isperhaps worth emphasizing again. This importance rests on thefact that here are to be found westward extensions of pottery horizonsmuch more fully represented to the east and south, as well as non-pottery manifestations with relationships to the west. If, from theresearches preceding actual flooding of the pool area, it becomespossible to v^^ork out a reasonably sure chronological and culturalsequence of the pottery and nonpottery complexes evidently present,then the somewhat nebulous picture of native man's occupancy of avast and little-worked region in the northern short-grass plains cen-tering in the Black Hills area, may be measurably clarified and 80 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154sharpened. The prospects for thus establishing geographically broadcorrelations seem promising enough to outweigh the relatively lowreturns from the current investigations in terms of artifact yield.Bixby Reservoir site.?The site of the proposed Bixby lieservoir,an irrigation and flood control project of the Bureau of Reclamation,is on Moreau River in northwestern South Dakota. The dam site isabout 1 mile west of Bixby, and the future reservoir area extendswestward up the Moreau about 24 miles. Dam specifications, reser-voir data, and construction plans have not been furnished the RiverBasin Surveys, and the project appears to be inactive at the presenttime.The proposed reservoir area lies approximately 100 miles west ofthe Missouri River, and less than half that distance northeast of theBlack Hills. Here the Moreau meanders through a relatively nar-row, bluff-bordered valley. Along its banks and those of its inter-mittent tributaries are clumps of cottonwood and box elder. Theadjoining uplands are treeless and, where not under cultivation, arecharacterized by a short-grass flora.In the 2 days available for preliminary survey at Bixby, only asmall portion of the area could be examined. Two sites of archeologi-cal interest were noted. One of these, 39PE10, is situated less thana mile upstream from the proposed dam site, atop a small knoll onthe south bank of the riv^r (pi. 14, &). Here were collected some50 fragments of pottery, sand- and grit-tempered, gray to buff incolor, and undecorated; a dentalium shell bead; a niano fragment;and a miscellaneous assortment of chips, blades, cores, and other odd-ments. The chronological position and cultural affinities of thismaterial cannot be determined from the evidence available at themoment.The second site lay about 2^/^ miles to the west, in a short arroyosouth of the river. A few small chipped flint artifacts, along withrefuse chips, flakes, and weathered animal bone, were found here onthe sandy, sloping surface of the arroyo. There were no sherds, andit is possible that a preceramic or nonceramic complex is represented.In view of the uncertain status of Bixby with regard to construction,no further work is recommended for the present. Should the projectbecome active, however, it would seem worthwhile to initiate testexcavations at the two sites found, and also to investigate thoroughlythe remaining portions of the future reservoir area.At present, there appears to be no need for paleontological salvagework at Bixby, unless construction discloses material not now known.Philip Reservoir site.?The site of the proposed Philip Reservoir,an irrigation and flood-control project of the Bureau of Reclamation,is on the North Fork of Bad River, in southern Haakon County, pip No^2t"^' ARCHEOLOGICAL SURVEY IN 1949 WEDEL 81S. Dak. The dam, for which alternate sites are proposed, is to be21/^ to 3 miles west of the town of Philip. The reservoir will floodportions of the North Fork and Mexican Creek Valleys for a distanceof about 514 and 2 miles, respectively, above the dam. The projectis inactive at present.The reservoir site lies about 100 miles west of the Missouri Kiverand approximately the same distance east of the Black Hills, along anatural route for hunting parties traveling from the main stemvillages to the hunting grounds of the upper Cheyenne River Basin.The stream valleys are narrow and thinly timbered with hardwoods ; the adjoining uplands are covered with short grass and cactus. Itis not a locality that would be expected to invite prolonged occupancyby semisedentary peoples.In the 2 days devoted to search of the proposed reservoir area, nosites or materials clearly of archeological significance were seen. Oc-casional chips collected on the ground surface may have been pro-duced by natural rather than human agencies. No further archeologi-cal work seems called for here, unless future construction activitiesbring to light promising materials of whose existence there is now novisible evidence or other record.The 1947 paleontological survey suggests that further search forfossils should be made in the Pierre shale exposures at Philip if theproject becomes active.Rockyford Reservoir site.?Rockyford Reservoir, a Bureau of Rec-lamation project for irrigation and flood control purposes, is proposedfor a site on White River, near the northern boundary of the PineRidge Indian Reservation in Shannon County, S. Dak. The dam isto be situated about 1 mile north of the town of Rockyford, whencethe reservoir will extend some 17 miles upstream. Dam specificationsand additional reservoir data were not available at time of the survey,and the project appears to be currently inactive.In the locality of the proposed reservoir, the valley of White Riveris cut into the fossiliferous Oligocene Wliite River formation whichhere, as elsewhere in the drainage of the White, gives rise to a sparselyvegetated and rather rugged badlands terrain. Erosional terracesare extensively developed in the locality, and include a flood plainmuch wider than would be expected from such a comparatively smallstream as the present White.The 1949 survey was restricted to an examination of both sides ofthe valley in the vicinity of the dam site, plus a search of the floodplain, terraces, and ravines on the east side of river upstream to thebend above the mouth of Wounded Knee Creek, some 4 or 5 miles abovethe dam site. The west, or left, bank of the stream in the future poolarea remains unexplored. 82 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154Three sites of archeological interest were recorded, and sample col-lections of surface materials were made at all three. Nonpottery ma-terials only are represented.Two of the sites, 39SH1 and 39SH2, are situated in terraces; thethird, 39SH3, lies atop a small flat-topped butte comprising the rem-nant of one of the higher terraces. Especially promising is site39SH1, which is on the east side of the valley approximately 3 milessouth of the dam site (pi. 15, a) . Here, about a mile west of a promi-nent feature known locally as Chimney Butte, a well-marked occupa-tional stratum covered by as much as 3 to 4 feet of sandy soil has beenexposed around the terrace edges. Eroding from this stratum areanimal bone refuse, chips, flakes, worked stone artifacts, and similarvestiges of former human activity. In the containing matrix may beseen bits of charcoal and occasional streaks of ash. In the two visitsmade to the site by River Basin Surveys personnel, in June 1948, andagain in May 1949, specimens collected included corner-notched pro-jectile points, blade and knife fragments, end scrapers, drills andgravers, mano (?) fragments, and refuse bone. The consistent ab-sence of pottery, as noted by all observers to date, and the topographicposition of the deposit at some depth beneath the terrace surface,suggests the possibility that a moderately early preceramic complexmay be represented ; but there are insufficient data at hand to warrantassignment of the materials to any of the known archeological com-plexes or horizons of the western plains.At site 39SH2 a few flakes, cores, and chips were collected on theeroded slope below the terrace surface, and in a nearby draw somebone fragments and chips were noted about 6 inches below the terracesurface. Neither this nor the third site, 39SH3, where worked stoneand refuse were noted only on the present ground surface, seem quiteso promising as the first one described.In the event that the Rockyford Reservoir project becomes active,it will be advisable to develop a program for further investigations inthe area. Because of their promise of yielding additional informa-tion on the little-known preceramic occupations of the Badlands lo-cality, extensive systematic testing and perhaps large-scale excavationwould seem advisable at site 39SH1 and possibly at 39SH3. Furthersurvey is also needed in those parts of the reservoir area not yetsearched, which lie mainly on the west side of the White River.In view of the extensive exposures of White River deposits andtheir high fossil content, further paleontological work may be neededhere if Rockyford becomes an active project.IShadehill Reservoir site.?The Shadehill Reservoir, an irrigation,flood-control, and water-supply project of the Bureau of Reclamation,will be located at the junction of the North Fork and South Fork of Pap. No^2^"'^' ARCHEOLOGICAL SURVEY IN 194 9?WEDEL 83the Grand Kiver in northeastern Perkins County, S. Dak. An earth-fill dam and dike, now under construction about 14 miles south ofLemmon, will have a maximum height above stream bed of approxi-mately 120 feet and a total crest length of some 13,000 feet. Theconservation pool, at an elevation of 2,272 feet, will cover approxi-mately 4,850 acres; superstorage at 2,312 feet will inundate approxi-mately 12,250 acres. At maximum pool level, the reservoir willoccupy about 3.5 miles of the main valley, plus some 23 miles in theNorth Fork Valley and 18 miles in the South Fork Valley.A preliminary visit was made to Shadehill by a River Basin Surveysparty in August 1946. In less than two days' work, the followingfeatures were noted: a possible tipi-ring site with flint chips, threeareas marked only by flint chips, two stone piles and two depressionsof unknown significance, and three deposits of bone that were thoughtto be of possible paleontological interest. No sites of great size orobvious importance came to light, but the desirability of further in-vestigations was indicated.The 3 days spent at Shadehill in May 1949 were devoted to a carefulrecheck of the area about the dam site and to a search of the previouslyunexplored portions of the pool area on the North Fork. A fewspecimens were collected, including four side-notched projectile points,two knives or scrapers, a hammerstone, chips, flakes, cores, and animalbones ; but there were no concentrations of material or other evidencesof occupational or burial sites.The apparent scarcity or absence of sites that would seem to warrantfurther investigation is rather surprising, since the valley of theGrand appears to offer a logical natural route of east-to-west andwest-to-east travel. The reservoir area lies less than 100 miles west ofthe Missouri, and would seem to be within the range of territorydoubtless visited by the Arikara, Mandan, and perhaps other tribesformerly resident along the main stem in northern South Dakotaand southern North Dakota. Ludlow Cave, from which chipped-stone implements, bone and shell artifacts, coiled basketry fragments,and potsherds suggesting "a generalized Mandan-Hidatsa type" havebeen reported, is approximately 60 miles to the west, on the head-waters of the South Fork. Surface deposits in the same site yieldedmetal projectile points, glass beads, and other objects evidently leftby hunting peoples of the historic period. One would suppose fromthis that the valley of the Grand would have been a frequently usedroute of travel in prehistoric and historic times, and that the banks ofthe stream would show plentiful evidence of that fact.It is possible, of course, that such sites as may once have existedhere have been destroyed by cultivation or are obscured by water- orwind-deposited soils and vegetation. At the moment, however, no 84 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154further archeological work seems to be called for; nor, judged by the1947 paleontological survey, will there be need for further investiga-tions along that line. Should construction activities or related opera-tions uncover archeological or paleontological remains not nowknown, the above recommendation would, perhaps, be subject torevision. WYOMING AND MONTANAArcheological researches by the Eiver Basin Surveys in Wyomingand Montana during 1949 consisted exclusively of reconnaissance atfour proposed Bureau of Reclamation water-control projects. Proj-ects visited, in all cases for the first time by archeological survey units,included : Moorhead Reservoir site, on Powder River in Montana andWyoming; and three small projects?Onion Flat, Raft Lake, andSoral Creek?in the upper Big Horn River Basin, in west centralWyoming.The proposed Soral Creek and Raft Lake Reservoirs lie within theWind River Indian Reservation. Grateful acknowledgment is heremade to the Business Council of the Shoshones and Arapahoes forgranting permission to the River Basin Surveys for the surfaceinvestigations.The work at these four projects was carried out by a field unit con-sisting of Richard P. Wheeler, archeologist, and Neil Isto, studenthelper. This party operated on the following schedule: Moorhead,June 19-23; Onion Flat, June 28-29; Soral Creek, July 1-2; andRaft Lake, July 5-7.Moorhead Reservoir site.?The site of the Moorhead Dam and Reser-voir, an irrigation and flood-control project now under constructionby the Bureau of Reclamation, is on Powder River in southeasternMontana and northeastern Wyoming. The dam, situated less than2 miles south of Moorhead, Mont., will be an eartli-fill structure 220feet high, with a crest length of 2,700 feet exclusive of a 550-footdike section at the west abutment. The flood control pool (elevation3,542.5 feet, m. s. 1.) will have a surface area of 18,300 acres, extendingapproximately 32 miles upstream from the dam through PowderRiver County, Mont., and into Sheridan and Campbell Counties, Wyo.Initial storage of water is scheduled for the spring of 1952.The Powder River rises in the arid plains of central Wyoming and,through its western tributaries, drains much of the eastern slope ofthe Big Horn Mountains. In the reservoir area, it is a silt-ladenstream meandering sluggishly through a wide, alluvial valley.Erosional terraces are well developed ; occasionally, in cut banks abovethe present flood plain, buried soil zones may be observed. Thevalley margins are hilly to rugged in character, often with "badlands"areas. Straggling stands of cottonwood, willow, box elder, and ash pip.No^2^"''" ARCHEOLOGICAL SURVEY IN 194 9?^WEDEL 85grow on the sand bars and along the stream banks, while juniper andyellow pine cover tlie "breaks" flanking the valley. Short grass ischaracteristic of the locality generall}^In the 5 days devoted to preliminary survey at Moorhead in 1949,a careful inspection was made of both sides of the river to a distanceof about 5 miles above the dam site, including also the lower portionsof Dry and Bitter Creeks. In addition, spot checks were made hereand there along the east bank of the stream to a point about 20 milesabove the dam site. The investigations resulted in the locating offour archeological sites.Two of the sites, lying on opposite sides of the river just above themouth of Clear Creek in the upper part of the future pool area,Sheridan County, Wyo., consisted of boulder circles or tipi ringsranging in diameter from 141/^ to 21 feet. No cultural material wasfound on or near either of these sites.Two other sites lay about a mile southwest of the dam site, on thewest side of the valley (pi. 14, a) . Cultural debris collected from thesurface included fragmentary projectile points, broken blades, scrap-ers, chips, spalls, and bones. No hearths or other features were noted,and limited tests disclosed no evidence of subsurface occurrence ofrefuse or artifacts. Both sites are thought to represent temporarycamps ; but in the absence of a definitive sample of the archeologicalcomplexes represented and until a chronological and cultural frame-work has been worked out for the region, none of the remains can beassigned to a definite time period or cultural horizon.None of the four sites seen are believed to merit further investiga-ation. In view of the relatively small portion of reservoir area exam-ined, however, and because the reservoir is situated in an area aboutwhich virtually nothing is known archeological] y, it is recommendedthat a comprehensive survey of the entire pool area be made beforeflooding takes place. The project lies in an area where little evidenceof semisedentary pottery-making peoples would be expected, but therewould seem to be an excellent chance of finding remains of preceramicor other nonceramic hunting and gathering complexes. Indicationsof such remains have been reported from the Yellowstone Valley, intowhich the Powder empties a few miles southwest of Terry, Montana,and also to the southeast in Keyhole and Edgemont Reservoir sitesin the Belle Fourche-Cheyenne drainage. It is possible that a pains-taking search of all cut banks and other exposures in the MoorheadReservoir area would disclose significant remains not now known toscience.In the limited time spent at Moorhead during the 1947 paleonto-logical reconnaissance, no identifiable fossils were obtained.Onion Flat Reservoir site.?The site of the proposed Onion FlatDam and Reservoir is on Devil Dive Creek, a small dry tributary of 86 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154the Little Popo Agie River about IQi^ miles southeast of Lander,Fremont County, Wyo. A dam some 1,680 feet long, with a dike sec-tion at each end, will impound water to be diverted from Little PopoAgie to form a pool extending approximately 1^/2 miles upstream.The future pool site is a broad valley with gently sloping sides anda shallow meandering stream channel, which is usually waterless.Trees are absent; the valley slopes and bottoms are covered with aluxuriant growth of sagebrush and short grass. In its present condi-tion, the locality would seem to offer little inducement to occupancy bynative peoples.Two days' search of the reservoir area by the River Basin Surveysfield party resulted in discovery of a single site of archeologicalinterest. This is situated about 225 yards northwest of the down-stream face of the proposed dam. Here, in a relatively restricted areasome 40 yards across, cultural materials were collected as follows:three projectile point fragments, two scrapers, a drill point, one com-plete and five fragmentary manos, five hammerstones, a chopper,cores, chips, and animal bones. No hearths or other structural or ha-bitational features were noted and limited tests disclosed no culturalmaterials below the surface. It is presumed that the site represents atemporary camp; its cultural affiliations and chronological positioncannot be determined from evidence at hand.No further archeological or paleontological salvage work is recom-mended for Onion Flat Reservoir site, unless construction activitiesdisclose important materials not now apparent.Raft Lake Reservoir site.?Raft Lake, source of the North Forkof the Little Wind River, is picturesquely situated about 19 milesdue west of the town of Wind River, in western Fremont County,Wyo. It lies just east of the Continental Divide, at an elevation ofabout 10,000 feet, and is about li/^ miles long. The proposed dam,in a deep gorge just below the present outlet, will raise the presentwater surface somewhat.Much of the lake margin consists of rugged-to-precipitous granitewalls, with little ground suitable for sustained human occupancy. Atthe northeast end there is a small sand-dune area. Narrowleaf cot-tonwoods, willows, and aspen fringe the lake, and the nearby slopesbear heavy stands of juniper, lodgepole pine, spruce, and fir.Two days were spent in examining the locality by boat and afoot.No evidence of aboriginal occupancy was noted.No further archeological salvage work is recommended for RaftLake Reservoir site.Soral Creek Reservoir site.?Site of the proposed Soral CreekReservoir is on the North Fork of the Popo Agie River, about one-quarter mile above the mouth of Soral Creek and some 18 miles byroad south of the town of Wind River, in Fremont County, Wyo. Pap.Na2^]"'^' ARCHEOLOGICAL SURVEY IN 194 9?WEDEL 87The reservoir will be a small one, extending upstream about 1.6 milesfrom the dam site. The valley to be flooded is narrow, and is bor-dered by sandstone cliffs and steep bluffs. The valley floor and slopesare covered with grass and sagebrush; a few pines may be foundalong the cliff tops and clumps of willow are scattered along thecreek banks.Search of the future pool area and adjacent slopes disclosed a fewscattered flint chips, two groups of animal bones without associatedcultural material, and some markings on the cliffs. The bones havebeen identified as those of dog, bear, deer or antelope, bison, cottontailrabbit, and possibly sheep. The chips may have been caused bynatural agencies. The cliff markings appear to be modern, perhapsthe initials of ranchers ; no Indian petroglyphs were found.There would seem to be no need for further archeological or paleon-tological salvage operations in the Soral Creek Reservoir area.FIELD WORK BY COOPERATING AGENCIESIt is a pleasure to report that several of the State-supported agencieswhich had participated in scientific salvage work during 1948 wereagain active in 1949. This cooperative work was particularly timelyin view of the very limited scope of operations that could be under-taken by the River Basin Surveys. In Nebraska there were three co-operating organizations working in as many different water-controlprojects. On a smaller scale the same sort of work was done in Kan-sas, Montana, and North Dakota. As in previous years, no Federalfunds were available to support these researches; all the operationswere on a voluntary and unsubsidized basis. It is gratifying to reportthat at year's end there w^ere indications of a substantial improvementin this latter respect, with procedures under consideration whereby theNational Park Service would provide financial assistance on a contrac-tual basis to qualified cooperating agencies.In the pages which follow, I have sought to present brief summariesof the outstanding findings by the several cooperating agencies in1949. It is to be hoped that more extended and definitive statementswill be offered in due course by those directl}^ connected with the work.My remarks here are based primarily on data provided in progressreports made from time to time by these agencies, in accord with theiragreements with the River Basin Surveys. In those few instanceswhere no such reports were furnished, my comments will rest largelyon bits of information gotten from field personnel and thus will beeven less complete than would be desirable. In a few instances it waspossible for River Basin Surveys representatives to visit the scenesof activity by cooperating groups. I am grateful to the agencies 88 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154concerned and to their representatives for making available such ad-vance data as are here offered. KANSASCooperative archeological investigations in Federal vs^ater-controlprojects in Kansas were carried on in 1949 as part of a larger fieldresearch program sponsored by the Museum of Natural History of theUniversity of Kansas. The work was done by students under thedirection of Dr. Carlyle S. Smith. It involved rechecking of certainsites in two proposed Bureau of Reclamation reservoir areas previouslyvisited briefly by River Basin Surveys personnel, namely, Cedar Bluffon the Smoky Hill River and Wilson on the Saline River ; and prelim-inary reconnaissance in a small portion of a third proposed reservoir,Glen Elder on the Solomon River, where no survey had previouslybeen attempted by the River Basin Surveys.Cedar Blujf Reservoir site.?The Cedar Bluff Reservoir, a projectof the Bureau of Reclamation, will be situated on Smoky Hill Riverabout 18 miles south of Wakeeney in Trego County, Kans. It is anearth-fill structure with a height of 204 feet and a crest length of12,515 feet. When full, the pool will be about 14 miles long and willhave a maximum width of about 2 miles. The terrain which willbe directly affected is part of the short-grass plains; small clumpsof Cottonwood and willow occur along the river banks, and juniper isto be found on the valley slopes.Of the four sites recorded here by the River Basin Survej^s partyin 1947, two were briefly reexamined in 1949. They are situatedabout one-half mile apart and approximately 2 miles upstream fromthe proposed dam site. One is designated 14T02, the other 14T03.The surface of the first was littered with flint chips, besides whichfire-cracked stones and occasional chipped-stone artifacts were noted.Among the artifacts were broken end scrapers, side scrapers, andknives, as well as a crudely made projectile point with broad stem andthick blade. Several small test pits revealed undisturbed subsoil im-mediately below the plowed zone ; no layer of refuse or occupationaldebris was manifested. It is possible that additional tests on a largerscale, made when crop and soil conditions are more favorable forsearch than at the time of the two brief visits in 1949, would disclosemore remains.At 14T03, chips and a few fragments showing secondary work werefound on the surface. No recommendation for further investigationhere has been made.In general, it does not seem likely that materials of great scientificimportance are threatened by the Cedar Bluff Dam and Reservoir. pip.No^'2^]"'' ARCHEOLOGICAL SURVEY IN 1949 WEDEL. 89In view of the virtual absence of data on the archeology of the locality,however, it seems advisable that a thorough search be made of allremaining unexplored portions of the future pool area.Glen Elder Reservoir site.?The site of the proposed Glen ElderDam and Reservoir, an irrigation and flood-control project of theBureau of Reclamation, is on the Solomon River in Mitchell County,north central Kansas. The dam site is just above the town of GlenElder. Specifications are not available, but it appears that the reser-voir will flood the Solomon Valley for some 10 or 12 miles to the westfrom Glen Elder, including both the North Fork and South Forkinto eastern Osborne County.With the assistance of a local collector, the University of Kansasparty located and recorded two sites that are subject to destructionand six outside the future pool but in the immediate locality.Both of the former are pottery-bearing sites, although pretty cer-tainly of distinct cultural affiliations and different temporal position,and both lie very close to what will be the work area if and when damconstruction gets under way. Site 14ML1 is on a lofty hilltop atthe west edge of Glen Elder, overlooking the Solomon River. Al-ready partly destroyed by a gravel pit at its southern edge, the sitewill undoubtedly suffer further damage and possibly complete de-struction from the dam, the north end of which will reach the presentgravel pit.In the limited time available for survey work here in 1949, onlytwo small test trenches were attempted. One disclosed a small basin-shaped pit containing limestone fragments and a few potsherds. Theother yielded a cache of planoconvex end scrapers and prismaticknives. A painstaking search of the surface, part of which was cov-ered with growing corn and alfalfa, netted several hundred sherdsand a number of stone artifacts. The work in general confirmed theobservations of previous investigators, also based on surface collec-tions and very limited test pitting, that the site is rather distinctiveceramically by comparison with other known village sites of Kansas.There are apparently close similarities to site 25HN39, recorded in1946 by a River Basin Survey field party on Prairie Dog Creek inHarlan County Reservoir site, Nebraska, and to various small sitesscattered through other reservoir areas in the Nebraska region. Ifconstruction is undertaken at Glen Elder, further excavation at 14ML1is urgently needed to define more clearly the nature of the archeo-logical complex here represented.Site 14ML8 is located south of Glen Elder and southeast of thepreceding site. It occupies part of a low terrace on the south bankof the Solomon. An abundance of mussel-shell fragments and oc- 90 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154 casional sherds of prehistoric Upper Eepublican type were foundhere. No test digging was attempted.Wilson Reservoir site.?Wilson Reservoir, a flood-control and ir-rigation project of the Bureau of Reclamation, is proposed for theSaline River in Lincoln and Russell Counties, Kans. The proposeddam site is at the eastern edge of Russell County ; no specifications forthe structure or for the resulting reservoir are available. It is esti-mated, however, that the reservoir will extend westward up the Salineapproximately 20 miles to a point north of the town of Russell. Anarm of the pool will extend southeastward up Hell Creek into thesouthwestern corner of Lincoln County.A very brief 2-day reconnaissance by a River Basin Surveys fieldparty at Wilson in 1947 had disclosed the location of six sites ofarcheological interest. At three of these, petroglyphs were noted.At a fourth, approximately one-fourth mile south of the proposed damSite, charcoal and animal bone, some of the latter burned, were foundat a depth of 56 inches below the terrace surface. At another loca-tion, on the north side of the valley near the head of the future pool,tests in one of a series of grass circles disclosed bone fragments andcharcoal at a depth of 8 inches. Near the mouth of Paradise Creek,which enters the Saline near the head of the reservoir, worked stone,shell and bone fragments, and a few small potsherds of unidentifiabletype were found. From these findings, made at a time of year whensurface conditions were not particularly favorable for collecting, itappears that further search will be in order before dam constructionis completed.The very limited operations of the University of Kansas party herein 1949 added little to the earlier record. Uncertainty regardingroads prevented a visit to site 14RU2 ; and a heavy fall of rain broughtto early termination an effort to recheck the nearby petroglyphs. Ifand when the construction project is finally set in motion, furtherarcheological investigations would seem to be in order.MONTANAIn Montana, a small field party from the Montana State Universityunder supervision of Carling Malouf devoted several weeks in Augustand early September to investigations in the Canyon Ferry area, some15 miles east of Helena. Preliminary surveys here in 1946 and 1947by field units of the River Basin Surveys had disclosed a total of 32sites in and near the reservoir area, not all portions of which could beexamined in the time then available for field work. A summary ofthe principal findings here to the end of 1947, together with a briefstatement as to the character of the region involved, appears in thepublished report of the Missouri River Basin Survey for that year. plp.No^"2T'^' ARCHEOLOGICAL SURVEY IN 194 9?WEDEL 91The 1949 investigations of the Montana State University here weredevoted largely to a recheck of sites reported previously in the futurepool area, plus reconnaissance downstream in search of comparativematerials. Approximately 12 sites not hitherto reported were ex-amined, and small quantities of surface material were collected. Asin the previous operations, it was found that most of the sites yieldedvery little surface material. Chips were plentiful in many localities ; but many of these were identified by the investigators as of natural,i. e., nonhuman, origin. A number of tipi-ring sites were located;numerous chips and a few projectile points associated as surface refusewith one of these groups raises the hope that a more or less distinctivepoint type may eventually be identified with the boulder circles.The surface remains recovered consisted mostly of lithic materials,i. e., of chipped and other stone. No traces of pottery came to light,despite careful search ; nor did the work uncover signs of any aborigi-nal structures, of rock carvings, or of stone alinements other than tipirings. Tests in several supposed rock cairns suggesting burial struc-tures yielded only negative results.The general impression left by the findings of the university partyAvas that of an area occupied only sporadically and for short periodsof time, perhaps primarily in connection with seasonal hunting andgathering activities. At no point were remains found in such abun-dance or under such circumstances as to suggest intensive or prolongedand continuous occupation by communities of any appreciable size.NEBRASKAThree State agencies were engaged in cooperative scientific salvageoperations in Federal water-control projects in Nebraska during 1949.During June and July, the second summer field session of the Uni-versity of Nebraska Laboratory of Anthropology, under the leader-ship of Dr. J. L. Champe, continued excavations begun in 1948 atarcheological sites in Harlan County Eeservoir area on the Republi-can River. From June until early September, the University ofNebraska State Museum carried on excavations at archeological sitesin Medicine Creek Reservoir under the immediate supervision of E.Mott Davis ; and in addition, had paleontological parties at MedicineCreek and Harlan County Reservoirs. The Museum's field operationswere under the general supervision of Dr. C. B. Schultz and W. D.Frankforter. The Nebraska State Historical Society, between June 5and August 13, tested several archeological sites in the proposed Mul-len Reservoir area on the North Loup River in Hooker County. Mar-vin F. Kivett, formerly on the River Basin Surveys staff and nowDirector of the Museum for the Historical Society, was in charge ofthis latter work. 92 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154A brief resume follows of investigations and findings by these agen-cies at eacli of the areas concerned.Harlan County Reservoir site.?This locality is situated in southernHarlan County on the Republican River just north of the Kansas-Nebraska State line, about 200 miles west of the Missouri River.It is of especial archeological interest because within the relativelyrestricted area to be flooded there are village sites evidently attrib-utable to four cultural complexes representing as many distinctaboriginal occupations of the region by pottery-making peoples.These cover a time span of at least seven or eight hundred years, andpossibly considerably more. Some of the earlier horizons appear tobe correlated with buried soil zones or old surfaces which, since aban-donment of the sites, have been covered with wind-deposited soils. Theprobability would seem to be very good that through the archeologicalremains some of these buried soil surfaces may eventually be dated,and that the suspected climatic fluctuations will then be fitted into alonge-range picture of prehistoric weather conditions.The 1949 operations of the University of Nebraska Laboratory ofAnthropology at Harlan County Reservoir were mainly a continua-tion of excavations begun in 1948 at site 25HN37. This site is lo-cated on the left bank of Prairie Dog Creek 3 miles from its junctionwith the Republican River and 5I/2 miles southeast of Alma, Nebr.Two house sites, resembling in most particulars the three previouslyopened, were excavated. One of these had evidently burned andcollapsed; portions of the charred timbers and other structural ele-ments lay on the floor in approximately their original relationshipto one another, thus providing exceptionally good information on thesize and construction of the original edifice. Noteworthy, too, wasthe finding of an iron axhead partially embedded blade down in theash-filled fireplace of this structure. This piece, of course, furnishesincontrovertible proof of contacts, direct or indirect, between the for-mer Indian occupants of the site and white men, and thus confirmsthe relative lateness of the site as judged from its relationship toother similar sites to the south and west. The source of this tradepiece, whether Spanish, French, or English, must await its exami-nation by a specialist. Other artifacts and debris recovered duringthe investigations were in general similar to materials recovered dur-ing the 1948 work. The site is assigned to the Dismal River culture,and probably represents a settlement of Plains Apache of approxi-mately the late seventeenth or early eighteenth centuries.Other sites in the vicinity were also investigated, including espe-cially an Upper Republican house ruin at 25HN11, a half mile east of25HN37. The evidence here, though somewhat confusing, suggeststhe possibility of more than one structure on the spot. Refuse and Pap.No^'sl"'^' ARCHEOLOGICAL SURVEY IN 1949?WEDEL 93cultural material of the usual prehistoric Upper Republican type, andconsiderable quantities of mussel shells, were also found here.The evidence obtained by the Laboratory party at Harlan Countyin 1949, added to that from the preceding year's work at the samelocality, will be a noteworthy increment to Central Plains prehistory,particularly as regards the new data for the Dismal River culture.Two representatives of the Univereity of Nebraska State Museumspent approximately two months in Harlan County Reservoir areasalvaging fossils from the core trench for the dam and also from theborrow pits. Some fish and reptilian remains were recovered fromthe Cretaceotis Niobrara chalk in the core trench ; elsewhere, Plioceneand Pleistocene deposits yielded fragmentary mammalian specimens.These latter are reported to be of much promise in terrace studies nowunder way by the Museum in the Republican Valley.Medicine Creek Reservoir site.?This locality is on Medicine Creek,a small but perennial northerly tributary of the Republican River inFrontier County, some 25 miles northeast of McCook. Extensiveexcavations were made here in 1948 by the River Basin Surveys incooperation with the Bureau of Reclamation. These resulted in ac-cumulation of a very large body of data concerning the prehistoricWoodland and Upper Republican occupations, the only two pottery-including complexes so far found within the reservoir area itself.In 1949 archeological and paleontological investigations were car-ried on in the same locality by the University of Nebraska StateMuseum, under the immediate field supervision of E. Mott Davis.Sites investigated include 25FT41 and 25FT42, both situated on LimeCreek, and site 25FT60, on the right bank of Medicine Creek belowthe mouth of Lime Creek. All of these are prepott?ry stations longantedating the pottery-bearing sites previously worked by the RiverBasin Surveys and the Nebraska State Historical Society in the samelocality.At site 25FT41, the lower occupational zone, now designated zoneC, which was worked by the Museum in 1947, was again explored.The limits of the zone of occupancy were extended somewhat, butlittle or no material of diagnostic value was recovered. At one place,bones and flint flakes were found about 4 feet above the lower zone,presumably indicating another habitation level. Approximately 8feet above zone C and perhaps 40 feet below the surface of the terrace,still another refuse-marked zone was found. A hearth with associ-ated burned bone, flint spalls, and several artifacts were recoveredhere. Most significant finds were two points apparently correspond-ing in all significant particulars to points found at Plainview, Tex.,in association with fossil bison, and since named Plainview points.Also present were the midsection of an obliquely flaked point, several953842?53 7 94 BUREAU OF AMERICAN ETHNOLOGY [Bull. 151 nondiagnostic scrapers, and knives. Pieces of charred wood foundin the terrace fill, both above and below the occupational zones, have jbeen prepared for Carbon 14 dating tests. According to the Univer-sity of Nebraska State Museum paleontologists in general charge ofthe work, the two (or three?) occupational levels at site 25FT41 occur "at the base of a Republican River Terrace 2 fill, tentatively dated aspre-Mankato (prior to climax of the Mankato glaciation)." No con-firmatory faunal or paleontological evidence has yet been adduced insupport of this dating.At site 25FT42, a hearth area with bones, scrapers, and blades waspartially uncovered. No diagnostic artifacts identifying the culturerepresented have yet come to light. Indications are that the site mayhave been occupied and reoccupied many times. Geologically, "thesite is believed to be contemporaneous in part with 25FT41."In the south bank of Medicine Creek some 3 miles north of 25FT41,a concentration of flint and bones was found about 7 feet below theterrace surface. Ten feet below this concentration was noted a soilzone with a hearth but without apparent artifact associations. Thepossibility of a third zone some 10 feet still lower down is suggestedby shell fragments.At site 25FT50, excavated in 1948 and subsequently reported inprint, some time was devoted to the gathering of faunal materialswhich might give further clues to age of the cultural deposits. Twolarge portions of bison skulls, a nearly complete deer antler, andother materials were recovered south of the portions worked for arche-ological remains in 1948. Flint chips were noted here, but there wereno diagnostic artifacts.Paleontological work at Medicine Creek included investigationsat two fossil quarries, 25FT40 and 25FT49, by students under super-vision of Loren M. Toohey of the Museum. Results were not as goodas those in previous seasons. Some new evidence was recovered, how-ever, including a complete skull of a fossil beaver. These quarriesare said to be of late Pliocene age. A test trench at site 25FT49 onMitchell Creek yielded bones of carnivores, rhinoceros, camel, horse,rodent, etc. Additional work has been recommended for this locality.Mullen Reservoir site.?The locality proposed for this project bythe Bureau of Reclamation is on the Middle Loup River in north-eastern Hooker County, in the heart of the Sand Hills region. Tenta-tive plans called for a dam at one of several possible sites 5 or 6 mileseast of the town of Mullen and north of State Highway 2. The pro-posed reservoir would extend approximately 7 miles northwest up theriver valley.Preliminary reconnaissance here by a River Basin Surveys fieldparty in the spring of 1947 had disclosed eight sites of varied age and pip No^^^"*^' ARCHEOLOGICAL SURVEY IN 1949?^WEDEL 95 origin, several of which appeared worthy of further and more detailedinvestigation. When the Nebraska State Historical Society was pro-gi-aming its cooperative salvage work in the spring of 1949, theMullen project carried a fairly high construction priority, and earlyresearch at archeological localities near the dam site seemed calledfor. Subsequently it was learned that the Bureau of Reclamationwork schedule had been revised and that construction at Mullen hadbeen indefinitely postponed. It was decided, nevertheless, that theHistorical Society would continue with its plans for work here, sincethe region is little known archeologically but has been visited fromtime to time by Society archeologists.The 1949 investigations of the Society at Mullen consisted of exten-sive tests at several Dismal River village sites and at one very promis-ing Woodland site. At 25H021, on the south bank of the riverdirectly on the axis of proposed dam site No. 2, several fireplaces,scattered post molds, and occasional small pits, but no house struc-tures, were uncovered. Stonework and pottery were not especiallyplentiful, but bone artifacts were relatively abundant. Four milesupstream at site 25H07, also on the south bank, artifacts of stone,bone, and pottery were collected, and also a single piece of iron. Thecultural zone here was covered by 1 to 5 feet of sterile sand and clay.Tests were also made at sites 25H09 and 25H024, both on the southbank of the stream between the two preceding sites. All of theseare assigned to the Dismal River culture, which is believed to repre-sent the Plains Apache of the late seventeenth and early eighteenthcenturies. It will be interesting to see whether and to what extenttlie material culture inventory at these Dismal River complex sitesin the Sand Hills region differs from that at such culturally relatedsites as 25HN37 in Harlan County Reservoir area. The distance be-tween the two localities is not great, being less than 150 miles air-line ; but the environmental setting of the Hooker County sites differsrather markedly from that of the Republican River Valley. Whenthe recent findings at both localities have been reported in full, ourunderstanding of the early historic Apache groups of the CentralPlains will be very materially enriched.The 1949 work at Mullen disclosed remains of at least two otherprehistoric complexes in the locality. One of these is a Woodlandmanifestation represented by an interesting site lying on proposeddam site No. 1, slightly more than a mile below proposed dam siteNo. 2 and the Dismal River site 25H021. Here were found heavygrit-tempered cord-roughened sherds, apparently from large vessels,a few points and other flints, occasional bone objects, and lenses ofcharcoal and ashy soil. The cultural stratum, first located in a road-side cut, sloped upward as the excavations continued back from the 96 BUREAU OF AMERICAN ETHNOLOGY I Bull. 154face, but the slope was more gradual than that of the present grass-covered ground surface. When work was discontinued, the sherdlevel was at a depth of 7 feet or more underground. Laterally, toeast and west, its distribution is restricted by shallow swales ; its ulti-mate limit southward under the rising valley slope, is undetermined.Surface remains seen during the surveys carried out from time totime suggest that Upper Republican materials, presumably inter-mediate in time between the prehistoric Woodland and the proto-historic Dismal River are also present here. No work was done inany sites that can be attributed to this horizon. Local collections seenincluded many and varied point types, including fragmentary Folsompoints. It seems certain that when thoroughgoing studies of thislocality are made a long sequence of occupations, probably beginningwith early hunter or lithic manifestations, and continuing through aseries of later prehistoric and protohistoric pottery and hunting hori-zons will emerge. Thus, despite their superficially uninviting ap-pearance, the Nebraska Sand Hills promise a rich return for anyonewilling to seriously investigate their archeological resources.NORTH DAKOTAOrganized archeological salvage operations proposed by the Uni-versity of North Dakota did not materialize on the scale of the pre-ceding season, owing to deficiencies of transportation and pendingpersonnel shifts. Dr. Gordon Hewes of the University's staff devotedseveral days to investigations in Garrison Reservoir above Sanish.Limited test trenching was done at site 32MN9 on the left bank of theMissouri River, about 11/2 miles above Little Knife River. This sitewas selected for further testing because preliminary reconnaissancein 1947 by a River Basin Surveys field party had disclosed evidenceof possible stratification of cultures. Pottery samples were collectedat that time from two zones?one in the upper 8 inches, the other from12^16 inches deep. These indicated at least two wares of ratherdistinct and different character, each associated with a definite type ofprojectile point. The markedly different character of the remainsfrom each of these two levels, and the clear separation of the twozones, strongly suggests two occupations. Since the site lay wellbelow the conservation pool level, further investigation seemedcalled for.The location of the test trenches opened by Dr. Hewes, in relationto those made previously by the Surveys party, has not been de-termined. However, the test trenches opened by Dr. Hewes failedto disclose any stratification comparable to that noted in the earlierwork, and no pottery was found. Other remains were so limited inamount that additional digging did not appear to be warranted. Sur- pip. Na ^l""^' ARCHEOLOGICAL SURVEY IN 1949?^WEDEL 97face observations at other sites in the vicinity were equally unre-warding. It would appear that the later investigations by Hewes mayhave been in another part of the site from that in which the first testswere made, and it is possible that extension of his explorations, whichtime did not permit, might have yielded different results more in linewith those of the earlier observers.CONCLUSIONSIn appraising the work of the Missouri River Basin Survey in 1949,as compared to that of preceding years, several considerations mustbe borne in mind. For one thing, there is a great and growing urgencyabout the scientific recovery program as a result of the rapidly ex-panding and constantly accelerating water-control program. On thepart of the construction agencies, projects already begun are beingpushed toward completion, and in some instances impoundment ofwater is well advanced. Elsewhere, new construction is getting underway at all too frequent intervals. This, of course, was to be expected ; and so when the River Basin Surveys program was set up, it was hopedthat the recovery operations could expand and go into large-scaleexcavation more or less simultaneously with expansion of the construc-tion work. Unfortunately, however, funds for field work other thansurvey have been difficult to get, and the proposed excavation projectsfor which money has been sought regularly have materialized only ina few exceptional cases. In 1949 the major obstacles to accomplish-ment on the scale called for were again factors largely beyond ourcontrol; namely, the allotment of funds inadequate to set in motionthe detailed plans that had been worked out in good time. It is tobe hoped that a substantial portion of these plans can be realized inthe coming year ; but since the program of the construction agencieswas not slowed down as drastically by fiscal uncertainties as was thatof the River Basin Surveys, the bleak fact remains that one fieldseason was largely lost to the organization specifically charged withresponsibility for the scientific recovery work in the Missouri Riverwatershed.The setback thus imposed was felt in some measure in all portions ofthe Missouri Basin, but has hit hardest in those sections where destruc-tion of archeological remains, at least, will be most complete and far-reaching?the valley of the Missouri main stem in North and SouthDakota. Here the work of the Corps of Engineers at Fort Randalland Garrison Dams is well advanced, and perhaps three or four moreseasons at most will see the structures completed to the point whereimpoundment of water and flooding of archeological sites will takeplace. A third major project, Oahe Dam near Pierre, S. Dak., is alsoexpected to get under way shortly. In all these, impoundment of 98 BUREAU OF AMERICAN ETHNOLOGY [Bull. 354water will result in the complete obliteration of hundreds of villagesites, burial grounds, and other features that cannot be duplicatedanj^where else in the region or outside it. No one even slightly ac-quainted with the many and varied remains in this area can viewtheir impending submergence except with grave concern and the hopethat a determined effort may yet salvage a worthwhile segment of thematerials facing destruction.Recognizing, then, the factors that operated to prevent a large-scale excavation program on the Upper Missouri and at certain smallerprojects in the Montana-Wyoming region, we may still say that theyear's activities produced some worthwhile results. The reconnais-sance of hitherto unsurveyed proposed reservoir sites in the Dakotas,Wyoming, and Montana disclosed, for one thing, several additionallocalities that will bear more intensive examination if constructionplans become active, and indicated at the same time that certain otherlocalities can probably be written off as unpromising from the stand-point of human and cultural remains. The intensive researches car-ried on at Angostura by the River Basin Surveys, and those whichwere conducted at Harlan County, Medicine Creek, and Mullenthrough the welcome cooperation of State agencies, have added signifi-cant data to our understanding of various phases of the pre-Whiteoccupancy of those localities. To these and other accomplishments inthe field must be added the less obvious and less easily appraisedachievements in the laboratory, where progress was made in the prepa-ration of scientific reports and in the organizing and preservation ofbasic data pertaining to previous and current researches.With reference to current field work and researches stemming there-from, the year's activities in the Missouri River watershed by the RiverBasin Surveys and the cooperating agencies again represented a widerange of interests temporally and spatially. Sites that are demon-strably early and may well have a multimillenial antiquity were in-vestigated by the River Basin Surveys at Angostura and by theUniversity of Nebraska State Museum at Medicine Creek. Materialsprobably or certainly on a later time level were examined at Angos-tura, Harlan County, Mullen, and at a number of other proposedand active projects in Kansas, South Dakota, North Dakota, Wyoming,and Montana. Most of the 1949 work in these latter reservoirs, asnoted in an earlier section, was of very brief duration and preliminaryin character, and so its evaluation must await completion of more in-tensive investigations than have yet been possible.Materials evidently bearing on the problem of "early man" in theMissouri Basin came under observation at two widely separate locali-ties. At one, the Long Site (39FA65) in Angostura Reservoir, theartifact yield has been distressingly low, but surface and subsurface Kno^^I""^' ARCHEOLOGICAL survey in 1949?^WEDEL 99finds, some of the latter in association with buried stoneless hearths,suggest relationships with the Plainview, Tex., finds, and with othersin the northern plains that apparently have a respectable antiquity.Unfortunately, there is virtually no bone refuse and so no direct com-parisons are possible with the excellent series of extinct bison remainsfound at the Plainview discovery site. As indicated in a previoussection, it would appear that a considerable part of the original LongSite has been destroyed by gullying, but it is still possible that ad-ditional information could be obtained by large-scale stripping of theoverburden on the stream terrace between the known outcrops ofarcheological materials.In Medicine Creek Reservoir, the discovery of projectile points ofunquestionable Plainview type in an occupational horizon at site25FT41, on Lime Creek, is of interest. That the site is as old as hasbeen claimed in press releases and preliminary reports, which give ita "pre-Mankato climax" dating, has not been demonstrated to the gen-eral satisfaction of archeologists and geologists. The fact thatevidence exists of a statification of cultural deposits, some of whichmay be older than the level in which the Plainview points were foundhere, lends added importance to the site.Very likely generally later in time than the Lime Creek and LongSite remains are the numerous nonceramic sites reported from manyparts of the Angostura locality, frequently in association with buriedsoil surfaces of unknown antiquity. These sites are marked byhearths of various kinds and by thinly scattered stone artifacts ; bonerefuse is scarce or absent. Under ordinary circumstances, the ma-terials that can be recovered from these stations are usually so limitedin quantity and kind as to preclude convincing segregation into mean-ingful sequential series. Through the limited use of earth-movingmachinery as well as hand labor, and the stripping of sizable parts ofthe old occupation surfaces, it has been possible at Angostura to gathersomewhat larger samples of the remains. To what extent these willpermit the delineation of distinct cultural horizons will become evidentonly after detailed laboratory analysis of the data recovered. Mean-while, we may note that more or less similar sites and materials havebeen recorded previously by River Basin Surveys personnel at manyother reservoir areas throughout the Wyoming-Montana region. It ispresumed that they represent the remains of a series of native huntingand gathering peoples who successively, but perhaps not continuously,inhabited the region through a span of time that probably coveredseveral millenia. It is perhaps unnecessary to observe here that notall nonceramic remains in this region are necessarily to be regarded aspreceramic or of great antiquity ; some of them doubtless represent thetraces of seasonal hunting camps of peoples of relatively recent periods 100 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154who dwelt at other times in more permanent villages situated else-where. There seems to be nothing in the archeological findings todate to contradict the view that the western plains in much of presentWyoming and Montana were occupied throughout most of pre-historic time, as in the historic period, by small bands of roving huntersand gatherers who had no fixed settlements comparable to those of thesemihorticultural peoples in the eastern plains.The investigations at pottery-bearing sites in Angostura, HarlanCounty, and Mullen Reservoir areas, has contributed useful informa-tion concerning some of the later prehistoric and protohistoric peopleswho resided in this region within, roughly, the last thousand or fif-teen hundred years. Discovery on Horsehead Creek at Angosturaof a site (39FA30) yielding heavy, cord-roughened sherds suggeststhe presence here of an early pottery-making people presumably re-lated to the Woodland culture, or cultures, so widely representedthroughout the trans-Missouri plains. Other buried material of evi-dent Woodland affiliation came to light at Mullen. At neither ofthese locations was it possible to obtain a very extensive series of arti-fact types or a clear picture of the sort of community and peoplerepresented.On a later time level are the materials disclosed by investigations inan Upper Republican village site at Harlan County and the perhapsrelated camp-site remains at site 39FA23 at Angostura. In HarlanCounty, the prospects seem excellent for getting not only a good defi-nition of the local Upper Republican horizon but also for establishingits association with an old buried soil surface which might eventuallybe datable thereby. Fortunately, the University of Nebraska Labora-tory of Anthropology which is carrying on the cooperative archeo-logical work at this locality, is cognizant of these prospects, and wemay hope that continued inquiries will throw further light on the in-terrelated problems of prehistoric archeology here and its inferredclimatic setting. At Angostura, it is not yet clear whether the UpperRepublicanlike site or sites represent only temporary hunting campsreoccupied briefly from year to year or alternatively are to be regardedas locally specialized manifestations of a prehistoric village-dwellingpeople who elsewhere subsisted in considerable measure by horticul-tural pursuits.The work done in 1949 at Dismal River culture sites in HarlanCounty Reservoir and at Mullen is significant because it ties in theupper end of an archeological sequence with early historic documenteddata bearing on Indian-White contacts. There remains little doubtthat these sites are assignable to a late seventeenth or early eighteenthcentury Plains Apache occupancy, and that the White contacts indi-cated were in the period when Spanish and French interests were im- pip.N^o^.'2T'^' ARCHEOLOGICAL SURVEY IN 1949?^WEDEL 101portant in the region. The data now at hand ought to contributematerially to a more accurate characterization of the mode of life andthe cultural equipment of the early historic Plains Indians, and per-haps also to a clearer perception of the antecedents from which the de-fined archeological horizon was derived.It is apparent from the foregoing remarks that the all-too-limitedarcheological field work conducted in 1949 in the Missouri Eiver Basinhas resulted in no startlingly new views or scientific concepts, nordid it disclose any hitherto totally unknown or unsuspected archeo-logical complexes. At the same time, it can truthfully be said thatnumerous details have been added here and there to help fill in thestill sketchy outline of human prehistory which has been graduallytaking on more substance as field work goes forward. There can beno question that our knowledge of Missouri Basin prehistory, demon-strably a long and involved story, will continue to advance as addi-tional data are gathered in the field. These data are badly neededfrom all portions of the Basin, from the sparsely peopled west as wellas from the more heavily populated east. With no intention of mini-mizing the scientific usefulness of what has been done or professionalinterest in the localities that have been worked intensively to date, itis still timely to record the sad fact that in the section of the MissouriBasin where the largest and richest remains of prehistoric peoplesoccur in greatest profusion, a large-scale and continuing excavationprogram still remains a thing of the future?a future, be it noted, thatwith a few more years of neglect will become archeologically forevera little-known and dimly understood past. 953842?53- BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 13 .-<*.; a, Siic I'U \1(), Trench 15, slKAvinu' archeological remains at various \c' ? : .'Stura Damin backtrround. Fall River County, S. Dak. (Neg. 39FA10-30.) o, I'.asi test trenchat site 39FA30, on Horsehead Creek, Angostura Reservoir, Fall River Countv, S. Dak.Neg. 39FA30-34.) BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 14 h a, Site 24PR2 at edge of sandstone cliff (foreground); looking southeast toward PowderRiver and Bitter Creek. Moorhead Reservoir, Montana-Wyoming. (Neg. 24PR2-2.)h. Site 39PE10 (on knoll in foreground), looking west up Moreau River Valley. BixbyReservoir site, Perkins County, S. Dak. (Neg. 39PE10-2.) BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 15 9f-^ - .J .^rI - ,| -- v-'-^'-'- ?t7 ? ** ^%m Chimney Butte from Manderson-Rockyford Road, with low terrace and site 39SH1(arrow) in middle distance. Rockvford Reservoir site, Shannon County, S. Dak.(Photo by R. G. Slattery.) h. Site 39ST14 (Scotty Phillips Ranch), a fortified Arikaravillage site in the work area just below Oahe Dam; Stanley County, S. Dak. SMITHSONIAN INSTITUTIONBureau of American EthnologyBulletin 154 River Basin Surveys Papers, No. 3The Woodruflf Ossuary, a Prehistoric Burial Sitein Phillips County, KansasBy MARVIN F. KIVETT 103 CONTENTS PAGEForeword 107Environmental background 109Previous archeological work 111Description of Woodruff Ossuary (site 14PH4) 113Artifacts 118Pottery 118Work in antler and bone 119Work in stone 121Work in shell 123Rdsum^ 126Cultural relationships 130Skeletal remains of the Woodruff Ossuary 137Literature cited 141ILLUSTRATIONSPLATES16. a, View to southwest toward Woodruff Ossuary, site 14PH4. b, Look-ing northeast across burial pit, site 14PH4, after the removal ofplowed soil 14217. a, Looking southeast across burial pit, site 14PH4, showing sectionwalls, b, Discoloration which marked Feature 1, a pit, at depth of12 inches, site 14PH4 14218. a, Looking south across burial pit at site 14PH4, after removal of alldisturbed soil, b, Feature 16, site 14PH4, a partially articulated adultskeleton 14219. c, Feature 8, site 14PH4, an area of scattered human bones showingsome articulation, b, Feature 9, site 14PH4, a pit filled withdisarticulated human bones 14220. a, Feature 21, site 14PH4, looking south. 6, Feature 21, site 14PH4,after removal to the laboratory in a plaster cast 14221. a, View in east area of burial pit, site 14PH4, showing typical disar-ticulated condition of bones, b, Sherds from burial pit, site 14PH4_ 14222. a, Bone and antler artifacts from site 14PH4. b, Stages in manufactureof shell disk beads, site 14PH4 14223. a, Chipped-stone artifacts from site 14PH4. b, Stone artifacts fromsite 14PH4 14224. Artifacts from site 14PH4 14225. Marine and fresh-water shell ornaments, site 14PH4 14226. a, Artifacts from various sites, b, Sherds from Guide Rock Ossuary(site25WT3) 14227. Restored vessel of the Harlan Cord Roughened type from site 25FR8_ 14228. Four views of male skull (14PH4-1436) from Woodruff Ossuary 142FIGURES2. Plan of Woodruff Ossuary after excavation 1103. North-south profile across Woodruff Ossuary on El line 118105 FOREWORDThe report that follows is based primarily on data collected forthe River Basin Surveys, Smithsonian Institution, during the sum-mer and fall of 1946.Other comparative data were gathered by personnel of the RiverBasin Surveys during the field season of 1948 in Frontier County,Nebr. Data for many of the sites came from the collections madeby A. T. Hill, which are now at the Nebraska State Historical Society.Many people assisted in the investigation of those sites.The excavation party during the fall of 1946 included J. M. Shippeeof the River Basin Surveys, A. T. Hill, director emeritus of theNebraska State Historical Society who, because of his long interestin this problem, spent considerable time at the site, and the writer.One local man, Carol Franke, of Woodruff, Kans., assisted part time.I wish to express our thanks to John Horrell, of Woodruff, Kans.,who granted permission for the excavations and assisted in manyother ways. For criticisms and suggestions as this report progressed,I am indebted to Dr. Waldo R. Wedel, Paul L. Cooper, and Dr.John L. Champe. Robert B. Cumming, Jr., has provided a prelim-inary study of the skeletal material. Identifications of shell speci-mens were made by Dr. J. P. E. Morrison and Dr. Harald A. Rehder,Division of Mollusks, United States National Museum. I wish tothank Dr. James C. Olson, superintendent of the Nebraska StateHistorical Society, for permission to use certain data from the col-lections of the Society. 107 THE WOODRUFF OSSUARY, A PREHISTORICBURIAL SITE IN PHILLIPS COUNTY, KANSASBy Marvin F. KivettENVIRONMENTAL BACKGROUNDPhillips County is situated in the Loess Plain area of north-centralKansas and borders on Nebraska to the north. The county is part ofa broad, gently eastward-sloping, loess-mantled plain, which has beenmodified by the valleys of Prairie Dog Creek in the northwest sectionand of the North Fork Solomon River in the southern half of thecounty. Numerous north-south drainages tributary to these streamshave carved the surface of the plain into a series of long, nearly par-allel, north-south divides which extend at about right angles to themajor streams. To the north, between Prairie Dog Creek and theRepublican River in Harlan County, Nebr., drainage channels arenumerous and the upland surface has been carved into a rather intri-cate system of steep-sided canyons separated by narrow and, in manyplaces, sharp divides.Prairie Dog Creek, a perennial stream, heads in Sherman County,northwest Kansas, and flows in a generally northeast direction to itsconfluence with the Republican River in Harlan County, Nebr. Thecourse of the stream is meandering, but the current is rather swiftso that the channel is gradually being deepened. A few springs occuron the valley slopes along the Republican River and Prairie DogCreek.Alluvial lands, which include the terraces and flood plains, occur incontinuous strips along both Prairie Dog Creek and the RepublicanRiver. The width of the alluvial lands, which is greatest along theRepublican, varies from a few yards to 214 miles. Lower terraces orbenches are well developed and usually occur at heights of 10 to 25feet above the bottom lands. The flood plains occur in strips ofvarious widths bordering both sides of the chamiels, and are subjectto inundation in places during periods of high water.The Prairie Dog Yalley is now utilized mainly for the growing ofcorn, but some small grain is also grown. The uplands and steeperslopes are often grass-covered and are given over to grazing. Nativetimber borders the stream bed ; cottonwood, willow, elm, ash, and boxelder are the main species. Wild fruits such as plums, chokecherries,and grapes are rather abundant along the stream course. 109 110 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154For a description of the area as it appeared a century ago, we turn toan account of John C. Fremont's second westbound expedition of1843. The party camped for 1 night on the stream which still bearsthe name they gave it, "Prairie Dog" (Fremont, 1887, p. 174). Theyentered the area from the south, after traveling up the North ForkSolomon River. Fremont observed bands of antelope and immenseherds of bison to the south and, on June 23, 1843, noted the following : At noon on the 23d, we descended into the valley of a principal fork of theRepublican, a beautiful stream forty feet wide and four feet deep, with a denseborder of wood, consisting principally of varieties of ash. It was musical witht"he notes of many birds, which, from the vast expanse of silent prairie around,seemed all to have collected here. We continued during the afternoon ourroute along the river, which was populous with prairie dogs (the bottoms beingentirely occupied with their villages), and late in the evening encamped on itsbanks.The prevailing timber is a blue-foliaged ash {Fraxinus near F. Americana),and ash-leaved maple. With these were Fraxinus Americana, cottonwood, andlong-leaved willow. We gave to this stream the name of Prairie Dog River.On June 25, while traveling west along the high smooth ridges southof the Republican River, the party observed "buffalo in great numbers,absolutely covering the face of the country" (Fremont, 1887, p. 174) . Since the advent of white men, large game such as bison and ante-lope have disappeared but considerable small game, including rabbits,opossums, raccoons, skunks, coyotes, prairie dogs, and pheasants, iscommon in the area today. There are some beaver and waterfowlalong the stream.It is evident that the area offered many inducements for primitivehorticultural peoples, as well as for those wholly dependent on theresults of the chase. Tributary streams such as Prairie Dog Creek,with a supply of wood, shelter, game, and arable ground, were readilyavailable. Flood-free benches and terraces along the streams pro-vided safe sites for villages. Sufficient timber was at hand for fire-wood and for lodge construction. The rich alluvial soils in the val-leys, sufficient rainfall, and a frost-free growing season adequatefor the maturing of such native-grown crops as maize, beans, andsquash, made the area especially attractive for groups dependentin part on horticulture. The vast herds of bison which roamed thearea would have provided a plentiful meat supply.During and probably for several centuries before the historicperiod, the area which includes the lower sections of the Prairie DogValley made up part of the Pawnee hunting grounds. Their nearesthistoric village appears to have been the Hill site (25WT1) ^ on the ^ Site designations used In this report are trinomial in character, consisting of symbolsfor State, county, and site. The State Is Indicated by the first number, according to thenumerical position of the State name in an alphabetical list of the United States ; thus,for example, 25 indicates Nebraska, 14 indicates Kansas. Counties are designated by atwo-letter abbreviation ; for example, HN for Harlan County, PH for Phillips County, etc.The final number refers to the specific site within the indicated State and county. N40E3 N N30E3 PLAN OF WOODRUFF OSSUARYAFTER EXCAVATION Scola in FaelNZ.^ LEGENDLine ot bose of plow zona ___ Line ot bottom of bosinX Locotion of feature MVP9695:!842 0-52 (Face p. 110 J PLAN OF WOODRUFF OSSUARYAFTER EXCAVATION Scol* in Fatt __ Lin* at bose o' plow ton* __ Lin* at bottom of botinX Location of fcotur* Figure 2.?Plan (.f Woodruff Ossuary after excavation. i?r>.H842 0-52 (Face |>. 110) "Bas.^Sur. WOODRUFF OSSUARY?KIVETT 111 ^">ublican Eiver in Webster County, Nebr., some 60 miles to the east.village, so far as can be determined, was visited in the summer180G by Lt. Zebulon M. Pike while he was en route to the head-iters of the Arkansas (Wedel, 1936, p. 17). Other historic groups ' ch as the Oto and Cheyenne also made some use of the valley.At least one village site, 25HN37, located approximately 8 milesdownstream, has been examined and assigned to the Dismal Riveraspect (Champe, 1949, pp. 285-292). This complex is believed tobe assignable to a late prehistoric and early contact period and maybe attributable to a Plains Apache group.PREVIOUS ARCHEOLOGICAL WORKThe area which comprises Harlan County and the valleys of theRepublican River and its tributaries to the east in Franklin andWebster Counties, Nebr., has been the scene of considerable archeolog-ical work. Prior to 1930, numerous sites were found in the Republi-can drainage by A. T. Hill, later director of the Nebraska State His-torical Society Museum.During the summer of 1930 the Nebraska State Archeological Sur-vey under the direction of Dr. W. D. Strong excavated sites in Web-ster, Franklin, and Harlan Counties, as well as in several othercounties throughout Nebraska. Excavations by the Survey in Har-lan County were confined largely to two burial sites on the southbank of the Republican Eiver. The first of these, the Graham Site(25HN5), is situated on the summit of a rounded hill, southwest ofthe junction of Prairie Dog Creek with the Republican River. Here,in a nearly circular basin, which was 23 to 24 feet in diameter and3 feet 6 inches deep in the center, were found human remains asso-ciated with a considerable number of artifacts. The majority of theremains were disarticulated, but two of the burials, both of infants,appeared to be primary interments. Scattered throughout the pitwith the human remains were numerous pottery sherds; rocks ofvarying sizes; fragments of charcoal; some bone, stone, shell, andantler artifacts ; and the remains of two copper-covered wooden disks.The cultural remains from the ossuary exhibited a close similarityto materials from sites of the Upper Republican aspect in the areaand were assigned by Strong to the Lost Creek focus of that aspect(Strong, 1935, pp. 103-114).One additional site of a similar nature in the area, the Alma Os-suary (25HN2), was excavated during the summer of 1931 by A. T.Hill and reported by Strong. The situation at this site, with itsdisarticulated human remains, is reported to have been comparable tothat at the Graham site. A comparison of the artifacts from thetwo sites indicated a very close similarity (Strong, 1935, pp. 122-123).&53842?53 9 112 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154A somewhat different type of prehistoric ossuary examined byStrong was the Marshall site (25HN1), located on a point of theriver bluffs 3 miles west and 2 miles north of the Graham site.Although the area had been badly disturbed by local collectors,Strong was able to find and excavate some small undisturbed areas.Disarticulated human bones, many of them broken and some burned,were associated in three areas or pits with an abundance of shell diskbeads and blanks. These pits were round or oval in outline and in eachcase were sunk at least a foot into the hard undisturbed soil at thebottom of the deposit. Some charcoal occurred in the fill, and manyof the beads were calcined.Shell disk beads were the predominating type ol artifact recoveredfrom the site; all of them appeared to have been made from theshells of fresh-water bivalves. They included finished beads, othersperforated but not ground on the edges, and rough unperforatedblanks. Other artifacts included two broken triangles of thin shell,with holes bored in two corners; a small bead with two intersectingperforations, one bored lengthwise and the other at right angles toit; and a fresh-vrater pearl with a single perforation. The only boneartifact was a broken and calcined object which Strong suggestedmay have been the head of an awl. Pottery, worked stone, copper, andmarine shells appeared to be absent (Strong, 1935, pp. 116-122).Another burial site, Holdrege 5 (25FR9), was investigated duringthe summer of 1934 by the Nebraska State Historical Society. It islocated on the left bank of Rebecca Creek about 2 miles south of theRepublican River in Franklin County. Like the Marshall site, itconsisted of a group of small pits which yielded broken and dis-articulated human remains, with which were associated nearly athousand shell disk beads and blanks. Broken shell pendants oftriangular form and a type reported to be similar to the "bear claw"form described by Wedel (1935, pp. 203-204) were found. The lattertype is represented only by fragments which also suggest a crescentform of pendant.Burial grounds similar in character to the Marshall and Holdrege 5sites appear to occur rather commonly throughout the upper Repub-lican drainage, both on the Republican River and on its tributaries.These sites, which are characterized by an abundance of finishedshell disk beads and a large number of blanks, have for the most partyielded few diagnostic artifacts which would aid in assigning theassociated burials to a cultural complex. A general relationship be-tween sites such as the Marshall Ossuary and those such as theGraham Ossuary is suggested by the fact that in both instances thereis reinterment of the mixed bones of the dead with offerings, andby the occurrence of some shell disk beads in the sites, like the GrahamOssuary, which also yield pottery. There is, however, considerable Riv. Bas. Sur. WOODRUFF OSSUARY KIVETT 113Pap. No. 3]difference between the types of artifacts from the shell bead ossuariescontaining thousands of blanks and those from the ossuaries whichcontain pottery of Upper Republican types and a few finished shelldisk beads. Only those ossuaries containing Upper Republican potterywere assigned by Strong to the Upper Republican aspect. The charac-teristic shell bead ossuary was placed in an unclassified category(Strong, 1935, p. 246).DESCRIPTION OF WOODRUFF OSSUARY (Site 14PH4)As it nears the Nebraska line in the vicinity of Woodruff, Kans.,Prairie Dog Creek has a meandering course in a general northeastdirection to its junction with the Republican River, some 12 milesdistant in Harlan County, Nebr. The valley here has a width ofnearly a mile, and numerous small tributary streams flow from thesouth across broad, level terraces. Little timber is present on theterraces and slopes, but the immediate banks of the stream supporta dense growth.The present stream channel is near the left valley slope where itis cutting into numerous spurs extending from the general terracesurfaces. It is on the point of such a terrace that site 14PH4 is sit-uated (pi. 16, a). At its south edge, the terrace drops some 30 feetto the channel of Prairie Dog Creek, while to the west and east, theslope is less abrupt to rather narrow but level flood plains, subject toinundation during periods of high water. The general terrace levelcontinues for more than a half mile to the north, where a transitionto the uplands is effected by means of a long gradual slope. Erosionis accelerated on this slope by cultivation, so that considerable soilhas been deposited on the terrace surface below.The Nebraska-Kansas State boundary line crosses the north edge ofthe site less than one hundred yards north of the ossuary. The fieldin which the site lies was in native grasses some 40 years ago, butsince that time it has been under cultivation, with corn the prmcpialcrop Less than a half mile to the east the stream has shortened itscourse by cutting off an ox bow, which tends to be swampy duringperiods of excessive rainfall. The general terrace level, with anelevation of 2,000 feet, is well above the limit of flooding and onlythe north edge would be subject to wash from the higher hills above.Although several archeological sites in the Prairie Dog Valley wereknown locally, none were recorded by research institutions above themouth of the stream prior to investigations by the River Basin Surveysduring the summer of 1946. At least one occupational site, 14PH5,which is situated less than a half mile east of 14PH4, had receivedsome attention from residents of the locality. This site appears to beassignable to a variant of the Upper Republican aspect. Although 114 BUREAU OF AMERICAN ETIESrOLOGY [Bull. J 54some surface materials have probably been collected at 14PH4, therewas no report or evidence of digging by local people.The site was first recorded during the course of a brief preliminaryarcheological reconnaissance of the proposed Harlan County Eeservoirarea during the month of August 1946 by the River Basin Surveys.Limited tests in a surface concentration revealed dark soil containingcharcoal, human bones, shell disk beads, and mussel-shell fragments toa depth of more than 4 feet over an area approximately 20 feet in diam-eter. Although surface examination revealed an area of darker soiland a concentration of cultural materials, there was no evidence of amound or depression. Such features, if present, may have been ob-literated by 40 years of cultivation.During the period October 17 to November 11, 1946, an intensive in-vestigation of site 14PH4 was undertaken by the River Basin Surveys.After additional tests had indicated that the most promising locationfor excavation lay at the southwest edge of the terrace, the area waslaid out in 5-foot squares. A north-south base line was established,starting 25 feet south of the area of dark soil and continuing northbeyond its apparent limits for a total distance of 60 feet. The southend of the line was designated "N zero." At intervals of 5 feet fromthe "N zero" stake, the line extending north was set off by stakes andnumbered consecutively in 5-foot units preceded by the letter N ; thussquare "N25" would have its southeast corner 25 feet north of the "Nzero" stake. These points served to designate the 5-foot squares westof the line and each square was designated by the number on the stakein its southeast corner. Sections in the first row of 5-foot squareswest and east of the base line were designated by the N series of num-bers to which were added Wl, W2, W3, and El, E2, E3, respectively,depending on the location of the southeast corner 5, 10, 15, or more feetwest or east of the base line. Thus, the southeast corner of a squaredesignated N35 E2 would lie 35 feet north and 10 feet east of the"N zero" point. Similarly, the southeast corner of square N25 W3 lies25 feet north and 15 feet west of the "N zero" stake.Within each 5-foot square the soil was troweled out in 12-inch levelsto within 2 inches of the outer limits of the square. This 2-inch sectionwas left standing and, together with the 2-inch section from theadjacent square, produced a central block 4 inches wide. Dis-articulated bones and scattered shell beads and blanks, which were rela-tively abundant, were sacked according to the 12-inch level in whichthey occurred. Less common materials, such as concentrations ofhuman bone, stone and bone artifacts, sherds, and significant shellartifacts, were sacked separately with exact horizontal and verticalprovenience recorded.The total area excavated measured slightly more than 500 squarefeet, and varied in depth from 12 inches at the outer edges of the Rlv. Bas. Sur.Pap. No. 3] WOODRUFF OSSUARY?KIVETT 115 ' excavation to slightly more than 6 feet in the deeper sections. Profileswere drawn and photographs were taken of the 4-inch control sectionsleft at the edges of the squares. Prior to any deeper excavation ofany squares, the entire area of discoloration was cleared of soil dis-turbed by cultivation. This loose soil, which varied in depth from 6to 8 inches, contained a quantity of human bone fragments, brokenmussel shells, and shell disk beads, as well as partially decayed cornstalks. The general outlines of a roughly oval area of dark soil, whichwere then discernible, were photographed and mapped (pi. 16, 6).Table 1. ? Summary of dimensions and contents of pits at site I4PH4 116 BUREAU OF AMERICAN ETHlSrOLOGY [Bull. 154in the west side of square N40 and in the southwest corner of squareN25. A similar system of excavation was extended throughout thearea of discoloration (pi. 17, a). Upon the removal of all disturbedsoil, the limits of an oval basin were discernible (fig. 2) . The presenceof some individual pits which occurred beneath the basin was suggestedby a greater concentration of human bones and shell beads traceablefrom the surface to the bottom of several pits (pi. 17, h). The exactlimits of the majority of the pits could not be determined, however,except where the bottoms extended some distance into the yellow loesssoil below the general area of mixture. The depth to which these pitspenetrated into the subsoil varied considerably and it is possible therewere other pits which did not extend into the yellow sterile soil andwere not discerned. The walls of several pits which intersected thelarge basin and extended beyond its general limits were discerniblejust below plow level. Cultural materials from individual pits belowthe basin level were segregated by levels and sacked separately.Feature 15 (table 1), an oval pit, was located outside the limitsof the large basin at the southwest edge (pi. 18, a; fig. 2). Testsaround the edge of the basin, as well as elsewhere on the terrace,failed to reveal other disturbed areas.With the exception of one skeleton (Feature 21), all of the burialswere secondary. A second burial (Feature 16) exhibited some articu-lation, but certain of the bones were missing (pi. 18, h). In otherisolated instances, lower arm and hand bones, vertebrae, and bonesof the foot were articulated (pi. 19, a) . Skulls were few in number and were in most cases in a poor state ofpreservation, so they were of little aid in determining the total numberof individuals present. A count of individual bones indicates thatsome remains of at least 61 individuals were placed in the basin andpits, and it seems likely that the total number was somewhat greater.Some remains were evidently destroyed during the last 40 years bycultivation, while others may have decayed.With the exception of the one fully articulated individual and a par-tially articulated burial, all of the bones appeared to have been de-posited in a haphazard manner; there was no evidence of bundleburials (pi. 19, h). Many of the bones were broken and some wereburned. Calcined shell disk beads, blanks, and mussel shells werecommon, and several bone artifacts show evidence of fire. Skeletalremains represent individuals of various ages from infants to oldadults.Bones of the single flesh burial, an adolescent, were well preservedand showed no evidence of fire (pi. 20, a). This may be accountedfor by its greater depth on the floor of the basin in square N25 El, 35inches below the surface. The skeleton was semiflexed, and lay on the '^H PfipNo^T"' WOODRUFF OSSUARY KIVETT 117 el; left side facing to the west, with the skull to the south. Extending e!' around the skeleton in the pelvic area were rows of shell disk beadsin alinement. Other rows of beads extended up the chest and aroundthe neck. Many of the rows were of well-ground and evenly matchedbeads, while others included bead blanks that had been perforatedand evidently strung but had not received the final smoothing whichcharacterizes a finished bead. Triangular shell pendants occurredon both the upper and lower sides of the skeleton, particularly in thevicinity of the skull (pi. 20, &) . Specimens associated with this burial,other than the abundant fresh-water shell disk beads, include workedsections of marine shells and one bone implement. This implement,which projected from beneath the distal end of the left humerus, isfashioned from the tibia of a deer. It is perforated near the proximalend and suggests a shaft-wrench type of tool or possibly a diggingstick handle.The general outline of the basin was symmetrical except where itwas interrupted by smaller pits. The slope of its walls was rathersteep on the north but more gradual on the other three sides. Al-though charcoal fragments?some of which appear to be elm?rangingup to 6 inches in diameter were common, there was no evidence of postmolds. Charred timber sections lying horizontal were commonthrough the basin and most of the pits, but small charred twigs andbranches were more common in the lower sections of these features(fig. 3). Some areas of burnt earth occurred in the upper portionsof the basin fill where the charcoal fragments were most plentiful,but there was no evidence of prepared hearths or prolonged burning.The calcined condition of the bones and artifacts appeared to haveresulted from burning within the basin with no indication of plannedcremation.Many disarticulated human bones were scattered throughout theburial area (pi. 21, a). These apparently represent secondary burialswith which shell disk beads, ornaments, and other artifacts may havebeen associated originally. The majority of the human remains arewell preserved ; others are in a fragile condition which may have re-sulted from exposure of the corpse on scaffolds or in trees. Afterdecomposition of the flesh, the scattered beads and weathered boneswere presumably gathered up and deposited in the various pits. Otherbones which show little evidence of exposure may have been exhumedfrom graves elsewhere and brought to a central burial pit for finalinterment. The single articulated skeleton lying on the floor of thebasin may represent an individual who was placed directly in the com-munity burial area without undergoing exposure or previous separateinterment. The abundance of shell beads and other artifacts scat-tered throughout the basin fill suggest that other corpses may at onetime have been comparatively well adorned. 118 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154The occurrence of some bones in their natural alinement resultedfrom sufficient tissue remaining on the bones at the time they wereplaced in the pit to maintain their articulation. The rather numerousoccurrences of from 3 to 20 evenly matched beads in close alinementprobably indicates that a majority of the beads were originally instrings like those with Feature 21 and were disarranged during thesubsequent period of exposure and reinterment. The unperforatedshell bead blanks, which could not have been strung, may have beenscattered over the various individual burials or placed in bags.^ Thereis some indication at the Woodruff Ossuary, as well as at other sitesof a similar nature, such as Guide Rock (25WT3), that a majorityof the shell ornaments were associated with infants and adolescents.ARTIFACTSStrong has emphasized the absence of pottery and artifacts of mate-rial other than fresh-water shells in the ISIarshall Ossuary. "It [theMarshall Ossuary] was truly a 'shell bead burial' and therefore differ-ent from the Graham Ossuary, both in content and structure"(Strong, 1935, p. 121).The Woodruff Ossuary was apparently similar in construction andcontent to the Marshall Ossuary with its predominance of shell arti-facts; in particular, both contained bead blanks as well as finishedshell disk beads. The Woodruff excavation, however, yielded a greatervariety of artifacts. These include pottery sherds ; artifacts of stone,bone, and antler; and objects fashioned from both fresh-water andmarine shells. Artifacts occurred throughout both the large basinand the smaller pits. POTTERYThe rather scanty pottery remains occurred at various depths in thegeneral basin fill, as well as in the individual pits. The sherds,which are all from the bodies of vessels, are uniform in appearanceand appear to represent a single pottery type (pi. 21, b). They con-sist of 10 sherds which vary in diameter from 7 to 50 mm. There issufficient variation in the characteristics of the sherds to suggest thatmore than one vessel is represented.Tempering in all of the sherds includes crushed calcite, used ratherabundantly. There is a limited amount of limestone in two of thesherds; sand, grit, and other aplastics do not appear to be present.Calcite crystals are found in various shales which are exposed alongthe Republican River and can be freely extracted on the weathered ^A burial custom which may be comparable in some respects to that represented by theremains at 14PH4 is described for the Huron Tribe by Jean de Breboeuf in the year 1636(Kenton, 1927, pp. 297-308). ?^5n1-g 1.letsIt3rid N30EI Surfoc* of groundDatum lino T^ sterile yellow loessSi^iKi Loess with slight mixture of chorcool NORTH -SOUTH PROFILE ACROSSWOODRUFF OSSUARY ON El LINEWM pip No%1"'' WOODRUFF OSSUARY?KIVETT 119 slopes. There may be other sources in the immediate vicinity. Thecalcite inckisions range in size from fine (less than 0.25 mm. diam.)to very coarse (over 7 mm. diam.) , with most falling between 1 and 3mm. The paste is buff to dark gray in color. The interior surfaceof the sherds has a thick carbon coating which may be scratched offeasily. When viewed in cross section, approximately one-fourth ofthe sherds' thickness adjacent to the interior surface is black. Freshlybroken surfaces are very irregular and have a granular appearancewith a tendency to shatter. Hardness ranges from 3 (calcite) to 4.5(chabazite), with 6 of the sherds about 3.5 (celestite). Thicknessranges from 9 to 13 mm., with 7 sherds having thicknesses of 11 to 13mm. The exterior surfaces of all the sherds are cord-roughened.The cord marks are medium fine to coarse and lie parallel ; impres-sions of individual twisted cords are visible in several instances.Although the soot-encrusted interior surfaces are difficult to observe,they appear to have been smoothed. There is no evidence of decora-tions, handles, or other appendages. The few body sherds, all ofwhich are small, give no indication as to vessel shapes.WORK IN ANTLER AND BONEArtifacts of bone and antler, although not common at the site,were rather well preserved. There are 8 specimens of antler whichappear to represent two types of tools. Five of the implements, oneincomplete, are made from the proximal end of the antler with aremnant of the "burr" evident as a slight swelling at one end (pi. 22,?, 5). The length of the antler sections varies from 47 to 75 mm.The original rough surface has been retained on the sides, but theends are rounded and moderately smoothed. The distal end of onespecimen has a rolighly circular cavity with a diameter of 11 mm.and a depth of 19 mm. The specimen is in a fragile condition andit cannot be determined whether the cavity is the result of decay or isdue to modification by human agency. These implements are some-what suggestive of the antler tapping tools or "cylinders" of theUpper Republican and other Central Plains aspects (Cooper, 1936,pi. 20, 1-3). In general, however, the length of the specimens from14PH4 is considerably less and the specimens are less carefully fin-ished. Similar objects have been described as "drifts" for workingstone (Webb, 1939, p. 32, pi. 13, h).Three antler tines were also recovered from the excavation (pi.22, a, 4) . The proximal ends of all three are roughly finished. Thedistal ends of two are scarred and nicked in a manner which suggeststheir use in pressure flaking ; one is worn diagonally to a chisel pointwhile the other has a rather blunt tip. The distal end of the otherspecimen is lacking. The sides of all three sections are smoothed and 120 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154moderately polished. The specimens vary in length from 72 to 80mm., and the greatest diameter, at the base, varies from 18 to 24 mm.Tubular bone beads are represented by 28 specimens, 9 of whichcame from the floor of Feature 12, a pit (pi. 24, 3). The remainderwere found at various levels throughout the excavation. All butone are small and well made ; they were probably fashioned from smallbird bones. Their lengths range from 9 to 14 mm. and the diametersvary from 3 to 6 mm. Characteristically, they have a circular crosssection throughout. The nature of two of the beads which are notfinished on the ends indicates that the small bone was cut part waythrough and then snapped off. Subsequent polishing has eliminatedevidence of this process on most of the beads. A somewhat largertype of tubular bone bead is represented by a single calcined specimen(pi. 24, 4), which is slightly curved longitudinally. It is 30 mm.long and has a diameter of 11 mm. Its ends are smoothed androunded and its sides appear to have been well polished prior to burn-ing. None of the beads are incised or otherwise decorated.Bone awls are represented by a single incomplete specimen (pi. 22, ?, 3) . Its original length cannot be determined, but a considerable sec-tion of the base appears to be missing. The surface of the break isrough and unpolished, and shows no evidence of use subsequent to frac-turing. The present length is 16 cm. and the greatest width, near thebreak, is 18 mm. The specimen appears to have been fashioned fromthe split metapodial of a deer. The point has been rounded and pol-ished, while the shaft retains the original exterior contour of the bone.The cancellous tissue has been removed from the interior.From a depth of 14 inches in square N35 El came a deer metapo-dial from which the distal end has been removed by cutting and break-ing (pi. 22, a, 1) . A circular hole 9 mm. in diameter appears to havebeen made near the center of the proximal articular surface and thecancellous tissue has been removed so that a cavity extends the lengthof the specimen. The exterior surface appears to be otherwise un-modified. The specimen has a length of 19 cm. Wliether it had afunction in its present condition or was in the process of being fash-ioned into a tool is not known. Bones similarly treated, except forthe hole in the articular surface of the proximal end, are found in sitesattributable to the Upper Republican aspect and appear to representa stage in the making of awls (Wedel, 1935, p. 200) . Beneath the lower left arm of the articulated burial (Feature 21)in section N25 El was an implement made from the metapodial of adeer (pi. 22, a, 2). A second very similar implement was found at adepth of 39 inches in Feature 12, a pit. A calcined fragment of athird specimen of this type lay surrounded by burnt earth and char-coal in section N30 El, at a depth of 24 inches. The distal ends of Pap No^sT"^' WOODRUFF OSSUARY KIVETT 121the two specimens which appear to be complete have been removed,apparently by breaking, as the end, although well polished, is uneven.The proximal end of the specimen excavated from Feature 12 hasbeen partially removed and the end has then been rounded. At adistance of 19 mm. from the modified proximal end of this specimenis an oval hole which averages 14 mm. by 10 mm. in diameter. Op-posite edges of the greater diameter are slightly beveled. One com-plete specimen, associated with Feature 21, has a length of 13.5 cm.while the other has a length of 11.6 cm. There are two shallow in-cised intersecting lines near the end opposite the perforation. Thecalcined fragment, which consists of the proximal end extending ap-proximately halfway through the perforation, appears to have beensimilar to the other two. Three short parallel diagonal lines havebeen incised on one side in the vicinity of the perforation. The twocomj)lete specimens are highly polished on the sides and the cancel-lous tissue appears to have been removed from the interior of thebone. These objects are similar to the so-called shaft wrenches whichoccur in nearly every Central Plains complex. They are more com-monly made from rib bones of large mammals, but specimens madefrom leg bones of cervids do occur in the Upper Republican aspect.Wliether these specimens served a function similar to those made ofribs is not known.Various burnt animal bone fragments which appear to have been cutor polished were found. At least one such fragment, which may befrom the scapula of a rather large animal, is well smoothed and pol-ished. Its original use cannot be determined. Large numbers ofrabbit teeth occurred scattered near Feature 16. They were unmodi-fied and their purpose is unknown.WORK IN STONEArtifacts of chipped stone were not abundant at the site, but theyrepresent considerable variety. Jasper appears to have been utilizedto the near exclusion of other materials for making various chippedstone artifacts. This material outcrops locally in the Republicandrainage and occurs in various shades of yellow, brown, red, and gray.A few chalcedonic flakes have surfaces suggestive of water polishingand may have been picked up as stream pebbles and utilized. Darkred quartzite was used for making one specimen.A total of five projectile points sufficiently complete to permit classi-fication came from the excavation (pi. 23, a, 3). The original formof four broken points cannot be determined. Two of the largerpoints exhibit relatively coarse flake scars on the flat surfaces but theedges are finely chipped. The three smaller specimens show ratherfine over-all chipping. In quality of workmanship, the smaller points 122 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154 equal the finely chipped triangular, side-notched points of the UpperRepublican aspect. The edges of one broken point are rather finelyserrated. In cross section the blades of all points are double-convexwith the greatest thickness (3-T mm.) along the midline.On the basis of form,^ the five projectile points appear to representthree types, all variants of a stemmed form. All have expandingstems and are shouldered and barbed. Three have convex bases(SCbl), one has a straight base (SCb2), and one has a concave base(SCb3). The three with convex bases vary in length from 21 to 26mm. and in width from 13 to 15 mm. One of the other two speci-mens (SCb2) has a length of 48 mm. and a width of 38 mm., whilethe length of the other (SCb3) is 55 mm. and the width, 31 mm.Six crudely chipped specimens which appear to fall into one classcame from various levels throughout the excavation and may be classi-fied as celts (pi. 23, &, 2). All are made of yellow to dark brownjasper and are roughly triangular in shape. The bit at the widestend is rounding and has been thinned by secondary chipping. Thepoll is less well shaped and may be either rounding or slightly pointed.A lime matrix, which probably represents the surface of the originaljasper slab, is present on one lateral surface, and sometimes on both.The specimens vary in length from 5.9 to 8.2 cm. ; the greatest width,which is just above the bit, varies from 3.8 to 5.2 cm. Maximum thick-ness is in the vicinity of the greatest width and varies from 1.5 to2 cm. These specimens are very similar to the chipped celts fromthe Upper Eepublican aspect, but as a series average much smaller.There are 35 rough flakes with one or more retouched edges (pi. 23,a, 2). They exhibit no uniformity in shape and appear to be hastilyimprovised implements. The retouched edges on many are very fine,while on others they are coarse. They may have had a scraping orcutting function.Six end scrapers, one of which came from the surface, are of theplano-convex type (pi. 23, a, 1). They vary to some extent in shapebut common to all is a plane or slightly curved ventral surface. Thesmallest is 27 mm. by 24 mm., while the largest is 54 nmi. by 31 mm.They do not appear to differ significantly from specimens occurringin nearly every Central Plains archeological complex.A few broken, roughly chipped sections, worked on both faces anddouble-convex in cross section, may be knife fragments. A roundingend is suggested by two fragments. Two complete specimens havelengths of 45 and 38 mm. and widths of 25 and 28 mm. The brokensections appear to be from specimens of somewhat greater size.Similar specimens were found at 25VY1 (Hill and Kivett, 1941,p. 172). 'After Strong, 1935. Pap.No^'sT"^' WOODRUFF OSSUARY?KIVETT 123A brown jasper slab measuring 13 cm. by 6 cm. by 2.5 cm. is roughlyflaked and may represent raw material in the form in which it wasbrought from the quarry or stream bed. A lime matrix visible onthree sides indicates the thickness of the original seam.Three stream-polished quartz pebbles show evidence of use as peck-ing or hammer stones (pi. 23, 6, 1) . They do not appear to have beenintentionally shaped but have been slightly modified by use. One,from the surface, is roughly rectangular in shape and has its twoopposite ends rounded through use. It has a length of 6.5 cm. anda width of 3.4 cm. The other two specimens, which were excavated,are somewhat round in outline. One shows evidence of batteringon only one end but the other has been used on all edges. The largerhas a diameter of 8.2 cm. and a thickness of 4.5 cm., while the otherhas a diameter of 6 cm. and a thickness of 4 cm.WORK IN SHELLArtifacts of shell were very abundant on the surface as well asthroughout the excavation. The majority of the thousands of speci-mens are fashioned from fresh-water mussels, but some were madefrom marine forms.Disk beads and blanks of shell representing all stages of manu-facture were extremely abundant throughout the excavation. Groupsof as many as 30 evenly matched, ground and perforated beads werefound in alinement just as the strings of beads were presmnably throwninto the pit. The single complete articulated skeleton appears tohave been wrapped in the pelvic area and up the chest with stringsof beads. All the beads and blanks have been bored but many arenot ground down. The unfinished beads appear to have been strungseparately from the ground and finished specimens. A few unboredand unground blanks were scattered above the lines of beads (pi. 22,5, 1) . Many of the individual strings contained more than a hundredbeads. The process of making the beads appears to have been asfollows : The mussel shells were broken into sections which were thenground or broken into a roughly circular shape. The blanks werethen drilled from the interior surface and the outer surface wassmoothed and flattened to enlarge the drill hole (pi. 22, 5, 2.). Theywere probably then strung before they were ground and polished totheir final matched form (pi. 22, 6, 3). Strong (1935, p. 120) hassuggested this method for almost identical specimens from the Mar-shall site. He describes the process as follows : Apparently, the shells of a fresh-water bivalve were broken up,ground approximately round on sandstone or broken to shape, bored, strungon some stiff wood flber, and then rolled on a slab or between two slabs untilperfectly smooth and round. This is the method employed by the Porno and 124 BUREAU OF AMERICAN ETHNOLOGY [Bull. 134 other California tribes, and the same sequence is suggested by the completeseries from the present [Marsliall] site.Although some of the thicker finished beads from 14PH4 have beendrilled from both sides, the majority, like those from the Marshallsite, were drilled only from the interior side. There is considerablevariation in the size of the finished beads (pi. 22, &, 3). They rangein diameter from 6 mm. to 18 mm., with the majority being not morethan 9 mm. across. The holes are usually well centered but in somecases they are very near the edge, particularly on those specimenswhich are ground down to a diameter of 7 mm. or less.There are 9 nearly complete crescent-shaped pendants with a per-foration at each end for suspension and fragments which appear torepresent 14 additional specimens of the same type (pi. 25 ; 11) . Theyare cut from the ventral margin of fresh-water bivalves along thepallial line. Their length appears to have been determined by thesize of the shell. The outer curved surface follows the ventral marginof the shell, wdiile on the inner edge the shell has been cut along thepallial line. Three complete specimens which were found associatedare of graduating sizes and suggest that several may have been worntogether. They vary in length from 6.2 cm. to 12.5 cm. The maximumw'idth, taken near the midpoint, varies from 1.2 cm. to 1.5 cm. Thespecimens taper to a rounded point at each end. Several were re-drilled at the ends after the original perforations for attachment worethrough or were broken. The edges of all specimens have beensmoothed and polished to a rounded surface, but the sides show littlemodification. Perforations vary in diameter from 2 to 4 mm.; theyappear usually to have been drilled from the interior surface of theshell.A second type of pendant is represented by a single specimen whichcame from square N40 El at a depth of 42 inches. This object wasfashioned from a fresh-water mussel shell and is subrectangular inshape wdth rounding corners (pi. 25; 1). A hole drilled from theinterior is located near each of two corners. The shell has been cutlongitudinally so that the upper margin of the gorget was originallynear the center of the shell and the lower margin retains the pallialline and ventral margin. The edges are well rounded and smoothed.The specimen has a length of 9.5 cm. and a width of 4 cm. A similarobject has been reported by Strong (1935, p. 115) from an unassignedburial near the mouth of Prairie Dog Creek, in Plarlan County, Nebr.Two similar pendants were also recovered, during River Basin Sur-vey's excavations in 1948, from a burial assignable to the Keith focus(Woodland) in the Medicine Creek Reservoir area in Frontier County,Nebr. (Kivett, 1949, fig. 69, 1, c). RiT.Bas.Sur. WOODRUFF OSSUARY KIVETT 125Pap. r^\r-r^ ?C^"-.^-^ fi*,~^ - . '!i 'si ^ikQ BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 23 C7 .^.^.:f!Sn ^ i|IIIW|lllljllll|ll|III|llll|llll[lllljllli|i'lijl!;i|lllijl,ll|m a, Chipped-stone artifacts from site 14PH4. Scale is metric, h. Stone artifacts from site14PH4. /, Hammerstones; 2, chipped-stone celts. Scale is metric. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 24 s ei ? I 9 i I9 ? ? ? % f i i I I I II ? ???? Artifacts from site 14PH4. 1 2 5, made of shells; 3, 4, of bone. Scale is metric. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 25 Marine and fresh-water shell ornaments, site 14PH4. Scale is metric. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 26 a. Artifacts from various sites. /, Triangular shell pendants from Robb Ossuary-, site25WT4; 2, perforated canine teeth from Guide Rock Ossuary, site 25VV1J; J,Guide Rock Ossuary, b. Sherds from Guide Rock Ossuary, site 2SWT3. Left, Bodysherd. Right, Rim sherd. Scale is metric. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 27 ^\>?,v??ipi^^^ Restored vessel of the Harlan Cord Roughened type from site 25FR8. Scale is metric. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 28 ite.**^ Four views of male skull (14PH4-1436) from Woodruff Ossuary, SMITHSONIAN INSTITUTIONBureau of American EthnologyBulletin 154 River Basin Surveys Papers, No. 4The Addicks Dam Site I. An Archeological Survey of the AddicksDam Basin, Southeast TexasBy JOE BEN WHEAT 143 953842?53 11 CONTENTS PAGEAcknowledgments 149Introduction 151Environment 152Topography 152Geology 154Drainage system 154Climate 155Biota 156Ethnohistory 157Early exploration of the Galveston Bay area 158Founding of the missions 160The historic Indians around Galveston Bay 160The culture of the Galveston Bay Indians 161Descriptions of Addicks Basin sites 163The Addicks "Mound" (42/66A6-1) 164Site 42/66A6-6 166Site 42/66A6-7 166Site 42/66A6-8 167The Doering site (42/66A6-2) 167Site 42/66A6-5 173Site 42/66A6-4 174The Kobs site (42/66A6-3) 176The Grisbee site (42/66A3-1) 182Material culture 184Pottery complex 184Goose Creek Plain 184Goose Creek Incised 189Tchefuncte Stamped 190Unidentified wares 190Miscellaneous objects of clay 190Ceramic stratigraphy 190Discussion 192Lithic complex 196Projectile points 201Upper level types 201Miscellaneous upper level types 203Middle and lower level projectile points 204Unplaced projectile point types 211Projectile point stratigraphy 214Discussion 218Miscellaneous lithic artifacts 220Knives 220Scrapers 224Choppers 225Drills 225Miscellaneous chipped stone 227145 146 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154Material culture?ContinuedLithic complex?ContinuedMiscellaneous lithic artifacts?Continued pageGround stone 227Miscellaneous minerals 228Stratigraphy and discussion of miscellaneous lithic artifacts. _ 228Bone, antler, and shell complex 230Bone 231Antler 232Shell 233Stratigraphy and discussion of bone, antler, and shell artifacts 233Subsistence 234Animal bone refuse 236Summary and conclusions 238Appendix : Trait list for Addicks Basin sites (table 9) 247Literature cited 248ILLUSTRATIONSPLATES29. Doering site and profile 25230. Kobs site and view of excavation showing Burial 6 2523L Pottery, restored vessels 25232. Pottery, decorated sherds 25233. Pottery, decorated sherds 25234. Projectile points: Perdiz Pointed Stem, Scalhorn Stemmed, EddyStemmed 25235. Projectile points: Alba Barbed, Kobs Triangular, and ProvisionalTypes Nos. 1, 2, 3, and 4 25236. Projectile points : Gary Stemmed points 25237. Projectile points: Provisional Types Nos. 5, 6, 7, 8, 9, 10, 11, 12, 13,14, and Wells Contracting Stem 25238. Projectile points: Provisional Types Nos. 15, 16, 17, 18, 19, 20, 21, 22,23, 24; Pedernales Indented Base, and Clovis Fluted (?) 25239. Projectile points, unplaced chronologically: Provisional Types Nos. 25,26, 27, 28, 29, 30, 31, 32; Copena point, Plainview point, and BairdBeveled point 25240. Knives: Types Nos. la, 16, Ic, Id, 2, and 3 25241. Knives: Types Nos. 4, 5, 6a, and 66 25242. Scrapers 25243. Choppers 25244. Drills, gravers, and eccentric flint 25245. Ground stone: Atlatl weight, sandstone abraders and saws, millingstone, and handstone 25246. Miscellaneous minerals and paint pigments 25247. Artifacts of bone, antler, and shell 252FIGURES4. Map of general area 1535. Map of site distribution with inset showing relation of Addicks Dam,Barker Dam, and Houston, Tex 1636. Sketch map, Addicks "Mound" 1647. Sketch map, Doering site (42/66A6-2) 168 pip.?k.'/r- CONTENTS 147PAGE8. Contour map and plan of excavation, Doering site (42/66A6-2) 1699. Profiles, Doering site (42/66A6-2) 17010. Burials, Doering site (42/66A6-2) 17211. Sketch map, site (42/66A6-4) 17512. Stratigraphic section, site (42/66A6-4) 17513. Sketch map, Kobs site (42/66A6-3) 17614. Contour map and plan of excavation, Kobs site (42/66A6-3) 17715. Profiles, Kobs site (42/66A6-3) 17916. Burials, Kobs site (42/66A6-3) 18017. Sketch map, Grisbee site (42/66A3-1) 18218. Stratigraphic section, Grisbee site (42/66A3-1) 18319. Pottery : Rim and lip profiles 18620. Pottery: Vessel shapes 18721. Pottery: Base shapes 18822. Graph showing comparative projectile point stratigraphy of Doeringand Kobs sites 21523. Relative chronology of the Addicks Basin sites 243 ACKNOWLEDGMENTSIt is an often repeated statement that any work of research is notthe product of the student alone, but of many individuals whosecontributions vary in character and extent. In verification of thisadage, I append the names of those who have, through their gener-ous help and encouragement, aided in the writing of this report.I wish to acknowledge gratefully the help of the personnel ofthe Department of Anthropology of the University of Texas. Dr.J. G. McAllister, then chairman of the department, made availablelaboratory and storage space for the use of the Smithsonian Institu-tion River Basin Surveys. Encouragement and direct help in arche-ological problems have been given freely by Alex D. Krieger andDrs. J. Charles Kelley and Thomas N. Campbell.Col. D, W. Griffith, then commanding officer of the Corps of Engi-neers, Galveston District, made possible the excavation of the Doeringand Kobs sites by making available a crew of laborers. John M. Simsgreatly facilitated the work by his cooperation in administration ofthe crew and by securing certain field equipment. All employees ofAddicks and Galveston offices of the Corps of Engineers were help-ful and interested in the progress of the field work. I wish also tothank the various crew members who fulfilled their jobs with inter-est and competence.Marshall Black and Charles Galenkamp aided in the field as timepermitted, and Mr. Black's carefully kept field notes on the Addicks"Mound" were of considerable use.The cooperation and help of Dr. Frank H. H. Roberts, Jr., in hisover-all direction of the project, lightened the burden of field workand made the subsequent task of writing the report easier.For the identification of various materials I am indebted to theseworkers: Dr. Joseph P. E. Morrison, of the Division of Mollusks,United States National Museum, for shell; Dr. Theodore E. White,staff paleontologist. River Basin Surveys, Smithsonian Institution, forbone; Mr. John W. Anthony, Arizona Bureau of Mines, and Mr.William V. Ward, Department of Geology, University of Arizona,for minerals ; Mr. George I. Quimby, Chicago Natural History Mu-seum, and Dr. James B. Griffin, Museum of Anthropology, Univer-sity of Michigan, for pottery.I am grateful to Dr. Emil W. Haury, Head of the Department ofAnthropology, University of Arizona, whose encouragement, criticism,149 150 ACKNOWLEDGMENTS advice, and aid have contributed greatly to whatever clarity andvalue this report may have. Mr. E. B. Sayles, Curator of the Ari-zona State Museum, aided in the photographing of lithic materials.Other members of the faculty, staff, and students of the Departmentof Anthropology, University of Arizona, have likewise shared theirtime and ideas with me concerning this work.Finally, I wish to acknowledge with affection and gratitude theassistance of my wife, Pat, who helped me in the field and who, byher constant aid and encouragement, has in a very real sense madethis work as much her own as mine.This paper in its present form was submitted to the GraduateCollege of the University of Arizona in partial fulfillment of therequirements for the degree of Master of Arts. I wish to acknowl-edge its release for publication by the Graduate Study Committee.To all these my thanks are due?to them must belong much ofthe credit for this work?the errors of omission and commissionare mine. AN ARCHEOLOGICAL SURVEY OF THE ADDICKSDAM BASIN, SOUTHEAST TEXASBy Joe Ben WheatINTRODUCTIONThe archeological survey of the Addicks Dam Basin was initiatedin March 1947 by the writer as a part of the River Basin Surveys. Theinitial phase of the operation involved the locating, testing, and evalu-ation of all sites likely to be damaged or lost through the constructionof the reservoir and associated channel rectification and subsequentinundation, silting, etc. This phase culminated in the partial excava-tion of two sites, begun late in May and concluded on July 15, 1947.This paper constitutes a report on the survey and excavation of theAddicks Dam Basin sites. The area is poorly known historically andalmost unknown archeologically.^ Sporadic collecting has been doneby amateurs in plowed fields, stream channel exposures, as well as someminor digging into two or three known middens, but the existence ofnumerous small middens located along the drainage ways was unsus-pected. Since this research marks the first major archeological re-search done in the area, the results have been presented in some detail.The construction of a local chronology and correlation with other areashas been carried as far as the data warrants. The lack of strictlycomparable published material has made such correlation difficult.The Addicks Dam, on South Mayde Creek, is one of two flood pre-vention dams designed to protect the city of Houston, Tex., from flood-ing by the waters of Buffalo Bayou, a tributary of the San JacintoRiver. This project, under the direction of the United States Corpsof Engineers, consists of the already completed Barker Dam, on themain channel of Buffalo Bayou, and Addicks Dam, under constructionat the time of the survey, which will impound the flood waters of SouthMayde, Langham, Horsepen, Bear, and Turkey Creeks, permitting the * Pearce, 1932 a and 1932 h, briefly mentions the area under consideration, basing bisremarks on the collections of several local amateurs and upon a few short visits to some ofthe sites around Galveston Bay. Sayles, 1935, combined survey data and ethnohistoricalmaterial to propose an Attacapan Phase chiefly historic in time. These data will befurther considered below. 151 152 BUREAU OF AMERICAN ETKNOLOGY [Bull. 154main channels to maintain a normal flow. A third project, WliiteOak Dam, to be constructed on \Vliite Oak Creek, is under considera-tion at the present time.Addicks Dam lies about 16 miles west of the city of Houston, to thenorth of highway U S 90 and the Missouri, Kansas, & Texas Railway.It is of earthen construction, the earth having been taken from borrowpits on either side of the embankment. The construction of this damwas initiated in 1943 and continued intermittently during the waryears. At the time of the survey it was nearing completion.The numbering of sites follows the system in use by the Texas Coun-cil of Archeology. This numbering system is as follows : The Stateof Texas is divided into a number of quadrangles, each circumscribing1 degree of latitude and longitude and numbered consecutively, begin-ning in the northwest corner and running from west to east. Theseare divided into four equal subquadrangles, lettered A, B, C, and D,also beginning in the northwest quadrant. The lettered subquad-rangles are further divided into 9 units, each comprising 10 minutesof longitude and latitude. Sites are numbered serially, as reported,within the smallest unit; e. g., 66A6-1: 66 (major quadrangle) A(subquadrangle) 6 (subunit) ?1 (serialnumber of site). The prefix,42, is placed before the site number to indicate the State of Texas inthe alphabetical order followed by the Smithsonian Institution.Projectile point names are, for the most part, those in use by theUniversity of Texas, and the present specimens have, in each case, beenchecked by Krieger, Kelley, or Campbell, of the university staff.Other named types will be discussed below.The cultural manifestations of the Addicks Dam Basin have beenarbitrarily divided into three periods or levels of occupation, thesebeing termed lower, middle, and upper levels respectively. The basisof this division will be discussed below.ENVIRONMENTThe Coastal Plain Physiographic Province is a well-defined low-land that borders the coasts of the Gulf of Mexico and the AtlanticOcean from Yucatan to southern New England (Atwood, 1940, p. 25),That part of it which extends from the Mississippi Alluvial Plainwestward and southward along the Texas coast into Mexico is termedthe West Gulf Coastal Plain (Fenneman, 1938, p. 100). It is thissubarea with which we are presently concerned, for it is in this regionthat the Addicks Dam Basin is situated.TOPOGRAPHYThe topographic features closely reflect the geological substructure.The coastal plain proper is nowhere very wide, not exceeding 100 milesin most areas. Sand reefs parallel the entire Texas coast, broken Riv. Bas. Sur.Pap. No. 4] ADDICKS BASIN WHEAT 153occasionally by shallow inlets to the lagoons that lie between themand the mainland. One of these expands into Galveston Bay (At-wood, 1940, p. 56) (fig. 4) . From this young coast the land grades Figure 4.?Map of general area.gently inland to a mature coastal plain which is a nearly featurelessbroad lowland. Shallow, wind-eroded pits, locally termed "blow-outs" or "hog-waUows," and small clustered knolls of disputed originoccur in the area, but neither is of sufficient magnitude to relievegreatly the monotony of this level land (Plummer, 1932, pp. 792-793) . 154 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154Inland from the low, swampy coast lies a belt of prairie covered withtall, deep-rooted grasses. A few trees are now found in this area,but true forest cover begins to the north and east, continuing through-out the eastward extension of the Coastal Plain Province.The Addicks Dam Basin is situated about midway on this plain.Its mean altitude is about 30 m. (98.42 feet). A fringe of forestencompasses it, but a few miles farther south this finally gives wayto the open prairie by which the Indians had easy access to the coast.GEOLOGYGeologically the area is a homoclinal structural plain, composedof overlapping formations which dip gently toward the coast. Attheir inner border they outcrop as a series of belts or low cuestas,roughly paralleling the coast. The older formations are marine inorigin, having been laid down during periods of submergence, andare separated by erosional unconformities resulting from alternateperiods of emergence and submergence. Since late Tertiary timesthe area has been above water for the most part (Chamberlin andSalisbury, 1930, p. 734), and the subsequent formations are largelyof fluviatile deposition (Atwood, 1940, p. 57).The present land surface of the coastal plain is Quaternary inorigin. The Lissie formation, of Pleistocene age, unconformablyoverlies sands of Pliocene age and outcrops in a belt some 30 mileswide, bounded on the north by the Hockley Escarpment. Its south-ern boundary is formed by the unconformably overlying Beaumontclays of late Pleistocene age which extend nearly to the present coast.Unconformably overlying both the Lissie and the Beaumont forma-tions over most of the plain is a topsoil composed principally of riverdeposits, while the present coastline is formed by wind- and marine-laid sands. Along the stream courses there are terrace deposits, alsoof recent age (Plummer, 1932, pp. 781-795).DRAINAGE SYSTEMThe region is characterized by a young drainage system (Suther-land, 1908, p. 243). The older, pre-Pleistocene rivers, such as theSabine, Neches, Trinity, Brazos, and Colorado, head upon the olderformations of the interior and pass completely across the coastalplain, through which they have cut deep grooves (Fenneman, 1906,p. 12). Gravel bars of these streams furnished the Indians with aready source of poor-grade flints. However, the water courses ofthe San Jacinto River, Buffalo Bayou, Clear, Oyster, San Bernard,and Caney Creeks, are confined to the plain itself or head at strongsprings along its northern border, and hence are younger or post-Pleistocene (Fenneman, 1906, p. 12). Nevertheless, much of the pip.Na'//"" ADDICKS BASIN?WHEAT 155area is poorly drained, and small ponds and lakes hold runoff waterduring the rainy season.The stream courses are fringed by trees and are usually borderedby discontinuous natural levees in the form of sand and silt knollswhich permit overflowing into the bottomlands adjacent to the chan-nels. These sand and silt or clay knolls are quite noticeable alongthe water courses and occasionally occur on the plain where, asresidual levees, they exist as a remnant of the plain's aggradation.It was on these natural knolls that the Indians established their camps.Buffalo Bayou rises along the Waller-Harris County line, flowssouth and east through the northeastern part of Fort Bend County,and then eastward through a relatively deep and narrow channel tothe city of Houston. From this point to its confluence with the SanJacinto River, the stream widens and today forms the upper part ofthe Houston Ship Channel which opens into Galveston Bay. It hasa number of tributaries, the largest of which are Brays Bayou, WhiteOak Bayou, Greens Bayou, Turkey Creek, and South Mayde Creek.The Addicks Dam Basin controls the flood waters of Turkey Creekand South Mayde Creek. Although Turkey Creek is an independentstream, and before the construction of the dam emptied directly intoBuffalo Bayou, it is intermittent, and the volume of water it carriesis small. The main tributary of South Mayde Creek is LanghamCreek, which is itself fed by Horsepen and Bear Creeks. Thesestreams flow through tree-bordered channels, the banks of which are,in places, thickly overgrown with thorny vines and underbrush. Thechannels are cut through the recent alluvial fill and into the Beaumontclay. Wliile the streams are normally shallow and carry a small vol-ume of water, they nevertheless afforded a permanent water supplyfor the Indians. During flood periods, however, these streams fre-quently overflow their banks and inundate wide areas of bottomland,so that the natural knolls and levees are at that time the only landabove water. For this reason, they were important to the aboriginalinhabitants. CLIMATEThe West Gulf Coastal Plain has a humid subtropical climate (Blair,1942, p. 191). Long and hot summers are characteristic of the area,but summer winds from the Gulf relieve temperatures that occasion-ally soar above 38? C. (100? F.). During this period the relativehumidity is about 70 percent, and the high temperatures during mostof that time make for somewhat oppressive summers. During thewinters the winds are variable, but frequently blow from the conti-nental interior, bringing chilling weather in their wake. For themost part, however, the winters are mild and pleasant. 156 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154The climatological data from Houston may be summarized as beingalmost precisely that of the Addicks Dam Basin, since it lies only 16miles to the east, at the same altitude above sea level, and about thesame distance inland from the Gulf. The mean annual temperatureis 20.61? C. (69.10? F.) , with a range of 10? C. (30? F.) . The coldestmonth, January, has a mean temperature of 11.83? C. (53.30? F.),and the warmest, July, 28.50? C. (83.30? F.). Eighty-one days havetemperatures of 32.20? C. (90.00? F.) or above, and only 9 days havetemperatures of 0? C. (32.00? F.) or lower. The growing season is294 days. The wettest month is May, with 11.88 cm. (4.68 inches),and the driest is February, with 7.53 cm. (2.97 inches). Ninety-ninedays have 0.025 cm. (0.01 inch) or more of moisture, and the averageannual precipitation is 1.15 m. (46.00 inches) (Blair, 1942, p. 195).BIOTAThe Austroriparian Biotic Province is characterized by the pineand hardwood forests of the southeastern United States. Numerousswamps and marshes occur along the coast and the inland waterways.Both mammalian and avian fauna of the area are characteristicallyrather dark in color and none of the mammals are known to hibernatein winter. This province includes that part of the West Gulf CoastalPlain which lies east of the 95th degree of longitude, swinging westto include the area immediately north and west of Galveston Bay(Dice, 1943, map facing p. 4, and pp. 18-21). On the west it isbounded by the Texan Province. The Addicks Dam Basin, then, liesjust within the Austroriparian Province but within easy reach of theTexan. The fauna utilized by the aboriginal inhabitants, as borneout by the arclieological remains, is transitional between these twozones, being primarily that of the eastern province, but having, inaddition, some forms typical of the western. Ethnohistorical, as wellas arclieological, data testify that the Indians took advantage of tliistransitional position in their utilization of game.Buffalo Bayou is the approximate boundary between the BroomSedge and Water Grass Prairie fringing the coast and the oak-pineforest that extends north to the Red Eiver. A few miles to the north-west a tongue of prairie grassland extends into a belt of oak-hickoryforest that parallels, and forms the western border of, the oak-pineforest. The trees that primarily compose the oak-pine forest are theshort-leaf pine {Pinus echinata Mill.), yellow oak {Quercus velutinaLam.), bitternut {Hickoria cordiformis (Wang) Britton), but otheroaks and pines do occur (Shantz and Zon, 1924, p. 14) . The prairie cover is largely dominated by broom sedges {Andropo-gon glomeratuSf A. saccharoides^ A. furcatus)^ and switch grass {Pani-cv/m virgatum), which merges farther south with water grass pip.Na/]"'"' ADDICKS BASIN?WHEAT 157{Paspalum sp.) and finally with marsh grasses of the coast (Shantzand Zon, 1924, p. 17) . A varied fauna previously occupied the region but has been muchdecimated since the intensive settlement by Caucasians. The Vir-ginia deer {Odocoileus virginianus texanus) and black bear [Ursusamericaniis) have virtually disappeared since the 1880's. A few ofthe smaller animals are still present, though not common. Some ofthese are the bobcat {Lynx rufus texanus)^ the raccoon {Procyonlotor) , the gray and the western fox squirrels {Sciurus carolinensis, S.niger rufmenter) ^ and a number of rodents. The copperhead {Agkis-trodon mokasen Beauvois) , the eastern rattlesnake {Grotalus horriduaLinne), and the coral snake {Micrwus fulvius fulvius (Linne)), arepresent, as are many snakes of nonpoisonous varieties. Fish, clams,and turtles are present in most of the streams in the area, and whilemany birds are found, the Indians apparently made only limited useof them.The environment with which we are concerned may be considered arelatively favored one. The regularity of topography, together withthe stream and bay systems, made for easy passage from one area toanother by land or water, while sheltered sites for encampment wereplentiful. The native products of forest and plain were easily avail-able and many forms of wild game could be found in this transitionalzone. The exploitation of these varied resources was further aidedby a mild climate and must have assured the aboriginal inhabitantsof at least a minimal food supply.ETHNOHISTORYThe relationship of the historic Indian occupants of an area to theproto- and prehistoric inhabitants may be of either of two classes.If the ethnohistorical evidence indicates a cultural similarity to thearcheological evidence, not explainable entirely in terms of environ-mental factors, then it may be reasonable to assume an ethnic, or atleast a cultural, continuum. If, however, the evidence points to ahiatus or a severe modification between the archeological and theethnological sequence, one must consider that ethnic or cultural move-ments have transpired and that it is not safe to infer a direct relation-ship between the two. For this reason, it seems wise to ascertain thecultural attributes of the historic peoples of the area to as full anextent as they may be useful in determining whether such a culturalcontinuum does exist.Furthermore, an ethnohistorical insight into the nature of therelationships between the Indians themselves, and between them andthe Europeans with whom they came in contact, may shed light on 158 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154problems reflected in the archeological data only by tangible, imper-ishable artifacts without the human context.Although some of the tribes living about Galveston Bay were amongthe first of the native American peoples to come in contact with theEuropeans, the data preserved to us from the period are singularlyvague and scanty. Not until the beginning of the Mission Era in themiddle eighteenth century do we have full reports of the area, andthese are principally concerned with the governmental, political, andreligious aspects of the Spanish colonization and, therefore, shedlittle light upon the ethnological problems which face us. There are,however, many documents in the archives of Spain, France, Mexico,Texas, and Louisiana which, when translated, edited, collated, andpublished, may well illumine some of the lacunae in our knowledge.EARLY EXPLORATION OF THE GALVESTON BAY AREAOn November 6, 1528, a boatload of survivors of the Narvaez expe-dition was cast ashore on the western end of an island which theynamed Isla Malhado, purportedly the present Galveston Island. AlvarNuiiez Cabeza de Vaca, one of the shipwrecked men, in his narrative ofthe expedition, gives us the first report of the Indians of the region ; for although Alonzo de Piiieda, in the service of Francisco de Garay,had sailed along the Gulf Coast in 1519, making a map and daimingthe land for the Spanish Crown, he left no records of the inlandgroups (Castaiieda, 1936, pp. 7-8).For a period of 6 years, De Yaca lived among these peoples, partof the time as a slave, later as a trader and healer, finally to escapeto Mexico where he arrived, with three companions, in the year 1536(Hodge, 1907, pp. 5-7). It is difficult to use the data afi'orded byDe Vaca's chronicle because of its vagueness from an ethnologicalpoint of view. Some of his statements are at variance with later in-formation, assuming that the identification of Isla Malhado as Gal-veston Island is correct, and that the Indians called Han andCapoques were respectively the Akokisa and the Cocos, as suggestedby Hodge (1907, note 4, p. 54). The fact that the narrative waswritten several years after most of the action had taken place maywell explain some of the seeming discrepancies.Following the sojourn of De Vaca was a period of nearly a centurybefore the Indians in the vicinity of the Bay were again visited.During that time De Soto explored a considerable part of the south-eastern United States, and his successor, Moscoso, in the year 1542,visited the Hasinai territory to the north (Swanton, 1946, pp. 39-59). La Salle established his colonj^ on Matagorda Bay and in 1686,and again in 1687, passed near, if not across, the headwaters ofBuffalo Bayou in his attempt to reach the French settlements in Can- pip.Na/]"^" ADDICKS BASIN?WHEAT 159ada (Castaneda, 1936, map following p. 444). Alonso de Leon, Jr.,likewise passed through the area in his fourth and fifth entradas inthe years 1689 and 1690 (Castaiieda, 1936, map following p. 444).None of these expeditions resulted in information concerning thenatives of the Galveston Bay area.Simars de Bellisle, a young French officer, landed with four com-panions on the Texas coast at Galveston Bay late in the year 1719.During the ensuing weeks, all but De Bellisle perished of hunger.Following the death of his companions, he joined a group of Indianswho enslaved him and subjected him to many inhumanities. Aftersome time he was given to a widow as a husband, but neverthelesscontinued to serve as a slave. Through the intervention of someHasinai Indians, he was rescued and returned to the French settle-ment of Nacatoches on February 10, 1721 (Folmer, 1940, pp. 204-225) . From the account of De Bellisle, one gains a general picture of themanner of life of his captors. Like De Vaca's narrative, De Bellisle'sreport suffers from the paucity of specific data necessary for the re-construction of a full cultural picture. Nonetheless, most of his storyis borne out by later research.In August 1721, De Bellisle returned to Galveston Bay with Ber-nard de la Harpe in the ship Suhtile^ commanded by Jean Berenger.With De Bellisle as his guide, De la Harpe landed and contacted theIndians, requesting permission to establish a settlement. The denialof this request by the Indians, because they feared reprisal for theill-treatment meted out to De Bellisle, led De la Harpe to carry awaynine of them to New Orleans (Folmer, 1940, pp. 204-225). Simarsde Bellisle called these Indians among whom he had lived "Caux,"but their almost positive identification as Akokisa rests on the vo-cabulary taken down by Berenger from the nine Indians taken onboard the Subtile (Swanton, 1946, p. 85).In 1727, the engineer, Alvarez Barreyro, crossed the area in execu-tion of orders by General Rivera, who was then making an inspectionof the northern frontier of New Spain, but the trip had no permanentresults in the form of ethnographic data (Bolton, 1914, vol. 1, p. 52).Rumors of French incursion in the neighborhood of Galveston Bayat last stirred the Spanish to concerted action, and in 1745 Capt. DonJoaquin de Orobio Bazterra was dispatched to investigate. In theearly months of 1746 he arrived among the Bidais and a short timelater among the Akokisa. Here the rumors were confirmed to theextent that French traders annually came into the territory to buyskins and to supply the Indians with needed commodities. OrobioBazterra found no permanent French settlement, but the promise ofone perhaps prompted a second visit to the area later the same year(Bolton, 1915, pp. 328-332).953842?53 12 160 BUREAU OF AMERICAN ETHNOLOGY [BnlL 184FOUNDING OF THE MISSIONSThe founding of three missions on the San Gabriel River?SanFrancisco, Nuestra Senora de la Candelario, and San Ildefonso?tookplace during 1749. Although these missions are not in the immedi-ate region of Galveston Bay, the last named concerns us briefly, sinceit was at San Ildefonso that the Akokisa, Bidais, and Deadoses, allof the Attakapa linguistic stock, were first congregated. After sev-eral years of trouble and dissension between the religious and themilitary, the missions were finally abandoned in the summer of 1765(Bolton, 1915, pp. 50-55).During the previous year a French trader, Joseph Blancpain, andfour companions were arrested by the Spanish in the village of ElOrcoquisac on the lower Trinity River. Governor Barrios sent Do-mingo del Rio among the Akokisa and the Bidai with gifts and in-structions to investigate the reaction of the Indians to the arrestof the Frenchmen. The report of Del Rio caused Barrios to dispatcha garrison to El Orcoquisac, which took up its post late in 1755(Bolton, 1915, pp. 337-342).The following year a presidio named San Agustin de Ahumadawas established in the Indian town of El Orcoquisac, a few leaguesabove the mouth of the Trinity. A mission was established nearbyand given the name Nuestra Senora de la Luz. (It is interestingthat 50 Tlascalan Indian families from Mexico were then settledthere.) Fifteen years of hardship, turmoil, strife between religiousand military, and constant threat of relocation ended in 1771 withthe final abandonment of both presidio and mission (Bolton 1915,pp. 342-374).Following the withdrawal of the Spanish, French traders again cameinto the area, and the English began shortly thereafter to make over-tures to the Indians. By 1805 the main town of the Akokisa wasreported to be on the west side of the Colorado River, and they thendisappear from the records (Swanton, 1946, p. 86).THE HISTORIC INDIANS AROUND GALVESTON BAYThe identity of the Indian tribes inhabiting the lower Trinity Riverand the region to the north and northwest of the Bay may be inferredwith considerable surety. The Han, who occupied the eastern endof De Vaca's Isla Malhado, are believed to be the Akokisa, the name"Han" being a synonym for the Attakapa and Akokisa word for house(Swanton, 1946, p. 85). The Capoques, occupying the western endof the island, were probably the Cocos of Karankawa affiliation(Hodge, 1907, p. 54). That De Bellisle's captors were Akokisa seemscertain, since our sole vocabulary of this language was recorded by Pap.Na/]"'^' ADMCKS BASIN?WHEAT 161Berenger from the nine Indians of this group who were carriedback to New Orleans by De la Harpe (Swanton, 1946, p. 85). OrobioBazterra is explicit as to the identity of the Indians around themouth of the Trinity and to the north and west. From OrobioBazterra's data, Bolton identifies the location of the four or fiverancherias of these people as along Spring Creek, which parallelsBuffalo Bayou a few miles to the north, and, like it, empties into theSan Jacinto (Bolton, 1915, p. 333). The Bidai, who occupied ter-ritory farther inland to the northwest, told Orobio that the Akokisaranged from the Neches River to a point halfway between the Trinityand the Brazos (Bolton, 1915, note 14, p. 334). Most of the contem-porary maps are also explicit in the location of the Akokisa andother coastal and inland tribes (Pichardo, 1931, cf. maps followingpp. 350, 388, 474; Bolton, 1915, pp. 350, 382). The Attakapa, ofwhom the Akokisa are western relatives, extend eastward into south-western Louisiana. Swanton places the Patiri, Bidai, and Deadose,all related to the Akokisa, along the upper course and tributariesof the San Jacinto and Trinity Rivers (Swanton, 1946, map follow-ing p. 1). The Mayeye lived west of the Bidai, and south of theBidai, toward the coast, lived the Cocos, of Karankawan stock. By1779, according to Morfi, these tribes had united and established them-selves on the coast between the Colorado and the Brazos (Morfi,1935, p. 81). Occupying the coast, westward from the mouth ofthe Brazos, were the Karankawa (Gatschet, 1891, pp. 45-46, andmap following p. 46). Some place them as far east as the west sideof Galveston Bay. From the evidence cited above, it appears thatthe sites of the Addicks Basin fall within the geographic range ofthe Akokisa.THE CULTURE OF THE GALVESTON BAY INDIANSSwanton states that : The Attakapa, although upon the coast of the Gulf of Mexico, were actuallyjust beyond the southeastern area proper. Along with the tribes westward ofthem, to and beyond the Rio Grande, they were characterized by a loose organi-zation, a low culture, and the existence of cannibalistic practices in somethingmore than merely ritual form. [Swanton, 1928, pp. 712-713.]The annual economic cycle of the Akokisa and the related tribes wassimple, the search for natural food determining their location. Thewinter season was passed in permanent or semipermanent camps.De Vaca says that because of the weather they "... retire to theirhuts and ranches, torpid and incapable of exertion" (Hodge, 1907,p. 57). De Bellisle also infers this, and describes the summer as aperiod of wandering in search of food "because they possess no cabinsor fields" (Folmer, 1940, pp. 216-217). Bolton (1915, p. 334) says 162 BUREAU OF AMERICAN ETHNOLOGY [BuU. 154that they went periodically back and forth between coast and interiorwith the changes of the season, although they lived in relatively fixed \villages. Orobio Bazterra reported in 1745 that they lived in "villages"(Bolton, 1915, p. 330). It can be seen that, in general, these tribes ^were transitional between the southeastern pattern and that of thecoastal groups to the southwest.Hunting, fishing, and gathering formed the main subsistence pat-tern. These tribes appear to have had little or no agriculture, at leastuntil the time of Orobio. Deer, bear, and occasionally bison werehunted, and it is interesting to note that by 1719 they had adopted ibison hunting from horseback (Folmer, 1940, p. 219). De Bellislementions the gathering of bird eggs in quantity, and the use of shellfisliis noted by him as well as by De Vaca. Edible roots, "wild potatoes," ;,are noted by both (Folmer, 1940, p. 215 et seq.; Hodge, 1907, p. 45 Iet seq.), and many other kinds of wild plants were undoubtedlyutilized.Cannibalism is affirmed by De Bellisle and many other later writers(Folmer, 1940, p. 219 ; Swanton, 1928, pp. 712-713) , so that it is some-what surprising that De Vaca states that the Indians were appalled |when the Spaniards, of necessity, ate each other (Hodge, 1907, pp. ?49-50). It does not seem likely, however, that camiibalism was Ieconomically important. ;The type of housing used by the Akokisa may only be inferred, iDe Bellisle states that they had no "cabins," but this statement may irefer only to their summer rounds. Moreover, he may have used the |term in the European sense of a permanent wooden or stone structure. |De Vaca mentions "houses of mats," but the reference is so general |that it does not greatly clarify the problem (Hodge, 1907, p. 52). |Orobio Bazterra speaks of the Bidai winter habitation as bearskintents (Bolton, 1915, p. 330), and, since the Bidai are relatives of theAkokisa, it seems safe to infer that some similar form prevailed there.Pottery is mentioned by both Cabeza de Vaca and De Bellisle, butneither gives the provenience or details of shape, size, or decorationof the vessels. Swanton (1946, p. 737) states that the Attakapa re-ceived most of their pots in exchange for other products from eitherthe Karankawa or the Avoyel, and probably from the Caddo also.Flint was obtained from the Avoyel and passed along to the Karan-kawa. Trade was also carried on with some of the interior peoples,as De Vaca points out (Hodge, 1907, pp. 56, 61).Social organization and nomnaterial culture have been describedin some detail elsewhere, by Bolton (1914, 1915), Swanton (1911,1928, 1946), Gatschet (1891), De Vaca (Hodge, 1907), De Bellisle 1(Folmer, 1940), and others. Since these subjects are not of primary |concern for the purpose of this paper, the reader is referred to theoriginal sources. Riv. Bas. SurPap. No. 4] ADDICKS BASIN?WHEAT 163DESCRIPTIONS OF ADDICKS BASIN SITESNine sites were recorded during the survey of the Addicks DamBasin and the associated outlet channel construction (fig. 5). Three Scale In MeiersiflOO1,000 5,000Scale In Fetl i.qoo10,000 Figure S.?Map of site distribution with inset showing relation of Addicks Dam, BarkerDam, and Houston, Tex. Wide-hatched area on inset shows zone covered by mainmap. of these sites are on the main channel of Buffalo Bayou, three on SouthMayde Creek, and the remaining three on Langham Creek. No siteswere located on Bear Creek or Turkey Creek within the area surveyed,and it seems likely, because of their relatively small size and the re- 164 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154 stricted amount of water carried by them, that they were not com-monly used as permanent camping areas. The survey was largelyconcentrated in those areas where it seemed probable that damagefrom construction, silting, or indiscriminate digging might occur. |The three sites located on Buffalo Bayou are the Addicks or Haber- fmacher "Mound" (42/66A6-1), upstream from the Buffalo Bayou- 'South Mayde Creek confluence, and sites 42/66A6-6 and 42/66A6-7, .downstream from this confluence in the order named. JTHE ADDICKS "MOUND" (42/66A6-1) |The Addicks "Mound" (42/66A6-1) was not an artificial mound but irather an extensive midden deposit located on the south side of the '.main channel of Buffalo Bayou (figs. 5, 6). Before the channel rec- FiGURE 6.?Sketch map, Addicks "Mound" (42/66A6-1).tification program for Barker Dam changed its course, the streamlocally took a northeasterly direction for about a mile to its confluencewith South Mayde Creek. The midden accumulation, 1.28 m. (4.20 pip.Na/]"'^' ADOICKS BASIN?WHEAT 165feet) thick, rested on a natural clay knoll. The combined height ofmidden and knoll above the surrounding land surface was approx-imately 1.82 m. (6.00 feet). On its north-south axis the site measured33.52 m. (110.00 feet) and on the east-west axis 30.48 m. (100.00 feet).A few meters to the southwest of the midden was a small linear de-pression, probably a segment of an old stream channel. About 20.00 m.(65.61 feet) to the east, the main channel of the Bayou was joinedby a northeasterly flowing shallow slough. Several large oaks weregrowing on the midden itself, and the bottom land of the valley floorsupported a number of hardwood trees, pines, and some undergrowthof bushy shrubs.For a number of years past, the midden had been known as an Indian ''mound" by local farmers and collectors from Houston and neighbor-ing towns. Sporadic digging for collecting purposes had been carriedon by several individuals. Stanley Morse, one of the collectors,brought the site to the attention of W. M. Black, an engineering stu-dent of Houston.In 1941 Mr. Black made a small excavation to a depth of about 30 cm.(0.98 foot) which resulted in the finding of a few projectile pointspotsherds, and other midden debris such as animal bone and musselshell.Early in 1942, when it became obvious that channel rectificationwork would result in damage to the site. Black returned, surveyed thesite, and prepared a contour map (fig. 6). He then made a secondexcavation. This pit was 60 cm. (1.96 feet) wide, 1.80 m. (5.90 feet)long, and penetrated the sterile clay underlying the midden. Thiswas reached at a depth of 1.28 m. (4.20 feet) . The pit was excavatedin arbitrary levels averaging 15 cm. (0.49 foot), and the materialrecovered was segregated by the level. Some of this material wasgiven to the Smithsonian Institution by Mr. Black and is includedin this report.No definite stratification was noted by Black. However, he describesthe midden fill as "loose, black loam." Numerous bone fragments,clamshells, small gastropods, and pieces of turtle carapace and plas-tron occurred in the fill. Pottery was reported as abundant in theupper part, occurring less frequently toward the bottom of the midden.( The segregated sample donated to the Smithsonian contained no pot-tery from the lowest levels.) Projectile points in the upper 15-30 cm.(0.49-0.98 foot) included small arrow points?Perdiz Point Stem andScalhorn Stemmed?and the larger Gary Stemmed points. Noneof the smaller points were found below this depth.In May 1942, a third excavation by Mr. Black resulted in the findingof a burial. I IQQ BUREAU OF AMERICAN ETHNOLOGY [Bull. 164Burial No. 1:Location: Test pit 3.Position: Semiflexed on right side; head oriented to northeast, face twistedbacli and down.Arms: Folded in front of chest with hands to chin.Legs: Loosely pulled up at about right angle to body with feet closeagainst pelvis.Age: Adult.Sex: Male (?).Associated objects: A large gray flint knife 63 mm. (0.206 foot) wide by202 mm. (0.66 foot) long, placed between the parallel forearms. Re-mains of what appeared to be seed or perhaps wooden beads werefound in the chest region. ,'Remarks: Bone condition was good. IDuring the completion of the Barker Dam outlet works, the middenwas razed and the material placed in a dump heap. It was latercovered by the embankment of the rectified channel. The site is nowvirtually destroyed, only about 30 cm. (0.98 foot) of the midden base lremaining in situ exposed in the channel face.SITE 42/66A6-6Site 42/66A6-6 lies about 2 miles downstream on the main channelof Buffalo Bayou and almost directly east from the Addicks "Mound"(fig. 5). This is a small midden, 22 m. (72.17 feet) long on a north-south axis by 12 m. (39.37 feet) on an east-west axis. It rests on alow, broad, natural sandy rise on the south bank of the bayou wherethe stream turns into a north meander through open woods of oak,willow, and a few pine.A small test pit, made to ascertain the depth of deposit and natureof the fill, resulted in the recovery of several plain ware potsherds.No projectile points or other stone artifacts were found, although flintchips occurred in the deposit. No stratigraphic test was made becauseof the objection of the tenant; moreover, the midden would not bedamaged in the channel rectification since the new channel was pro-jected to pass some distance to the north, isolating the site on a drymeander.Two hundred meters (656.16 feet) to the west, in a formerly culti-vated field, a surface camp site was reported but could not be verifiedupon investigation. SITE 42/66A6-7A short distance downstream from site 42/66A6-6 Buffalo Bayoumakes a small horseshoe bend toward the north. Sixteen meters(52.49 feet) east of where the bayou turns south again, on the northside of the stream, is a low, sandy knoll, 22 m. (72.17 feet) on itsnorth-south axis by 16 m. (52.49 feet) on its east-west axis. The site pip.Na'4T'^' ADDICKS BASIN WHEAT 167number 42/66A6-7 was assigned to this knoll (fig. 5). It had beenintermittently used by the Indians as a camping place, but occupationwas never of sufficient duration or continuity to result in the formationof a midden.From the grass roots to a depth of 15 cm. (0.49 foot) , sparse campdebris occurred in the sand. One potsherd and a few unaltered flintchips comprise the collection made at this site.Below its confluence with Langham Creek, South Mayde Creekflows in a general southeasterly direction for about 1,850 m. (6,069.53feet), then curves sharply to the southwest to flow some 750 m.(2,460.62 feet) before emptying into Buffalo Bayou. Three sites arelocated on South Mayde Creek. These are: Site 42/66A6-8, theDoering site (42/66A6-2) , and site 42/66A6-5, upstream in the ordernamed. SITE 42/66A&-8Site 42/66A6-8 is a small surface camp on the east bank of thecreek approximately 100 m. (328.08 feet) north of the Missouri,Kansas, & Texas Railway bridge (fig. 5) . Scant camp debris is erod-ing from grass-root level over an area of about 5 m. (16.40 feet).Further testing was not undertaken because of the superficial natureof the deposit. THE DOERING SITE (42/66A6-2)The Doering site is located some 200 m. (656.16 feet) northwest ofsite 42/66A6-8 (fig. 5). A short slough joins South Mayde Creekat this point, and the stream veers to the northeast for about 100 m.(328.08 feet) before resuming its normal southeasterly course. Thenorth bank is somewhat loAver than the south. It is composed princi-pally of a reworked sandy deposit which is brush covered and supportsa few willow trees (fig. 7). Beyond these, on higher ground, areoaks and other hardwood trees. The south bank forms a low escarp-ment. The site consists of a midden deposit 1.20 m. (3.93 feet) indepth overlying the clay which is the main constituent of the escarp-ment. The midden rises about 75 cm. (2.46 feet) above the surround-ing area, the bottom of the midden being 45 cm. (1.47 feet) below thelevel of the adjacent bottom land. It is 36.50 m. (119.75 feet) longon an axis parallel to the stream, and 16.50 m. (54.13 feet) across onthe north-south axis. Open forest surrounds the site, while severalpines and a few oaks, some of which attain a diameter of 30 to 40 cm.(0.98 to 1.31 feet), grow on the midden itself. There is less under-growth on the site than is usual along the stream banks, probably be-cause of the better drainage afforded by the slope of the midden.The highest part of the midden is westward of the center and im-mediately adjacent to the creek. From here it slopes gently to the east 168 BUREAU OF AMERICAN ETHNOLOGY [BuU. 154 Figure 7.?Sketch map, Doering site (42/66A6-2). Hatched area denotes area excavated.and south, and somewhat more steeply to the north and west. Alongthe creek bank the midden deposit is exposed in a vertical face wherethe southward cutting action of the stream during flood conditions hasresulted in the dissolution of a considerable portion of the site. Themidden soil thus precipitated into the flood waters was transported insuspension. Heavier components, however, such as potsherds, groundand flaked stone artifacts, and a quantity of stone debris, found lodge-ment in irregularities of the stream bed. A search of the stream bedduring low water yielded about 700 artifacts which will be referred toas the "stream bed collection." "WHiile these are not useful for strati-graphic purposes, they extend the typological range of some of thematerial classes found in situ in the midden. Several of the types werenot duplicated in any of the excavations, however, and their exactstratigraphic relationships must await further research in the area.At one time the midden had been used as a barnyard, the highestpart being occupied by a pigpen. This structure was carried awayby a flood several years ago, and there remained no surface indicationof its ever having been there. Several iron nails found in the upperfew centimeters of the deposit, however, may be attributed to thissource.A large tree on the south side of the site was chosen for a datum point,and a north-south base line was established from this datum throughthe center of the site to the creek scarp (fig. 8) . The site was laid out Riv. Bas. Sur.Pap. No. 4] ADDICKS BASIN?WHEAT 169in a grid of 1 m. (3.28 feet) squares numbered from the datum northalong the base line and east or west along lines intersecting the baseline at right angles. The stake nearest the datum became the desig-nator for that square. Excavation was by arbitrary levels of 15 cm.(0.49 foot). All material from each 15 cm. (0.49 foot) level and 1m. (3.28 feet) square was screened and the artifacts segregated. Figure 8.?Contour map and plan of excavation, Doerlng site (42/66A6-2).Depth was measured from surface at the level of the designator stakefor that square. Because of the relatively gentle slope of the middenthis system, in effect, resulted in the peeling of the deposit in 15 cm.(0.49 foot) layers. Profiles were measured by the use of a carpenter'shand level, the drop or rise being measured from each stake with arigid steel tape. These profiles were then plotted on graph paper,the profiles along OE/W, 14N/S, and 17N/S being used as controlsfor the construction of a contour map (fig. 8). 170 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154The physical stratigraphy of the Doering site is simple (fig. 9).From the surface to a depth of 20 cm. (0.65 foot) the deposit consistsof midden soil, sandy in texture, light tan in color, and quite friable.The lightness of color may be the result of weathering out or bleachingof the humic content by oxidation in place. However, occupation,judging from the artifact content, seems to have been less frequent orof lesser density during the deposition of this material, and the colormay, therefore, be the result of less inclusion of organic matter. PROFILE AT 5 EAST PROFILE AT EAST -WEST ^E 5E PROFILE AT 14 NORTH PROFILE AT 17 NORTHFigure 9.?Profiles, Doering site (42/66A6-2). Light stippling, weathered midden zone; '^horizontal hatching, midden deposit; black, sterile clay knoll; diagonal hatching,unexcavated.Below 20 cm. (0.65 foot) the organic discoloration of the middenincreases to the depth of 75 cm. (2.46 feet) . There is less sand admix-ture, and clay content increases. From 45 to 75 cm. (1.47 to 2.46 feet)the midden is somewhat indurated, and this zone marks the period ofthe most concentrated occupation. Animal bone fragments, includ-ing human bones, were most frequent at this level, although they occurin considerable quantity in the lower levels as well. Pap.N^o^l]"'^* ADDICKS BASIN?WHEAT 171From 75 cm. (2.46 feet) to the bottom of the midden deposit the soilis dark in color, heavy in consistency, and contains much moisture,making it difficult to screen. It is primarily of clay with decreasingorganic content near the base of the deposit. A blending of the mid-den soil with the upper part of the clay knoll upon which it rests masksthe precise transition from the one to the other. Below 1.20 m.(3.93 feet) the clay, which is devoid of artifacts or other occupationaldebris, is dense and indurated, of yellow color, and contains ferrousinclusions in the form of pebbles and root molds. It underlies themidden to the bottom of the creek and to an unknown depth below.Artifacts were relatively fewer in the level from grass roots to20 cm. (0.65 foot). Projectile points of Perdiz^ Pointed Stem, withits related types, and Gary Stemmed were present in about equalquantities. Pottery was predominant over lithic material.Below 20 cm. (0.65 foot) and down to 90 cm. (2.95 feet), the GaryStemmed points numerically predominated over other types. At 90cm. they were replaced in frequency by expanding stem forms. Pot-tery was relatively abundant to a depth of 75 cm. (2.46 feet). Belowthis depth, however, it was absent or extremely rare.Four burials were found in the Doering site. Three of these wereflexed while the other was loosely semiflexed. Although all of theseinterments probably were made in pits, it was not possible to definethem, except in Burial 3, because of the homogeneity of the middendeposit. This burial was intrusive through the lower part of Burial2, and in that section where it cut through the earlier burial, an arcof the pit was defined by the bones removed. No artifacts weredefinitely associated with any of the burials. Bone condition wasuniformly poor, probably due to soil conditions.Burial No. 1 (fig. 10, a) : Location: 18N/0E and 18N/1E.Depth: 90 cm. (2.95 feet) to skull.Matrix: Midden earth.Position: Semiflexed on right side ; head oriented to east, facing north.Arms: Upper right arm along side, entire lower arm absent; left armloosely folded across the abdominal region ; hands missing.Legs: Right leg loosely flexed ; left leg on top of right leg; feet missing.Age: Adult.Sex: Female (?).Associated Ohjects: A number of deer bones and two broken flint artifactsoccurred near the burial, but probably represent midden debris ratherthan intentional burial offerings.Remarks: This burial was much disturbed, the skull broken and scatteredover an area about 1 m. (3.28 feet) square. Many bones were missingand the rest fragmentary. * See chapter on Material Culture for complete information on these named projectilepoint types. 172 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154 ^. I9N-IE C=^/^ \ s^^-^ ^ t^^ ^ I0N-4E Figure 10.?Burials, Doering site (42/66A6-2). A, Burial 1. 5, Burials 2 and 3. C,Burial 4, pip.Na/]"'" ADDICKS BASIN?WHEAT 173Burial No. 2 (fig. 10, 6) : Location: 15N/3E.Depth: 32 cm. (1.04 feet) to skull.Matrix: Midden earth.Position: Tightly flexed on back ; head oriented to west, facing south.Arms: Right arm folded closely over the right leg; left arm looselyfolded, probably over the left leg at time of interment; both handsmissing.Leys: Right leg pulled tightly back against body; left leg and left in-nominate bone removed at time of intrusion of Burial No. 3.Age: Adult.Bern: Female.Associated Objects: None.Remarks: The lower left part of the body of this burial had been removedby the intrusion of the pit for Burial No. 3.Burial No. 3 (fig. 10, 6) : Location : ir)N/3E-ir)N/4E.Depth: 33 cm. (1.08 feet) to skull.Matrix: Midden earth.Position: Flexed on left side; head oriented to east, facing south.Arms: Right arm i>laeed across the legs ; left arm incomplete, only afragment of humerus represented ; hands missing.Legs: Right leg on top of left leg, both pulled against body.Age: Adult.liex: Male.Assooiatcd Objects: None.Remarks: This burial was inhumed in a pit intrusive through the lower leftside of Burial No. 2. Large roots penetrating the skull had forced itapart at the sutures and elsewhere had broken the long bones.Burial No. 4 (fig. 10, c) : Location: 14N/0W.Depth: ii'-i cm. (1.08 feet) to skull.Matrix: Midden earth.Position: Loosely flexed on back ; head oriented to southwest, facing north-west.Arms: Right arm alongside the body and partially underneath the rightleg; left arm folded across the lower part of the body under thelegs.Legs: Both legs loosely flexed to the right side of the body.Age: Adult.Sex: Male.Associated Objects: A large double-pointed drill was found beneath the body,but was not certainly included as grave goods.Remarks: Some of the bones were penetrated by root growth, but the preser-vation was generally better than that of any other burial from thissite. Several of the long bones displayed pathological growths andlesions. SITE 42/66A6-5Neither the precise location nor the nature of site 42/66AG-5 isknown, since it was destroyed as a consequence of the construction 174 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154work of the Addicks Dam embankment (fig. 5) . A number of stoneartifacts were collected from a scraped surface over a relatively re-stricted area on the west side of South Mayde Creek, immediatelyadjacent to the stream and between the present dam embankment andthe borrow pit that parallels it on the inside of the dam. Inclusionsof dark earth were visible in the embankment fill. However, no arti-facts were found in them, and the question of whether they representmidden debris remains unanswered.Projectile points, fragmentary blades, scrapers, and worked flakes Icomprise the artifacts found at this site. No potsherds were recovered,although unshaped lumps of burnt clay were noted. During the con-struction of the dam, earth-moving machines had cast up and de-molished an unknown number of burials. No artifacts were noted atthe time, however, and the area was presumed to have been an un- 'recorded Caucasian cemetery.From the point where it is joined by Horsepen Creek, LanghamCreek flows in a general southerly to southwesterly direction to its 'confluence with Bear Creek (fig. 5) . Below this, it follows a meander-ing course, its general trend slightly east of south, until it debauchesinto South Mayde Creek which flows in from the west. Three sites .are located along Langham Creek. Two of these, site 42/66A6-4: and 'the Kobs site (42/66A6-3), are upstream in the order named from fthe South Mayde Creek-Langham Creek junction but below the mouthof Bear Creek. The Grisbee site (42/66A3-1) is on Langham Creekjust below the mouth of Horsepen Creek.SITE 42/66A6-4Site 42/66A6-4 is located on the south side of an east to west 3meander of Langham Creek, about 150 m. (492.12 feet) above its -confluence with South Madye Creek (fig. 5). A troughlike depres- ^sion lies between the midden and the stream some 10 m. (32.80 feet) |lto the north (fig. 11) . It is parallel both to the stream and the longaxis of the midden and probably represents the stream channel at |the time the site was occupied. About 20 m. (65.61 feet) east of the Imidden at the point where the creek turns west from its more normal ';southerly course it is joined by an intermittent contributary from the 1northeast. Rather open woods of oak and large pine lie to the east Iand the stream course is lined by willow trees. The midden is 27 m. t(88.58 feet) along its east-west axis, 15 m. (49.21 feet) across the ;north-south axis, and rises approximately 60 cm. (1.96 feet) above the 'flood plain. It is somewhat higher at the eastern end and slopes \gently off to the south and west. There is an irregular, shallow .|depression in the southwest quadrant of the site where the former marks the period of densest occupation.Conformably underlying the midden proper is a knoll of unknowndepth composed of very light-colored, water-laid sand. As in theDoering site, the precise transition from the knoll to the overlyingmidden deposit is masked by the blending of the sand composing theknoll with the lower part of the midden. Although artifacts occurbut infrequently in this stratum, they nevertheless indicate limitedoccupation of the site during a period when the knoll was stillaggrading.Artifacts from the Kobs site were most numerous in the dark mid-den zone. The majority of the projectile points in this stratum werePerdiz Pointed Stem, Alba Barbed, and Eddy and Scalhoni Stemmed.At the base of the midden deposit these were of about equal occur-rence with Gary Stemmed points, indicating that the transition fromthe use of the large points to the smaller ones was already well underway. Pottery occurred frequently in this layer.All the proje<;tile points from the sand knoll are large, the majoritybeing Gary Stemmed points. Pottery was quite rare in this stratum.In the upper part of the knoll several sherds of a Tchefuncte Stampedpottery jar were found, giving us one of our most important chrono-logical clues and affording a tentative cross-dating with the earlyTchefuncte cultures of the Lower Mississippi Valley.Three burials were found in the Kobs site. Two of these wereflexed, while the third was semiflexed. Two were buried in pits duginto the sand knoll, easily recognized because of the dark midden fill. Rlv. Bas. Sur.P;il>. No. 4] ADDICKS BASIN?WHEAT 179 fI i 180 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154 Figure 16?Burials, Kobs site (42/66A6-3). A. Burial S. B, Burial 6. C, Burial 7. pS.N^o^.'4T'^' ADDICKS basin WHEAT 181The third burial probably was also interred in a pit, but inasmuch asthere was no penetration into the white sand of the knoll, none couldbe defined. Burial numbers were continued serially from those ofthe Doering site.Burial No. 5 (fig. 16, a) :Location: 9N/7W.Depth: 72 cm. (2.36 feet) to skull.Matrix: Midden fill in oval pit 85 cm. (2.78 feet) long and of indeterminatewidth and depth. Pit well defined on south side by contrast ofmidden material with white sand of the knoll.Position: Flexed on back; head oriented to southwest, facing southeast.Arms: Right arm tightly folded against itself, hand to chin; left armfolded alongside chest, hand to left shoulder.Legs: The legs were pulled close to and on top of body with lower legsrotated medially so that the feet crossed each other above thepelvis.Age: Adult.8eip: Female.Associated Objects: Half of a small turtle shell with a few pebbles beneathit, found on top of the chest, may have been a rattle.Remarks: Bone condition fair; some breakage due to ground pressure.Burial No. 6 (fig. 16, 6) : Location: 11N/7W-12N/7W-11N/8W-12N/8W, center of burial pit almostdirectly under designator stake for square 12N/8W.Depth: 70 cm. (2.29 feet) to skull.Matrix: Midden fill in elongated oval pit, 120 cm. (3.93 feet) long, 67 cm.(2.19 feet) wide, 40 cm. (1.31 feet) deep, excavated from middeninto white sand of knoll.Position: Semiflexed on right side; head oriented to southeast, facing north-east.Arms: Upper right arm across chest to front of body, lower arm andhand missing; left arm folded alongside body, hand to leftshoulder.Legs: Upper legs form approximately a 140? angle with the body, thelower legs folded tightly against them.Age: Adult.Sex: Female.Associated Objects: None.Remarks: A number of bones were missing, including the left and rightradii, left ulna, the lower vertebrae, pelvis, hand and foot bones, andmost of the ribs. The leg bones were separated from the upper bodyby a distance much greater than normal. This, together with thenumber of missing bones, suggests the possibility that the body waspartly decomposed before burial and that some attempt at rearticula-tion had been made.Burial No. 7 (fig. 16, c) : Location: 10N/9W.Depth: 40 cm. (1.31 feet) to skull.Matrix: Midden fill. 182 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154Position: Flexed on left side, almost on its back ; head oriented north-north-east, facing southeast.Arms: Right arm folded along side, hand (missing) between rightshoulder and chin ; left arm tightly folded beneath the left leg,hand under left side of face.Legs: Pulled close to chest, feet at pelvis.Age: Adult.Sex: Male.Associated Objects: None.Remarks: Best preserved skeleton from the site. No burial pit visible be-cause of homogeneity of midden earth.THE GRISBEE SITE (42/66A3-1)The Grisbee site is located at the north end of a westward meanderof Langham Creek about 150 m. (492.12 feet) south of its junctionwith Horsepen Creek (fig. 5) . At the south end of the meander, pot-sherds, lithic artifacts, and other camp debris are eroding from astratum about 10 cm. (0.32 foot) in thickness underlying an average20 cm. (0.75 foot) of alluvial fill. The area encompassed by themeander supports a cover of small pine trees, many of them of secondgrowth. Along the stream are oaks and other hardwood trees, aswell as willows and a few larger pines. Approximately 20 m. (65.61feet) above the site an intermittent contributary joins the main channelof Langham Creek from the northeast.The site consists of a midden deposit 90 cm. (2.95 feet) thick over-lying a kidney-shaped knoll which parallels the bend of the streamon the southeast bank as it turns into the westward meander (fig. 17) . FiGUR E 17.?Sketch map, Grisbee site (42/66A3-1), showing test pit. Kiv. Bas. SillPap. No. 4] ADDICKS BASIN WHEAT 18^The site is 15 m. (49.21 feet) across its northeast-southwest axis and27 m. (88.58 feet) across its northwest-southeast axis. The north-eastern end rises sharply to approximately 45 cm. (1.47 feet) abovethe flood plain and the midden slopes gently to the south and west.The site is undergoing some lateral erosion on the edge next to thecreek in the form of a miniature guUey cutting headward into thecentral portion of the midden. The remainder of the surface is sta-bilized by an undergrowth of thorny vines and low bushes. Threelarge oak trees are growing on the lower slopes of the midden.A stratigraphic test pit 1.20 m. (3.93 feet) square was excavatedby 15 cm. (0.49 foot) arbitrary levels, approximately in the centerof the midden (fig. 18). From the surface to a depth of 15 cm.(0.49 foot) was an indurated crust of topsoil and sparse midden de-posit, compacted, perhaps, by the pressure of the hooves of grazingcattle (fig. 16, h). Below this crust the midden earth was cohesive. 61218243036 153045607590o3 Indurated Crust Midder) Deposit Sterile ClayUnexcavated Figure 18.?Stratigraphic section, Grisbee site (42/66A3-1).From 45 to 75 cm. (1.47 to 2.46 feet) the organic content of the depositgreatly increased, making it more friable and perceptibly darker incolor. Most of the artifacts obtained at this site come from this level,which probabl}^ represents the period of heaviest occupation. Belowthis level the midden continued to a total depth of 90 cm. (2.95 feet)where it rested conformably upon the sterile clay composing the knoll.The uppermost 60 cm. (1.96 feet) of the deposit contained PerdizPointed Stem and other upper level types, as well as a few GaryStemmed projectile points. Below this level only the larger pointswere found. Potsherds occurred in all levels of the midden but wereinfrequent in level 75-90. One sherd from the base of a flat-bottomedvessel and two small rodlike pieces of pottery that may have beenlimbs of pottery figurines were unusual finds in this site. 184 BUREAU OF AMERICAN ETHNOLOGY I Bull. 154MATERIAL CULTUREApproximately 7,000 artifacts were recovered from the AddicksBasin sites, the great majority deriving from the Doering and Kobsmiddens, and most of the remainder from those sites surveyed andtested by the survey. The group of artifacts from the Addicks"Mound" (42/66A6-1) were donated to the Smithsonian Institutionby Marshall Black of Houston.Twenty-five artifacts were made of bone, antler, or shell by cutting,grinding, drilling, and polishing; 45 of granular stone by peckingor abrasion, tubular drilling, and polishing; and 1,870 of flint, chert,mineralized wood, chalcedony, and other crypto-crystalline, siliceousmaterials by percussion and pressure flaking. A few specimens ofearthy minerals were, for the most part, of unaltered tabular ornodular form. The rest of the collection consists of about 5,000potsherds and a few miscellaneous objects of clay.POTTERY COMPLEXThe indigenous pottery of the region about the northern end ofGalveston Ba}' consists of a single, highly variable ware divided intotwo subtypes, decoration serving as the primary criterion for theseparation. These subtypes are tentatively named Goose Creek Plainand Goose Creek Incised (Campbell, n. d.). With the exception ofa few sherds of intrusive wares, described below, all ceramic ma-terial falls into the above categories.GOOSE CREEK PLAINPaste.Method of Manufacture: Coiled. Sherds frequently break along coil junc-tions and often show oxidation between the coils where they werenot sufficiently wedged before firing. Coils average 8 mm. (0.31 inch)in width.Temper: Usually consists of sand of rounded quartz particles of variablesize, although in some cases the sand is fine and of regular size. A fewsherds, probably from two vessels, show calcium phosphate temperparticles, and a number of sherds from the upper levels of the Kobssite appear to have clay or sherd temper. It seems probable that thecalcium phosphate tempered sherds are from intrusive vessels.Texture: Medium coarse to fine and compact, determined by the amountand size of the sand-temper particles. Some sherds seem to be almostpure sand, and the texture is like that of fine-grained sandstone. Afew, however, are compact and fine-grained, the sand-temper particlesfirmly embedded in the clay. In cross section many of the sherds showcoil laminae diagonal to the vertical axis of the vessel wall.Hardness: Varies from 1.5 to 3.5 (Mobs' Scale), the average being about 2.5.Color: Ranges from orange-red through reddish browns, brownish grays,and into very dark grays or black. Cores vary from reddish to black. pip.Na'/]""^" ADDICKS BASIN?WHEAT 185Ten sherds, selected at random and matched with the standard colorsof Maerz and Paul,^ show the following variations : 4A11, 5A11, 5E12,6H11, 12C2, 14E6, 15A4, 15C8, 16A2, and 16A3. While this by nomeans exhausts the range of nuance of color, it shows the generalvariation. All the variations in color apparently are the result ofuneven firing. A few sherds show remnants of a dark red wash orfilm applied on both interior and exterior.Surface finish.Usually smooth on both interior and exterior, with some variation of wallthickness resulting from uneven scraping and in some cases, perhaps,by finishing in the paddle-and-anvil technique. Marks produced by adenticulated instrument are occasionally visible on one or both sur-faces. Exteriors sometimes have the surface scraped diagonally to thevertical axis of the vessel; and the interiors of a few sherds showvertical polishing marks on a dense black surface. Floating of fineparticles of the paste to the surface was commonly used to produce asmooth finish. Such surfaces erode easily. Crackling of the surfaceoccurs occasionally.Form.Wall Thickness: Range is 2 mm. to 10 mm. (0.08 to 0.39 inch) ; average,6 mm. (0.23 inch).Rim: Usually vertical, a few slightly outcurving. Thickness of the wallis ordinarily decreased toward the lip, usually from the vessel's interior,but occasionally from both interior and exterior (fig. 19).Lip: Normally allowed to come to a relatively sharp edge (fig. 19, o-e),but sometimes flattened (fig. 19, m-t) or rounded (fig. 19, f-l). Lipnotching is common, usually on the interior edge, but occasionally onthe outer edge, and sometimes across the lip. Rim sherds from twovessels show notches formed by pinching with opposed thumb andfingernails creating a deeply scalloped effect. The lip of one rimsherd has been enlarged to form a lug (fig. 19, v, w).Body: No whole vessels were found, but two shapes have been determinedfrom partly reconstructed pieces. One of these is a deep bowl withfull round sides incurving slightly at the rim (pi. 31, d; fig. 20, d).The second form is a small jar with slightly outcurving rim and sidesthat curve gently to a full round bottom (fig. 20, a).Size: Forty-four rim sherds (including five partially restored vessels)have a range of diameters from 6 cm. (2.37 inches) to 44 cm. (17.50inches). Twenty-four of these fall in the 20 to 30 cm. (7.81 to 11.80inches) range, the remainder being distributed about equally. Depthmay be defined only in the five partially restored vessels and variesfrom circa 15 to 25 cm. (5.90 to 9.68 inches).Bases: Perhaps the most common base form is round ; but pointed or conoidalforms are of frequent occurence (Fig. 21). Bases are usually thickenedat the center and frequently form a distinct node. Two bases are roundand flat, the transition from base to wall marked by a distinct angle,and another apparently flattened squarish in outline. * Maerz and Paul, 1930. The notation system, refers to plate number, file, and rank,in the order given. 186 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154 a IP T I J m ni\i Iin u wFigure 19.?Pottery: Rim and lip profiles, a-e. Sharp edge. /-/, Rounded edge, m-t,Flattened edge. ?, Sharp outcurve. v, w, Lug. Interiors to left. Riv. Bas. Sur.Pap. No. 4] ADDICKS BASIN?WHEAT 187 Figure 20.?Pottery: Vessel shapes. 188 BUREAU OF AMERICAN ETHNOLOGY !Biill. 154 Figure 21.?Pottery: Base shapes. pip.Na'/]'"^' ADDICKS BASIN?WHEAT 189Distribution.Insufficient data are at hand to define the complete distribution of the type,but its present known range is a zone some 30 miles wide along thenorthern periphery of Galveston Bay and an unknown distance northeastand northwest. GOOSE CREEK INCISEDPaste.Paste features as described above for Goose Creek Plain.Surface finish.As described above.Decoration.The zone of decoration is confined to the rim section in all specimens recov-ered. Design elements are simple and are produced by incising the wetclay with a sharp implement. A few sherds combine incision withpunctation, and sherds from a single vessel show punctation used aloneas the decorative technique.The most common form of design is single, dual, or, infrequently, multiplelines around the circmnference of the vessel just below and parallel tothe lip (pi. 32, a-g). These vary from fine, evenly incised lines to crudeones drawn with a rough tool. A variation of this style is the inclusionof a wavy or zig-zag line between parallel framing lines (pi. 32, h).These latter elements are sometimes pendant from a straight top lineand occasionally appear singly or paired, independently of straightframing lines (pi. 32, i-m). The character of execution of these linesvaries considerably.Triangles, pendant from a straight line or alternating between doubleparallel lines, are frequently hatched and sometimes cross-hatched (pi.33, flh-c). Double concentric triangles occur infrequently (pi. 33, d, e).Ladderlike elements and parallel rows of punctation sometimes formtriangles or other geometric figures (pi. 33, g-i). The ladderlike ele-ments are formed by parallel incised lines which are connected by short,equidistant transverse lines at more or less regular intervals. Puncta-tions are usually quite small, about 1 mm. (0.039 inch) wide and 1.5mm. (0.059 inch) long, and are made with a small pointed tool. On onerim sherd, parallel paired rows of three punctuations each, made with asmall hollow instrument (perhaps a cane section) pressed into the clayat an angle so as to produce a crescentic punctation, are pendant fromthe lowermost of three parallel incised straight lines (pi. 33, ;).Form.Rim: As described above.Lip: As described above.Body: Two shapes have been distinguished for Goose Creek Incised from thepartial restoration of three vessels. One form is represented by twodeep, narrow jars with nearly vertical sides (pi. 31, a, h; fig. 20, c) . Thejar is slightly constricted below the rim, but the basic shape is semi-cylindrical. The bottom converges gently to a rounded or conoidal base.The second form is a wide-mouth jar which expands slightly from themouth to an elongate, globular body and a full round base (pi. 31, c;fig. 20, d) . 190 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154TCHEFUNCTE STAMPEDFive sherds have been identified as Tchefuncte Stamped (Ford and Quimby,1945, pp. 56-57 ; pi. 30, f) ware (pi. 33, k)." These sherds, all part of the rim of asingle vessel, vary somewhat from the usual Tchefuncte type in their relativethinness, the deepness of the lip notching, and in that the temper is sand. Itshould be stated that sand and grit tempering was not unknown, either in theTchefuncte period or in Tchefuncte Stamped ware itself. The rim is slightly ,flaring and the lip is roughly notched on the outside. Below the rim and parallel 'to it is a row of rocker-stamped impressions produced by a forked implement,the lower part being more heavily impressed than the upper. Pendant from thisare parallel vertical rows of rocker-stamped impressions. These rows measure j7 mm. (0.28 inch) wide, and the distance between rows varies from 5 to 10 mm.(0.20 to 0.39 inch). A single crack-lacing hole, drilled from the outside, is 25mm. (0.96 incli) below the lip.UNIDENTIFIED WARESFifteen sherds from the wall of a single vessel have not been identified as totype (pi. 33, I, m). The entire exterior surface is decorated by lines formedwith a denticulated implement and arranged in rows parallel to the rim of thevessel, then crossed by both vertical and diagonal rows of parallel lines.A single sherd has asphalt adhering to it and may be intrusive from the 'jRockport Focus (Sayles, 1935, pp. 35-40, table 5). iMISCELLANEOUS OBJECTS OF CLAYClay Rods.?Four small, round pieces of pottery range from 7 to 9 mm. (0.28 to0.36 inch) in diameter. Length cannot be determined accurately since allspecimens are fragmentary. Two are broken at both ends, one terminates in arounding point, and the other has a tapering point. They do not appear to havebeen handles or lugs for pots, and it is difficult to imagine a utilitarian functionfor them. Although no identifiable figurines were found, these objects may belimbs from broken specimens. These are from levels 30-45 and 45-60 in theGrisbee site, and levels 30-60 and 60-75 in the Kobs site.Two somewhat flattened rods of clay are twisted across each other and arepressed together to form a single object. It is from level 30-45 in the Kobs site.Worked Sherd.?A single sherd, roughly rectangular in outline, has a smallprojection on one end, formed by notches ground in from the sides. This mayhave served as a pendant, CERAMIC STRATIGRAPHYBecause the apparent paucity of sherd material in the lower levelsof the Doering and Kobs middens implied the possibility of a non-ceramic horizon, stratitests were made to determine the validity ofsuch an hypothesis. Five contiguous meter squares, each penetratingto sterile native soil beneath the midden, were chosen at both the Kobs ' George I. Quimby has examined these sherds and confirmed the Identification. Dr. .Tames B. Griffin does not fully concur In this identification. However, basically the sherdsso closely resemble Tchefuncte Stamped, and inasmuch as there is no other known typeto which they can be assigned, I have retained the original allocntion. Riv. Bas. Sur.Pap. No. 4] ADDICKS BASIN WHEAT 191and Doering sites to serve as check squares. Test pit 2 at site42/66A6^, and test pit 1 at the Grisbee site, while not equivalent incubature, are included to show comparative data from other sites.*Table 1 shows the sherd count from the stratitest squares by levelfrom each of the sites.Table 1. -Sherd count from stratitest squares at Doering, Kobs, 4^/66A6-4; andGrisbee sHes 192 BUREAU OF AMERICAN ETHNOLOGY fBuU. 154-distribution upward from level 75-90, depending upon the interpre-tation given the occurrence of sherds below 60-75, at the Doering'site. 'The actual percentage of incised to plain ware cannot be assessed jaccurately since decoration is confined to the rim and upper part ofthe body, and basal and lower body sherds may have come from either]plain or incised ware. However, judging from rim sherds, the plainfware clearly predominated. SAlthough sharp lipped profiles predominate numerically over bothlflattened and rounded forms at all levels, there is no indication that'they are appreciably older (table 2). There are too few rim sherds'at the lower levels to supply satisfactory chronologic criteria. \>Table 2 shows the vertical distribution of the various rim profiles. : Table 2. ? Frequency distribution by site and level of rim profiles Riv. Bas Sur. ADDICKS BASIN WHEAT 193Pap. No. 4 ]level represent material displaced downward by rodent disturbanceor the disturbance incident to dwelling on the site. The possibilityof dual occupancy of the site may be ruled out since there is no breakin the stratigraphy, and the lithic complex forms a continuum. Onthe whole, the third possibility seems the most likely explanation to thewriter. Further support is vouchsafed this interpretation by anexamination of the other materials from the lower levels. Leavingaside any typological considerations, the bulk of the lithic complexes,together with the bone and antler materials, effectively demonstratethis.An extraction of data from tables 3, 4, 6, and 7 (below) yields thefollowing results in comparison with the total number of sherds fromthe lower levels : Miscellaneousartifacts SherdsLevel 60-75 149 49Level 75-90 HO 3Level 90-105 84 5Level 105-120 7 1Thus, the clear predominance of bone, antler, and lithic materialstends to bear out the essentially nonceramic character of the lowerlevels.By way of contrast, in the earliest levels of the remaining sites,pottery is already strongly represented in comparison with the lithiccomplex. Further, the increase in frequency of sherds from the lowerlevels upward to the maximum seems correlated with the increasingfrequency of occupation of the sites as attested by a similar increasein the bone, antler, shell, and lithic remains. Taken together, thesedata show that pottery was already an important component of thetotal culture complex when occupation of the Kobs and Grisbee sitesand site 42/66A6-4 began. Upon this basis the lower stratigraphiclevels of these sites would correlate with the middle stratigraphic levelsof the Doering site.Chronologically, two factors are of importance. The first of theseis the presence of Tchefuncte Stamped ware, intrusive into the potterycomplex of the lower level of the Kobs site. This intrusion would seemto fix a relative point in time for the occupation of the Kobs site. Sinceit is found in an assemblage in which pottery is already significant, itis probable that the intrusion occurred comparatively late in theTchefuncte period. If one accepts the A. D. 500-900 dating assignedthe Tchefuncte by Martin, Quimby, and Collier (1947, p. 401), a datenearer the A. D. 900 terminus would seem most likely for the earliestdeposits at the Kobs site. Recently, radiocarbon dates have beenpublished for the Tchefuncte Culture (Arnold and Libby, 1951). 194 BUREAU OF AMERICAN ETHNOLOGY fBull. 154These dates, A. D. 1317? 150, and 717? 250 are confusing. The earlier |date falls near the middle range originally postulated, but the later ione is much too late to be accepted at face value. In terms of datesfor other southeastern cultures of a similar technological stage ofadvancement both seem too late. On the whole, it would seem best towithhold acceptance of either of the radiocarbon dates until furtherinformation is forthcoming. iThe second important factor is the apparent introduction of potteryto the established lithic complex of the lower levels of the Doeringsite, an event usually held to delimit the Archaic from the later South-eastern horizons (Sears, 1948, pp. 122-124). Considering the datafrom the Kobs site as our most secure chronological tie for the middleand early levels, it seems not unlikely that the lower, nonceramiclevels of the Doering site antedate those of the Kobs site by perhapsseveral centuries.Of possible chronological interest, although of minor importance, isthe stratigraphically late appearance of new aplastics, calcium phos-phate, and perhaps sherd temper. However, sand temper continues topredominate here as in the lower levels of all sites examined. It is \this essential uniformity of Goose Creek wares throughout both their . vertical and areal distribution that renders them, per se, of little value ''as time markers.Briefly, the most salient structural features of the Goose Creek jwares are : Sand-tempered paste ; vessels made by the coiling technique 'and usually thinned by scraping; surface normally smoothed by float- ' ing the smaller particles to the surface to form a mechanical slip, but ; occasional other treatments ; few and simple shapes, the most charac- 1 teristic being elongated globular jars with slightly constricted necks jand wide mouths, and large, deep, wide-mouthed bowls of simple con- ;tour ; bases are round, truncated conoidal, or conoidal, and frequently jterminate in a distinct node ; rims are simple and direct, for the most ; part, and have rounded, flattened, or relatively sharp-edged lips, inorder of their increasing importance.As a preliminary statement, it may be noted that these traits standin sharp contrast to the Mississippi culture pottery complex (Deuel, :1935, pp. 437 et seq.). The pottery complex of the Addicks Basin ;belongs to the predominantly smooth or plain surface tradition of the !Southeast.^ This tradition, developing out of the various fiber-tem- jpered plain types, consists of a series of clay and sand-tempered wares, ITchefuncte Plain, Mandeville Plain, St. Johns, and Weeden Island |Plain, and their associated incised and stamped Avares,^ to which may .now be added the Goose Creek wares. I * GriflSn, 1949, personal communication. rSi. N^o.' Ir* ADDICKS BASIN WHEAT 195The specific resemblance of the elongate globular jar with conoidalbase to jars of the Northern Woodland groups (Anonymous, 1943, p.396) raises an interesting problem. The shape is not common in theSoutheast, occurring only in Florida (Moore, 1902, pp. 159, 300; figs.52, 270). However, the surface treatment of all pottery recoveredfrom the Addicks Basin sites is typically Southeastern, and only theshape of these jars is reminiscent of the Woodland vessels. Paddlestamping, the prevalent surface treatment of Woodland pottery, wasnot found. The question of the ultimate derivation of this shape isone for a specialist and lies beyond the limits of this paper.The decoration of Goose Creek Incised is confined to the rim andupper wall of the vessel. Lip notching, incising, and punctation areused singl}' or in combination to form simple geometric designs.These decorative techniques are typically Southeastern, although theyalso occur elsewhere.Specific relationships at the early levels seem to be confined to theTchefuncte Culture of the Lower Mississippi Valley. From the mid-dle levels of the Doering site a single sherd which has a thick, flattenedglobule of bitumen or asphalt adhering to the surface may be intrusivefrom the Rockport Focus (Sayles, 1935, pp. 35-40; table 5) whichcenters along the Gulf Coast south and west of the Addicks DamBasin.? It is possible, however, that the asphaltum is accidental,since it was used to a considerable extent in the hafting of projectilepoints. On the late horizon an unidentified trade ware, characterizedby an all-over design produced by a dentate instrument, and a morecompact paste, is represented by sherds from a single vessel. It can-not be related to a known source. Also late in the occupancy of theAddicks sites calcium phosphate and perhaps sherd temper makean appearance, but whether these aplastics were used locally in themanufacture of pottery or represent trade relationships is not yetdetermined.Pearce (1932 a, p. 51) records much use of slips and paints onthe pottery of the Galveston Bay region. This was not confirmedby our excavations. Floating of the surface to produce a mechanicalslip is characteristic, but there was no evidence of a true slip. Paintappears to be confined to a wash or film of red occasionally presenton both interior and exterior of vessels. This trait may be relatedto the red filming technique of the Tchefuncte Culture, although thetempering and thickness differ somewhat (Griffin, 1949, personalcommunication).Gatschet (1891, pp. 24, 69) records the presence of pottery forthe Karankawa Indians but leaves no description, so we cannotexamine the possible affiliations of the ware with the Goose Creek ? Rockport Phase of Sayles. 196 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154 wares. Archeologically, sites presumed to be Karankawan have beeninvestigated by Sayles (1935, pp. 35-40), Martin (1929, pp. 50-57;1931, pp. 53-56), and Potter (1930, pp. 41-44). While asphalt, usedboth as a paint in decoration and for producing a waterproof coatingon the interior of vessels, may be considered the hallmark of Karan-kawa pottery, Martin (1929, p. 56) distinguishes one type which differsin paste and treatment. The paste is described as being harder andless porous, and seldom shows the use of asphalt. The aplastic isnot identified, however, and it is impossible to determine the likeli-hood of linkages with the Goose Creek wares. Both lip notching andincising occur, as well as asphalt painting. Potter (1930, p. 42),however, states that nearly all decorated vessels are very thin-walled(2-4 mm. (0.08-0.16 inch)) and are reinforced or waterproofed byan interior coating of asphalt. Furthermore, the lip notching ispreponderantly deep notching or grooving completely across the lipof the vessel, in contrast to the typical shallow interior or exteriornotching characteristic of the Goose Creek ware (Potter, 1930, p. 43,pi. 7, Nos. 5-10) . Jar and bottle shapes occur in the Karankawa series ibut are significantly different. The small, deep, simple-contoured, 'wide-mouth bowl, while much smaller in the asphalt painted group,is a logical counterpart, in shape, to the larger bowl forms in the }Goose Creek series (Potter, 1930, p. 44, pi. 8, No. 15). Thus, while :there are significant differences, there are, nevertheless, some similari-ties of shape and decorative technique which suggest possible inter- |change of certain traits, at least during the late period from which \all the reconstructed vessels from the Addicks Basin have come. 'The Attakapan Phase suggested by Sayles (1935, p. 41) is charac- |terized by sand-tempered pottery decorated with incised lines, non- jdistinctive lithic complex, and usually the presence of European jtrade goods. Since no details are given, it is not possible to correlatethis phase with the sites excavated in the Addicks Basin. Further, 'none of the Addicks Basin sites produced any evidence of historic jcontact. However, as defined by Sayles, this phase may represent jthe historic focus of the culture of the peoples of the Addicks Basin Iarea. On the whole, the general paucity of extra-areal contacts pointsto the relative isolation of the Addicks Basin peoples. ILITHIG COMPLEX IIf cultures predominantly lithic in complexion are to be correlated |satisfactorily with each other and with groups whose history is largelytold in ceramic terms, then we must pay the same attention to arti-facts of stone, bone, shell, etc., that has yielded such valuable results 'in the study of pottery. We must arrive at an intelligent classifica- I pL^p.^o/lr' ADDICKS BASIN?WHEAT 197tion based not upon one or two features, such as outline or flakingtechnique, but upon the totality of inherent traits that make one groupof artifacts distinct from another. The range of variation within thetype must be realized, and distribution in space and time recognizedand plotted. It is incumbent that the word description of nonpotteryartifacts be as clear and accurate as possible and be fully documentedby photographs and drawings. When we have accumulated such acorpus of data we shall be in a position better to define and under-stand the relations of groups of people who, while differing in theireconomic adaptations, are alike in leaving a heritage limited toimperishable bits of stone and pottery.In this paper projectile points have been classified according tothe Tyi^ological Concept as defined by Krieger (1944, p. 271-288).The binominal system of nomenclature for projectile points in useby the Texas Council of Archeologists, and recently adopted by theSouthwestern Archeological Conference, Point of Pines, Ariz., hasbeen used in this report. This system, long used in pottery designa-tion, combines the type-site name and a word or phrase descriptiveof the most characteristic or consistent feature of the projectile point.As previously stated, named projectile point types are those which,for the most part, have been established by Kelley, Krieger, orCampbell, of the University of Texas, and the present specimenshave in each case been checked by them.^ One new type, Kobs Tri-angular, has been established on the data furnished by the AddicksDam Basin excavations. The remaining named types have beendesignated by several different investigators and will be discussedunder the proper headings below.A number of provisional types have been set up. These have beennumbered rather than named, inasmuch as our data are altogethertoo scanty to establish them on a firmer foundation. They may beconsidered as "working groups" in the sense that this term is used byKrieger (1944, p. 280, fig. 25). It is entirely probable that some ofthem may be consolidated into more inclusive types. It is equallypossible that fuller information may require further subdivision. Inany event, it has been thought best to describe them in detail and toplace them on record as a basis for future work in this and neighboringareas.For conciseness in description, data are presented in expandedoutline form, and certain descriptive conventions have been adopted.Points whose width equals or exceeds their length are considered to ' Although references to the names of several of these types have previously appeared inprint, the present paper constitutes their first description at length. However, type de-scriptions based on greater areal sampling are to be published by the responsible individualscited above. 198 BUREAU OF AMERICAN ETHNOLOGY IBull. 154be of broad proportions. Long or narrow proportioned points arethose in which length is as much as, or more than, twice the width.Points whose length-breadth relations are intermediate are designatedas of medium proportions. "Contracting stem" is used to describetangs w^hich increase in width from the base to the point of inter-section with the blade ; "expanding stem" implies the reverse condi-tion. A "Remarks" section has been included to cover such data asdeviations from norm and temporal and areal range. No attempthas been made to exliaust the literature concerning distribution, sincethe varying treatments and the gaps in areal coverage make suchdistribution studies hazardous at present. Distribution, where given,is simply to point out certain basic similarities of form, technique,and size of typologically similar artifacts in other areas in an attemptto generalize upon areas of possible influence.The stratigraphic sequence of projectile points in the Addicks DamBasin sites is of particular significance, since it provides a relativechronological framework by which the different sites may be tempo-rally equated. This is the more important since pottery, usuallythe most sensitive criterion, did not here reveal significant variationsin style or structure which might serve to distinguish the wares ofone period from those of a later date.Projectile points may be divided into two major groups : the gen-erally smaller and lighter arrow points whose occurrence is confinedto the upper levels of all the middens tested ; and the larger, heavierdart points found chiefly in the middle and lower levels. Certaintypes, represented only in the stream-bed collection, cannot be placedprecisely in the relative chronology and have, therefore, been groupedseparately until further excavation shall reveal their correct place-ment.The followmg tabulation shows the breakdown of the lithic com-plex into its component categories:Artifacts of stone : Chipped stone:Projectile points:Upper Level types:Perdiz Pointed Stem II3Scalhorn Stemmed 12Eddy Stemmed I7Alba Barbed 34Kobs Triangular _ 34Provisional Type 1 4Provisional Type 2_ 3Provisional Type 3 1Provisional Type 4 1Total Upper Level types 219 pip No/]""^' ADDICKS BASIN WHEAT 199\.rtifacts of stone?ContinuedChipped stone?ContinuedProjectile points?ContinuedMiddle and Lower Level types:Gary Stemmed 518Provisional Type 5 2Wells Contracting Stem 1Provisional Type 6 25Provisional Type 7 23Provisional Type 8 13Provisional Type 9 8Provisional Type 10-- 9Provisional Type 11 22Provisional Type 12- _ 28Provisional Type 13 6Provisional Type 14 8Provisional Type 15 1Provisional Type 16 2Provisional Type 17 5Provisional Type 18 15Provisional Type 19 18Provisional Type 20 6Pedernales Indented Base 2Provisional Type 21 2Provisional Type 22 12Provisional Type 23 21Provisional Type 24 3Clovis Fluted (?) 1Total Middle and Lower Level types 751Unplaced types:Provisional Type 25 1Provisional Type 26- 1Provisional Type 27 2Provisional Type 28- 1Provisional Type 29. _ 4Provisional Type 30- 6Provisio7ial Tvpe 31- 1Provisional Type 32 1Copena Point (?) 1Plainview Point 1Baird Beveled Point 2Nontypable fragments 284Total Unplaced types 305Total projectile points 1, 275 200 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154 UArtifacts of stone?ContinuedChipped stone?ContinuedImplements:Knives:Type la 2Type lb 1Type Ic 1Type Id 8Type 2 100Type 3 53Type 4. 13Type 5 6Type 6a 1Type tib 1Total knives 186Scrapers : End Scrapers, Stemmed 5End Scrapers, Large 3End Scrapers, Small 7Side Scrapers, Large 5Side Scrapers, Small 20Flake Scrapers 260Total scrapers 300Choppers : One end unmodified 20Both ends modified 54Chipped Celt 1Total choppers 75Drills:Type la 5 'Type lb 4 jType Ic 4Type 2a 3 ,Type 2b 3 ;Types 1 1Total drills 20 !Gravers 13 jEccentric flint 1 34Total implements 595Total chipped stone 1, 870i pS-N^o^.I^" ADDICKS basin?WHEAT 201\^rtifacts of stone?ContinuedGround stone:Atlatl weights (banner stones) 2Sandstone abraders or saws 41Milling stone 1Handstone 1Total ground stone artifacts 45Miscellaneous stone:Paint pigments 18Total artifacts of stone 1, 933PROJECTILE POINTSUPPER LEVEL TYPESPerdiz Pointed Stem (pi. 34, a-n); 113 Specimens.Proportions: Usually medium but sometimes long and slender.Blade: Edges normally straight, occasionally convex or concave, frequentlyserrated.Shoulders: Prominent to barbed.Stem: Most frequently long relative to total length of point, and contractingto a pointed or slightly rounding base.Average dimensions: Length, 33 mm. (1.31 inches) ; width, 19 mm. (0.75inch) ; thickness, 4 mm. (0.15 inch).Provenience: See table 3.Remarks: This type of point is the most characteristic of the upper levelsin all sites tested and excavated. They are small, light in weight, andsince they are commonly made from a thin flake, one side often showsmost of the flaking. Workmanship varies considerably, but they areusually well made. The Perdiz Pointed Stem is one of the type projec-tiles for tlie Toyah Focus (Kelley, 1947 a, p. 122) which Kelley datesat circa A. D. 1500. It is a component of the lithic complex of theFrankston Focus of northeast Texas which Krieger dates between1400 and 1600,' when it is replaced by Cuney Stemmed, a type notoccurring in the Addicks Dam Basin sites. The Perdiz Pointed Stemtype has a wide distribution over most of Texas, occurring from Trans-Pecos in the southwestern part of the state, across a broad central beltto northeastern Texas. These flnds represent a southeastern extensionof their known distribution.Scalhorn Stemmed (pi. 34, o-s) ; 12 specimens.Proportions: Long and narrow.Blade: Edges straight and frequently serrated.Shoulders: Straiglit to barbed, but usually obscured because of wide base.Stem: Expanding to flaring, formed by diagonal notches from basal corneror just above ; base straight to convex.Average dimensions: Length, 36 mm. (1.43 inches) ; width, 17 mm. (0.68inch) ; thickness, 3 mm. (0.12 inch).Provenience: See table 3. " Krieger, personal conference. 202 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154Remarks: Appear stratigraphically later than Perdiz Pointed Stem, bothhere and in the type area. Distribution limited to central and easternTexas, centering around Tyler.^"Table 3. ? Provenience of Upper Level types of projectile points i pip.N'o.'lr" ADDICKS BASIN?WHEAT 203Kobs Triangular (pi. 35, i, j) ; 34 specimens.Proportions : Medium to long.Blade: Usually concave from base to midblade, convex to tip; a few arestraiglit to convex, sometimes serrated ; basically triangular.Base: Usually convex.Average dimensions: Length, 45 mm. (1.78 inches) ; width, 22 mm. (0.87inch) ; thickness, 4 mm. (0.15 inch).Provenienee: See table 3.Remarks: These points have all the sliape and size characteristics of theAlba Barbed type except the notching, which produces both stem andbarbs in the latter. Distribution is considered to be generally similar,as well, with perhaps a greater concentration of the Kobs Triangularin the eastern range of the type.^MISCELLANEOUS UPPER LEVEL TYPESProvisional Type 1 (pi. 35, p) ; 4 specimens.ProportiOHs: Long and narrow.Blade: Edges convex.Shoulders: Prominent but obscured by flaring stem.8t?7n: Wide expanding stem formed by deep U-shaped lateral or diagonalnotches ; strongly convex base.Dimensions (only one whole specimen) : Length, 50 mm. (1.96 inches) ;width, 21 mm. (0.87 incla) ; thickness, 5 mm. (0.20 inch).Provenience : See table 3.Remarks: These are larger than usual for upper level projectile points.The workmanship is good. Distribution in time and space is unknown.Provisional Type 2 (pi. 35, m-o) ; 3 specimens.Proportions: Long and narrow.Blade: Slightly convex edges and rather thick in cross section.Shoulders: Tend to squareness but are not prominent.Stetn: Relatively short and squarish with parallel edges and bases somewhatconvex.Average dimensions: Length, 65 mm. (2.56 inches) ; width, 19 mm. (0.75inch) ; thickness, 8 mm. (0.31 inch).Provenience: See table 3.Retnarks: Like Provisional Type 1, these are longer than is common inupper level types. Workmanship is above average for this area.Temporal and areal distribution unknown.Provisional Type 3 (pi. 35, fc) ; 1 specimen.Proportions: Medium; blade broad, stem relatively long.Blade: Edges straight.Shoulders: Poorly defined.Stem: Concave edges; contracts gently from the shoulders to midstemfrom wliere it expands to a slightly concave base ; stem and basal edgesheavily ground.Dimensions: Length 28 mm. (1.09 inches) of which 12 mm. (0.46 inch) isblade length ; width, 15 mm. (0.59 inch) ; thickness obscured by thermalfracture on both faces of the blade.Provenience: See table 3. " Krieger, personal conference. fi204 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154Provisional Type 4 (pi. 35, ; 1 specimen. iProportions: Medium. \Blade: Edges parallel to midblade, asymmetrically convex to tip; edges'ground smooth along parallel portion. ' .Base: Concave. |Dimensions: Length, 39 mm. (1.56 inches) ; width, 19 mm. (0.75 inch) Ithickness, 7 mm. (0.28 inch). 'iProvenience: See table 3. |Remarks: The end of the blade has been roughly reflaked to its presentasymmetric shape. It seems likely that this is an earlier piece which!had been picked up and reused by the later occupants of the site.lSpatial and temporal distribution unknown. iMIDDLE AND LOWER LEVEL PROJECTILE POlNTgGary Stemmed (pi. 36, a-y) ; 518 specimens.Proportions: Commonly medium, but a few are of broad or narrow proportions. ^Blade: Edges vary from convex to somewhat concave and are frequently}asymmetrical. .Shoulders: Prominent, rarely barbed; an occasional point has only oneishoulder, ,'Stem: Usually contracts from the shoulder to form a tapered or rounded!stem, although a very few have parallel or even slightly expanding sides..Base is normally pointed, but in some cases is convex or straight. The 'majority of these convex or straight base points are the result of leaving Ithe unmodified original striking platform of the flake to form the butt : of the stem. Asphaltum was often used on the stem as an adhesive 5for binding it to the shaft.Average dimensions: Length, 47 mm. (1.87 inches); width, 26 mm. (1.03 1inches); thickness, 8 mm. (0.31 inch). There is a considerable size ;range in these points, depending somewhat upon material and work- i manship. lProvenience: See table 4. IRemarks: The Gary Stemmed point is the most characteristic type of the imiddle levels. Workmanship ranges from extremely crude percussion iflaked examples to well-made pressure retouched points. This depends [in part upon the material : Flint, quartzite, mineralized wood, and one 'point of fine-grained sandstone. Typologically similar projectile pointshave a wide range both in space and time. Without attempting to givea complete distribution, a few pertinent occurrences may be noted. Itis similar to, if not identical with, the simple haft 1 B of "the Tchefuncteperiod (Ford and Quimby, 1945, fig. 8, 7(, i), and most, if not all, ofthe points designated as types simple haft 1 A, 1 C, 1 D, 2 A, 2 B, and2 C of the Marksville period (Ford and Willey, 1940, figs. 45, 46). Pov-erty Point in northern Louisiana (Webb, C. H., 1948, fig. 44 1 2 and 4) ?sites Lu' 65, Lu" 67, Lu" 72, Lu^ 92, and Ct? 27 in the Pickwick Basin(Webb, Wm. S. and DeJarnette, 1942, pis. 207, 226, 227, 236, 257, and 1293) ; site 5 in the Norris Basin (Webb, Wm. S., 1938, pi. 31 6) ; IndianKnoll (Webb. Wm. S., 1946, fig. 32, c) ; Lake Tohopekaliga, Florida (Wat- Ison, 1948, fig. 23, g, h) ; and Stalling's Island (Claflin, 1931, pis. 55, 56)all yield points of a similar pattern. It is common in pottery sites innortheast Texas (Krieger, 1946, p. 117). Furthermore, it is listed asone of the most common forms of the Woodland Pattern of the north- Riv. Bas. Sur.Pap. No. 4] ADDICKS BASIN?WHEAT 205ern United States (Anonymous, 1943, p. 396 and fig. 39, I b). Thetemporal distribution within this geographical range is Archaic to TempleMound II, perhaps later. Within the sites of the Addicks Basin, theGary Stemmed point earlier displaces a group of expanding stemmedpoints in favor, and is finally, itself, largely replaced by the smallarrow points of the upper levels. The variation within the type rangeof the Addicks specimens seems to have little, if any, chronologicalsignificance per se. The importance of the group lies in the predominantrole it played during the middle level of occupation.Table 4. ? Provenience of Middle and Lower Level types of projectile points ' 206 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154Wells Contracting Stem (pi. 37, s) ; 1 specimen (broken).Proportions: Narrow.Blade: Less wide than stem, edges convex (?) ; distal end broken.Shoulders: None.Stem: Long, narrow, contracts from midpoint of blade; base straight; stemand basal edges gi-onnd smooth.Dimensions: Length, circa 70 mm. (?) (2.75 inches) ; width, 19 mm. (0.73inch) ; thickness, 7 nmi. (0.28 inch). Leiigtli of stem, 32 mm. (1.28inches).Provenience: See table 4.Remarks: Basically a narrow leaf-shaped point. The stem is demarkedby grinding and by a greater width than the blade, forming, in effect,a reverse shoulder. The known range of the Wells Contracting Stemform centers in northeast Texas and may be fairly late, but it is cer-tainly precontact.Provisional Type 6 (pi. 37, d, e) ; 25 specimens.Proportions: Medium to narrow.Blade: Edges convex to straight, rarely concave.Shoulders: Usually poorly defined, sometimes square.Stem: Slightly to widely expanding, formed by relatively broad and shallow J;notches diagonally flaked from the basal corners ; bases mostly convex, \occasionally straight. |Average dimensions: Length, 45 mm. (1.78 inches) ; width, 21 mm. (0.84?inch) ; thickness, 6 mm. (0.23 inch). |Provenience: See table 4. |Remarks: Although somewhat similar points occur in the Poverty Point*site in northern Louisiana (Webb, C. H., 1948, fig. 44, 4), in general*their temporal and geographical range is not known. ;Provisional Type 7 (pi. 37, f-h) ; 23 specimens.Proportions: Medium.Blade: Edges convex.Shoulders: Obscured by width of stem and base.Stem: Flaring, as wide as blade; lateral notches are usually broad andshallow, sometimes narrow ; bases generally straight.Average dimensions: Length, 39 mm. (1..56 inches) ; width, 20 mm. (0.78inch) ; thickness, 6 mm. (0.23 inch).Provenience: See table 4.Remarks: Rough but not poor workmanship and somewhat smaller thanthe majority of Middle and Lower Level points. The complete rangein space and time is not known, but the shape occurs at site Ct? 27 inthe Pickwick Basin (Webb. Wm. S., and De.Jarnette, 1942, pi. 294, 1, typelisted as No. 47) ; at site Ma? 4 in the Wheeler Basin (Webb, Wm. S.,1939, pi. 108, ft) ; in Missouri (Wedel, 1943, pi. 12, c^e) ; and is listedas one of the basic shapes of the Woodland Pattern (Anonymous, 1943,fig. 39, d).Provisional Type 8 (pi. 37, i, /) ; 13 specimens.Proportions: Narrow to medium.Blade: Forms a compound curve, concave from shoulders to midblade,convex to tip.Shoulders: Prominent but not barbed. pip-N^o^.^T"^' ADDICKS BASIN?WHEAT 207Stem: Somewhat expanding, frequently ground smooth ; base is straight toconvex.Average dimensions: Length, 60 mm. (2.37 inches) ; width, 25 mm. (0.96inch) ; thickness, 8 mm. (0.31 inch). , Provenience: See table 4.Remarks: These points are larger than average for the site, and the work-manship is good. The type is quite similar to the simple haft 1 A ofthe Tchefunete and Marksville cultures (Ford and Quimby, 1945, fig. 8,c, 6, c, and e) and to some of the points from site Ct^ 17 in the WheelerBasin (Webb, Wm. S., 1939, pi. 27, a). Its complete range is not known.Provisional Type 9 (pi. 37, A:, /) ; 8 specimens.Proportions: Broad.Blade: Edges convex.Shoulders: Slightly barbed, occasionally straight.Stem: Gently to widely expanding, formed by notches from the basal corners ; bases are convex.Average dimensions: Length, 35 mm. (1.37 inches) ; width, 24 mm, (0.93inch) ; thickness, 5 mm. (0.20 inch).Provenience : See table 4.Remarks: Complete areal and spatial distribution unknown, but at least oneother southeastern occurrence is at Hiwassie Island (Lewis andKneberg, 1940. pi. 66B, upper row).Provisional Type 10 (pi. 37, m, n) ; 9 specimens.Proportions : Narrow.Blade: Edges gently convex.Shoulders: Obscured by width of stem.Stem: Expanding, formed by broad, shallow, lateral notches ; base convex ; basal and notch edges frequently smoothed by grinding.Average dimensions: Length, 40 mm. (1.59 inches) ; width, 17 mm. (0.68inch) ; thickness, 6 mm. (0.23 inch).Provenience: See table 4.Provisional Type 11 (pi. 37, a-c) ; 20 specimens.Proportions : Medium to narrow.Blade: Edges predominantly straight but some show slight convexity orconcavity.Shoulders: Not prominent, obscured by wide expanding stem.Stem: Expanding to flaring; formed by broad, shallow, irregular, lateralnotches some distance above the base; base usually convex but some-times straight.Average dimensions: Length, 54 mm. (2.12 inches) ; width, 22 mm. (0.87inch) ; thickness, 8 mm. (0.31 inch).Provenience: See table 4.Remarks: Appears in site Mg^ 2 in the Wheeler Basin, an early horizon(Webb, Wm. S., 1939, pi. 83, a, row 3, No. 2), and also in Mississippi(Brown, 1926, fig. 28, second row, last two specimens), but withoutcontext. Its complete distribution in space and time is not known.Provisional Type 12 (pi. 37, p, q) ; 28 specimens.Proportions: Narrow to medium.Blade: Edges form compound curve, concave from shoulders to midblade,convex to tip.Shoulders: Prominent, rarely barbed.953842?53 15 208 BUREAU OF AMERICAN ETHNIOLOGY [Bull. 154Stem: Vary from slightly contracting to somewhat expanding, frequentlyof a bulbous shape and smoothed by abrasion ; stem constitutes fromone-third to one-fourth of the total length of the point ; base convexto rounding.Average dimensiojis: Length, 65 mm. (2.56 inches) ; width, 25 mm. (0.96inch) ; thickness, 8 mm. (0.31 inch).Provenience: See table 4.Remarks: In general appearance this type is quite similar to ProvisionalType 8, the chief difference being in the consistently longer, occasionallybulbous stem of the present type. Temporal and spatial distributionunknown.Provisional Type 13 (pi. 37, r) ; 6 specimens.Proportions: Medium.Blade: Edges convex to straight.Shoulders: Slightly barbed to poorly defined.Stem: Relatively long, parallel-sided ; bases indented ; basal and stem edgessometimes ground.Average dimensions: Length, 55 mm. (1.65 inches) ; width, 23 mm. (0.90inch) ; thickness, 7 mm. (0.28 inch).Provenience: See table 4.Remarks: Points of similar shape but generally larger and inferior inworkmanship occur over a wide range to the west of the Addicks Basin.They seem to center on the Edwards Plateau, occurring as far northas Abilene (Sayles, 1935, pi. 14, 6, c, and g), west to the Pecos River(Pearce and Jackson, 1933, pi. 12, a, 6), east to the Black and GrandPrairies near Dallas (Bryan, 1930, pi. 24, 4), and on the plateau proper(Jackson, 1938, pi. 4, 1; Pearce, 1932 a, pi. 10, No. 20; and Kelley, 1947,pi. 14,1).Provisional Type 14 (pi. 37, ; 8 specimens.Proportions: Narrow.Blade: Edges straight to slightly convex.Shoulders: Barbed.Stem: Short relative to total point length, gently expanding ; base straight.Average dimensions: Length, 45 mm. (1.78 inches) ; width, 22 mm. (0.87inch) ; thickness, 4.50 mm. (0.17 inch).Provenience : See table 4.Provisional Type 15 (pi. 38, a) ; specimen.Proportions: Narrow.Blade: Convex.Shoulders: Indeterminate because of broken condition of point.Stem: Basically parallel-sided, edges slightly concave; base somewhat con-cave ; stem and basal edges ground smooth.Dimensions: Length, circa 80 mm. (?) (3.14 inches) ; width, 21 mm. (0.84inch) ; thickness, 7 mm. (0.28 inch).Provenience: See table 4.Remarks: Specimen broken and small central segment was not recovered.Workmanship excellent, the flaking of the blade is basically collateralwith minute retouching along the blade edges. This point somewhatresembles the Scottsbluff points which Webb (Webb, C. H., 1948, fig. 45,Nos. 1-5) reports from northern Louisiana. If it actually falls intothat category, the Addicks Basin find marks the southernmost occur-rence yet noted. P^.N^o^-lr' ADDICKS BASIN?WHEAT 209Provisional Type 16 (pi. 38, 6) ; 2 specimens.Proportions: Medium.Blade: Basically leaf-shaped, edges convex from tip to intersection with stem.Shoulders: None.Stem: Parallel-sided or slightly contracting; base concave; basal and stemedges ground.Dimensions (larger specimen) : Length, 67 mm. (2.64 inches) ; width, 31 mm.(1.21 inches) ; thickness, 9 mm. (0.36 inch).Provenience: See table 4.Remarks: Workmanship excellent with broad, parallel flakes on the facesof the blade. Typologically similar points occur at Nebo Hill in west-ern Missouri (Shippee, 1048, pi. 2, Ah, Ch) where they are consideredto be reworked blades from the more typical elongate, lanceolate variety.Others somewhat resembling these occur at site Ft-42 on Lime Creekin Nebraska (Schultz and Frankforter, 1948, fig. 12, specimen 7555).Provisional Type 17 (pi. 38, c) ; 5 specimens.Proportions: Medium to narrow.Blade: Convex edges describe parabola from shoulder to tip.Shoulders: Prominent, square.Stem: Square, edges parallel ; base straight.Average dimensions: Length, 59 mm. (2.32 inches) ; width, 25 mm. (0.96inch) ; thickness, 8 mm. (0.31 inch).Provenience : See table 4.Remarks: Workmanship is excellent. Points of this type occur in siteLu? 86 in the Wheeler Basin (Webb, Wm. S., 1939, pi. 12, &), in IndianKnoll (Webb, Wm. S., 1946, fig. 32, a, rows 1 and 2), and the type islisted as one of the basic Woodland Pattern shapes (Anonymous, 1943,fig. 39, le).Provisional Type 18 (pi. 38, d) ; 15 specimens.Proportions: Medium.Blade: Edges convex.Shoulders: Prominent to barbed.Stem: Slightly expanding, formed by diagonal notches from the basalcorners; bases convex.Average dimensions: Length, 55 mm. (2.16 inches) ; width, 32 mm. (1.28inches) ; thickness, 6 mm. (0.23 inch).Provenience: See table 4.Provisional Type 19 (pi. 38, e, f) ; 18 specimens.Proportions: Medium.Blade: Edges straight, occasionally serrated.Shoulders: Prominent.Stem: Large relative to total point size; edges usually parallel; base slightlyconcave to convex, generally straight ; basal and stem edges may belightly ground.Average dimensions: Length, 55 mm. (2.16 inches) ; width, 29 mm. (1.14inches) ; thickness, 7 mm. (0.28 inch).Provenience: See table 4.Remarks: Points of this shape occur in Mississippi (Brown, 1926, fig. 28,row 4, second and third specimens), at site Lu'' 92 in the PickwickBasin (Webb, Wm. S. and DeJarnette, 1942, pi. 257, 1, second row 210 BUREAU OF AMERICAN ETHNOLOGY [BuU. 154 i second specimen), and are a component of the collections from theMorhiss site " in Victoria County, Tex., southwest of the Addicks Basin.Provisional Type 20 (pi. 38, g,h);G specimens.Porportions: Medium to narrow.Blade: Edges convex.Shoulders: Usually obscured by flaring stem.Stem: Flaring, sometimes as wide as blade, formed by lateral notches immedi-ately above the base ; base straight.Average dimensions: Length, 50 mm. (1.96 inches) ; width, 25 mm. (0.96inch) ; thickness, 6 mm. (0.23 inch).Provenience: See table 4.Remarks: Points of this characteristic shape occur in a considerable sizerange, the larger ones appearing in the Southeast, both as knives andas projectile points (Webb, Wm. S. and DeJarnette, 1942, pi. 236, middlerow, third specimen from right), in the Central Plains area (Wedel,1943, pi. 12, o-fir), and the type is one of the basic Woodland Patternshapes (Anonymous, 1943, fig. 39, la).Pedernales Indented Base (pi. 38, i) ; 2 specimens.Proportions: Medium to narrow.Blade: Edges convex to straight.Shoulders: Straight to slightly barbed.Stem: Edges parallel ; base concave to deeply indented.Dimensions: Length, probably 90 to 100 mm. (3.53 to 3.92 inches) (both speci-mens broken) ; width, 37 and 40 mm. (1.46 and 1.59 inches) ; thickness,7 and 9 mm. (0.28 and 0.36 inch).Provenience : See table 4.Remarks: This type of point is characteristic of sites along the middle courseof the Colorado River of central Texas."Provisional Type 21 (pi. 38, /) ; 2 specimens.Proportions: Narrow.Blade : Broad relative to total length ; edges convex.Shoulders: Well defined.Stem: Long relative to total length ; edges parallel ; base straight ; stem edgesground.Dimensions : Length, 35 and 45 mm. (1.37 and 1.78 inches) ; width, 24 mm.(0.93 inch) ; thickness, 5 mm. (0.20 inch).Provenience: See table 4.Remarks: The distribution of this type is poorly known. One occurrenceof a similar point is at site No. 5 in the Norris Basin in eastern Ten-nessee (Webb, Wm. S., 1938, pi. 31, third row, second from left).Provisional Type 22 (pi. 38, n) ; 12 specimens.Proportions: Narrow to medium.Blade: Edges convex.Shoulders: Well defined.Stem: Slightly expanding ; base convex to straight.Average dimensions : Length, 36 mm. (1.43 inches) ; width, 16 mm. (0.63inch) ; thickness, 5 mm. (0.20 inch).Provenience: See table 4. " Collections in Museum of Anthropology, University of Texas, Austin, Tex." Kelley, personal conference. pS.N^o.'lr" ADDICKS BASIN?WHEAT 211Provisional Type 23 (pi. 38, k) ; 21 specimens.Proportions: Narrow.Blade: Edges strongly convex.Base: Convex to round.Average dimensions: Length, 55 mm. (12.16 inches) ; width, 20 mm. (0.78inch) ; thickness, 9 mm. (0.36 inch).Provenience: See table 4.Remarks: The leaf-shaped blade is one of the basic types of the WoodlandPattern (Anonymous, 1943, fig. 39, If). It is likely that these specimensare too small to have been used for anything but projectile points. Muchlarger pieces of the same shape also occur and are discussed below underthe heading Knives.Provisional Type 24 (pi. 38, m) ; 3 specimens (broken).Proportions: Medium to broad.Blade: Convex edges converge strongly toward tip.Base: Concave with several parallel longitudinal basal thinning flakes oneither face of the blade ; very small lateral notches immediately abovethe base give the effect of ears ; basal and notch edges ground smooth.Dimensions: Length, probably near 35 mm. (1.37 inches) ; width, 29 mm.(1.14 inches) ; thickness, 5 mm. (0.20 inch).Provenience: See table 4.Clevis Fluted (?) (pi. 38, ; 1 specimen (broken).Proportions : Indeterminate, probably narrow.Blade: Edges convex, expanding from base toward midblade; lateral edgesground ; fluted on both faces.Base : Concave, basal edge ground.Dimensions: Indeterminable ; present length, 40 mm. (1.59 inches).Provenience: See table 4.Remarks : Two fragments of this point were recovered, one from the streambed, the other from the Doering midden, and fitted together. It isinteresting to note that the piece from the stream bed is patinated atannish cream color while the fragment from the midden is colored avery light cream. This is an excellent demonstration of the effect ofimmediate environment on patination. The southern distribution of thefluted forms is poorly known, very few having been found in context.They occur in the Abilene region to the west (Bryan, Kirk and Ray,C. N., 1938) and in northern Louisiana (Webb, C. H., 1948, fig. 45, 6-8)to the north. While this occurrence may argue for a relatively recentsurvival of the type, it is the opinion of the writer that it representsa fossil point picked up and reused by the later occupants.UNPLACED PKOJECTILE POINT TYPESAll of the following projectile points, with the exception of No. c, Plate 39,from site 42/66A6-5, are from the stream bed at the Doering site. Since theywere not duplicated in the excavated series, their relative stratigraphic positionmay only be inferred at the present time.Provisional Type 25 (pi. 39, 6) ; 1 specimen.Proportions: Narrow.Blade: Edges convex.Shoulders: Well defined.Stem : Relatively long, contracting from shoulders ; base straight.Dimensions: Length, 62 mm. (2.45 inches) ; width, 28 mm. (1.09 inches) ;thickness, 11 mm. (0.43 inch). 212 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154Provisional Type 26 (pi. 39, c) ; 1 specimen.Proportions: Narrow.Blade: Concave to midblade and convex to tip (broken) ; beveled on oppositefaces.Shoulders: Poorly defined.Stem: Contracts from shoulders and then expands slightly to base; baseconcave ; stem and basal edges heavily ground.Dimensions: Length, indeterminate; width, 19 mm. (0.75 inch) ; thickness,7 mm. (0.28 inch).Remarks: This form resembles, superficially at least, the beveled points ofthe Brazos River Phase of Sayles (19.3.5, pi. 11, a-e).Provisional Type 27 (pi. 39, h); 2 specimens (broken).Proportions: Medium.Blade: Edges straight.Shoulders: Prominent, square.Stem: Large, contracting, with straight edges; base convex; basal and stemedges ground.Dimensions: Length, about 70 mm. (2.75 inches) ; width, 37 mm. (1.46inches) ; thickness, 8 mm. (0.31 inch).Remarks: Workmanship excellent. These points occur in the Morhiss sitein Victoria County, Tex."Provisional Type 28 (pi. 39, g) ; 1 specimen (broken).Proportions: Narrow.Blade: Edges convex.Shoulders: Well defined, square.Stem: Lai-ge, contracting, with convex edges ; base probably round or pointed ;stem edges ground.Dimensions: Length, about 95 mm. (?) (3.73 inches); width, 30 mm.(1.18 inches) ; thickness, 7 mm. (0.28 inch).Remarks: This form also occurs as a component of the Morhiss site lithiccomplex. Workmanship excellent.Provisional Type 29 (pi. 39, i, /) ; 4 specimens.Proportions: Medium over-all with broad blade.Blade: Edges straight to convex.Shoulders: Medium.Stem: Long and wide, edges parallel ; base convex.Dimensions: Length, 38 mm. (1.50 inches) ; width, 21 mm. (0.84 inch) ;thickness, 6 mm. (0.23 inch). Stems average 21 mm. (0.84 inch) inlength.Remarks: Because of the relative size of the stems, there is a possibilitythat these are reworked from larger points, although only one seemsactually to have been rechipped. On the whole, it seems likely thatthey constitute a type. A single example of the type is illustrated byMoorehead (1931, fig. 2, lower row, third specimen) from Yell County,Ark., and another similar point from site JO 2, an Adena site in easternKentucky (Webb, Wm. S., 1942, pp. 310, 315, and fig. 7, 3K), but thecomplete range in time and space is unknown. " Collections, Museum of Anthropology, University of Texas, Austin, Tex. pip.Na'lr* ADDICKS BASIN?WHEAT 213Provisional Type 30 (pi. 39, k,l);6 specimens.Proportions: Medium.Blade: Edges straight to somewhat convex.Shoulders: Medium.Stem: Edges parallel to slightly expanding; bases straight and tend toform a square with the stem.Average dimensions: Length, 38 mm. (1.50 inches) ; width, 24 mm. (0.93inch) ; thickness, 7 mm. (0.28 inch).Provisional Type 31 (pi. 39, d) ; 1 specimen (broken).Proportions: Medium.Blade: Asymmetric, edges convex.Shoulders: One side has a small barb, the other a broad, long, curved barb.Stem: Shape unknown, probably asymmetric in direction opposite to largebarb.Dimensions: Length, about 43 mm. (?) (1.71 inches) ; width, 33 mm. (1.31inches) ; thickness, 4 mm. (0.15 inch).Provisional Type 32 (pi. 39, e) ; 1 specimen (broken).Proportions: Broad.Blade: Edges convex.Shoulders: Prominent.Stem: Short, broad, with slightly expanding edges ; base convex ; stem andbasal edges ground.Dimensions: Length, indeterminate; width, 28 mm. (1.09 inches) ; thickness,5 mm. (0.20 inch).Copena Point (pi. 39, a) ; 1 specimen.Proportions: Narrow.Blade: Edges form compound curve, concave just above the base, convex atmidblade, and then concave to the tip.Base: Slightly convex.Dimensions : Length, 76 mm. (2.98 inches) ; width, 21 mm. (0.84 inch) ; thick-ness, 9 mm. (0.36 inch).Remarks: This point corresponds to the less finished type as reported byWebb and DeJarnette (1942, pi. 207, 2) for the Pickwick Basin.Plainview Point (pi. 39, ; 1 specimen.Proportions: Narrow.Blade: Parallel edges, asymmetrically convex to tip; end of blade resharp-ened by much inferior technique ; flaking basically collateral withparallel flake scars meeting along a slight dorsal ridge ; lateral edgesground.Base: Slightly concave ; edge ground.Dimensions: Length, 39 mm. (1.56 inches) ; width, 17 mm. (0.68 inch) atmid-blade ; thickness, 4 mm. (0.15 inch) . Remarks: Although its dimensions are somewhat less than those of pointsfrom the type station ( Sellards, Evans, Meade, and Krieger, 1947, p. 939and table 4, pp. 943, 944), this specimen probably should be considered aPlainview point.^* " Krieger, personal conference. 214 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154Baird Beveled Point (pi. 39, m, ?) ; 2 specimens.Proportions : Medium.Blade: Triangular, edges straight; longitudinal thinning flakes on bothfaces of one, on one face of the other.Base: One concave, one convex.Dimensions: Length, 41 and 44 mm. (1.61 and 1.75 inches) ; width, 26 and27 mm. (1.03 and 1.06 inches) ; thickness, 6 mm. (0.23 inch).Remarks: Points of this form occur both in central Texas (Ray, C. N., 1938,pi. 25, 1, row 5, Clear Fork Dart 3; and Kelley, 1947 b, pi. X, 6) andin the South Gulf Coastal area (Sayles, 1935, pi. 25, h). These pointsmay, therefore, represent trade relations with either area, although thecentral Texas area seems most likely since several other central Texastypes also occur as trade specimens.PROJECTILE POINT STRATIGRAPHYAn examination of table 5 and figure 22 reveals several trendssignificant in the establishm;ent of a relative chronology for theAddicks Dam Basin sites. Considering the Doering site alone forthe moment, these are : (1) The greater variety of miscellaneous pointtypes from the lower levels in relation to the Gary Stemmed group(except level 105-120 where the number of points recovered is toosmall to be statistically important) ; (2) the increasing dominanceof the Gary type to level 30-45, in which it constitutes nearly 85percent of all forms present; (3) the increasing importance of thePerdiz Pointed Stem and its associated forms in the two upper levelsof the site.As already mentioned above, the data of the lowest level are in-secure, since only four points were found. It is perhaps significantthat only one of these, or 25 percent of the total for that level, wasof the Gary Stemmed type ; but the smallness of the sample rendersunsound any definite conclusions based on the occurrence of this type.From level 90-105 up, however, the evidence is clear that the mis-cellaneous group of points, nearly all of which are expanding orparallel-sided stemmed varieties, give way, as a group, as the GaryStemmed type gains in favor and importance.The appearance of the Perdiz Pointed Stem, Alba Barbed, andassociated types marks the initiation of the rapid encroachment ofthe small projectile points. This event probably heralds the intro-duction of the bow and arrow. The occurrence of a single PerdizPointed Stem point in each of levels 60-75, 45-60, and 30-45 posesan interesting problem. It may be that the Doering site peoples,through trade, warfare, or otherwise, had come into contact withpeoples already using the type some time prior to its main introduc-tion. It is likewise possible that, although introduced earlier, it didnot gain a strong foothold until relatively late in the occupancy ofthe site. The most probable explanation, however, at least for thetwo lower occurrences, is that they represent dislocations due to dis- Riv. Bas. Sur.Pap. No. 4] ADDICKS BASIN?WHEAT 'Zli)Scale of Percentage25 50 75100 75 50 25 1000-.5 kV\\V\^SX\\\\\XNX\\\\N\\\\\\)jJ ,5-30 K\^^^^^^^v^^^^^^^^^^^^^^^l 130-45 K\\\VVV\VV\\^5^^^^^^45-60 kvvwvv^vww-^^^^60-75 K\\\V\M75-90 I90-105 [105-120 I KOBS SITE 0-15 KWVVVVVWWV15-30 l^^^333^ 90-105 I105-120 [ \^i:immm^^^^i^^f^i^D0ERIN6 SITEFigure 22.-Graph showing comparative projectile point stratigraphy of Doering andKobs sites. Diagonal hatching, upper level types; blank, Gary Stemmed type; dottedarea, other lower and middle level types. 216 BUREAU OF AMERICAN ETHNiOLOGY [Bull. 154 i^'^<^ t^ OtiO^ O ?^ O '^ ^ OINmiM> P^ >. >i-f:^ >.>>t>i(>>t>>>>>.t>>;*> f>> Co, f>> p^ p>> p>,T3 >> >-, b, >>^ o ^ K O O O O _0 _0 O O O _0 O 3 O O O O O O 03 O O O O pLH<^C)j2000000000000000'OOOOOj^M(X|(l|fl,pHOPH(X|fl|PHPHPHPlHeL|PMfL|FmlHAHPHfL|fL|fl. 1500, cannot beanswered by physical stratigraphy; but with the beginning of thedefinite midden deposit, the occupation was by upper-level peoples, asattested to by the lithic complex. Both the Grisbee site and site42/66A6-4 appear to have been occupied chiefly during upper-leveltime, but little may be said of them because their excavation waslimited to testing operations.A comparison may now be made between the prehistoric occupantsof the Addicks Basin sites and the historic inhabitants to determinethe nature and extent of their relationship. Swanton (1928, pp. 712-713) characterizes the Attakapa as a loosely organized group of lowculture who practiced cannibalism and who were definitely marginalto the Southeast proper. This would presumably hold, also, for theAkokisa, the westernmost group of the Attakapa. The Akokisa fol-lowed a simple annual economic cycle, passing the winter in perma-nent or semipermanent camps, and the summer in wandering to theseacoast and elsewhere. They followed a hunting and gathering 246 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154 subsistence pattern, including the collecting of shellfish. They hadpottery, but whether they made it or received it in trade is not clear.Swanton (1946, p. 737) states that they received most of it from theKarankawa, the Avoyel, and probably the Caddo. Trade was alsocarried on with the peoples of the interior.In general, the pattern described for the Akokisa might fit theprehistoric pattern as well. However, there are certain discrepancieswhich may be examined briefly. There is obviously no means ofjudging whether social or tribal organization was loose, and the deter-mination of a low or high culture is a relative one. That cannibalismwas widely practiced by the prehistoric inhabitants may be questionedso far as the evidence in the excavated sites is concerned. Further-more, the pottery of the prehistoric sites is much too uniform to havebeen imported from a variety of sources. It does not conform to thepottery known from the Karankawa nor from the Caddo. On theother hand, it seems likely that the economic cycle described for thehistoric peoples of the area was much the same as that of the pre-historic. Whatever form of shelter was made and used by the precon-tact inhabitants must have been perishable, and this may be generallyequated with the house types of the historic peoples. Further, thereis archeological evidence to support the story of trade with the peoplesof the interior. But the majority of the cultural factors which agreeare factors which might be attributed to an environmental basis.Therefore, it is impossible at present to derive a satisfactory answer.One solution to the problem lies in the excavation of known Akokisasites such as the site of the Mission Nuestra Seiiora de la Luz.It is likewise impossible to fit the Addicks Basin sites into the Attak-pan Phase of Sayles (Sayles, 1935, p. 41), inasmuch as there areinsufficient details given for a comparison. While the pottery appearsbasically of the same pattern, the statement that the sites usually showsigns of European contact removes them from direct association withthe sites covered by the River Basin Surveys. It is probable that thesesites represent, as Sayles points out, the early historic occupants?theAttakapans.The sites of tlie Addicks Basin, then, represent a people South-eastern in origin and basic affiliations, who, marginal to the mainstream of cultural achievement, seem to have developed slowly and inrelative isolation. This is attested to by the essentially homogeneousbut static culture complex. In late prehistoric times they came intothe sphere of influence of the peoples to the west and northwest. Dur-ing this time the bow and arrow was probably introduced, and cer-tainly new and distinctive forms of projectile points and knives.Nevertheless, the culture pattern remained Southeastern in characteruntil the final abandonment, perhaps around the year A. D. 1600. Riv. Bag. Siir.Pap. No. 4] ADDICKS BASIN?WHEATAPPENDIXTable 9.? Trait list for Addicks Basin sites i 247 Trait list LevelsUpper Middle LowerGeneral traits :Sites alongside stream on natural knoll.Midden depositBurial complex;FlexedSemiflexedOn backOn left side -On right side.. _Orientation eastOrientation westOrientation southeastOrientation southwestOrientation north-northeastOval pitWith burial furnitureMaterial culture :Pottery complex :Goose Creek PlainGoose Creek IncisedElongate jar, wide mouthWide mouth bowl, simple contourCalcium phosphate temperSherd temperSand temperDirect rimSharp lip..Rounded HpFlattened lip... -FigurinesWorked sherdsLithic complex :Projectile points :Perdlz Pointed StemScalhom StemmedEddy StemmedAlba Barbed -Kobs Triangular __Provisional Type 1Provisional Type 2Provisional Type 3Provisional Type 4Gary StemmedWells Contracting StemProvisional Type 5Provisional Type 6Provisional Type 7Provisional Type 8Provisional Type 9Provisional Type 10Provisional Type 11Provisional Type 12Provisional Type 13Provisional Type 14Provisional Type 15 _.Provisional Type 16Provisional Type 17Provisional Type 18Provisional Type 19Provisional Type 20Pedemales Indented BaseProvisional Type 21 _.Provisional Type 22Provisional Type 23..Provisional Type 24Clovis Fluted...Provisional Type 25Provisional Type 26Provisional Type 27Provisional Type 28Provisional Type 29Provisional Type 30Provisional Type 31Provisional Type 32Copena Point..Plainview PointBaird Beveled Point (2)? See footnotes at end of table. 248 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154Table 9. ? Trait list for Addicks Basin sites i?Continued Trait list LevelsUpper Middle Lower Material culture?ContinuedLithic complex?ContinuedMiscellaneous lithic artifacts:Knives:Type laType lb-Type IcType IdTvpe2TypesType 4Type 5Type 6aType 6bScrapers :Stemmed endLarge endSmall endLarge sideSmall sideFlakeChoppers :One end modifiedAll sides modliiedFlaked celtDrills :Type laType lbType IcType 2aTvpe2bTypesGraversEccentric flintGround stone :Atlatl Weights (banner stones).Sandstone saws and abraders. _ . Milling stone.HandstoneHammerstonesPaint pigmentsBone :Cut bone awlsUlna spatulaAntler :Projectile pointsBeadWorked tinesShell :Pendants ? pip No^'l]'"'^' ADDICKS BASIN?WHEAT 249Brew, J. O., and others.1947. Symposium on river valley archaeology, Amer. Antiq., vol. 12, No. 4.Brown, Calvin S.1926. Archeology of Mississippi. Mississippi Geol. Siirv., Uuiv. Mississippi.Bryan, E^ank.1930. Archeological remains in the Black and Grand Prairies of Texas,Texas Arch, and Paleont, Soc. Bull., vol, 2. Abilene,Bryan, Kirk, and Ray, Cyrus N,1938. Long channeled point found in alluvium beside bones of Elephascolumn. Texas Arch, and Paleont, Soc. Bull., vol. 10. Abilene.Oampreix, Thomas N.n. d. The Galveston Focus, (Ms,)Castaneda, Carlos E.1936, Our Catholic heritage in Texas, 1519-1936, The Mission Bra, 1519-1693, vol, 1, Austin.Ohamberlin, Thomas C, and Salisbury, Rollin D,1930. College textbook of geology. New York. Rewritten and revised byRollin T. Chamberlin and Paul MacClintock.Claflin, William H., Jr.1931, The Stalling's Island Mound, Columbia County, Georgia. Pap. Pea-body Mus. Amer, Arch, and Ethnol., Harvard Univ., vol. 14, No. 1,Cambridge.Collier, Donaij). See Martin, Paul S. ; Quimby, George I.; and Collier,Donald,DeiJarnette, David L. See Webb, William S. and DeJarnette, David L.Deuel, Thorne,1935. Basic cultures of the Mississippi Valley, Amer. Anthrop., vol. 37,No. 3,Dice, Lee R.1943. The biotic provinces of North America. Ann Arbor,Evans, Glen L. See Sellards, E. H. ; Evans, Glen L. ; and Meade Grayson E.Fairbanks, Charles H,1942. The taxonomic position of Stalling's Island, Georgia, Amer, Antiq.,vol. 7, No. 3,Fenneman, N, M.1906. Oil fields of the Texas-Louisiana Gulf Coastal Plain, U. S, Geol,Surv, Bull, 282,1938. Physiography of the Eastern United States. New York.FoLMER, Henri.1940. De Bellisle on the Texas Coast. Southwestern Hist, Quart,, vol, 44,No, 2. October.Ford, J. A,, and Quimby, George I., Jr.1945, The Tchefuncte Culture, an early occupation of the Lower MississippiValley. Mem. No. 2, Soc, Amer. Arch.Ford, J. A., and Willey, Gordon R.1940, Crooks site, a Marksville Period burial mound in LaSalle Parish, Loui-siana, Anthrop. Study No. 3, Dept. Conservation, Louisiana Geol,Surv, New Orleans,Fkankforter, W, D, See Schultz, C. Beetrand, and Fbankfoeter, W, D.Gatschet, Albert S,1891. The Karankawa Indians, the coast people of Texas, Pap. PoabodyMus. Amer. Arch, and Ethnol., Harvard Univ., vol. 1, No, 2. Cam-bridge. !250 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154 JHodge, Frederick W., Editor. |1907. Cabeza de Vaca (Alvar N'uiiez), the narrative of Alvar Nuuez Cabeza jde Vaca. Orig. Narratives Early Amer. Hist. New Yorls. jJackson, A. T. I1935. Ornaments of East Texas Indians. Texas Arch, and Paleont. Soc. ?Bull., vol. 7. Abilene.1938. The Fall Creek sites. Ann. Rep. WPA and Univ. Texas Arch. Res., ;Lake Buchanan, 1936-1937. Univ. Texas Publ. No. 3802, Authrop. jPap., vol. 3, No. 1. Austin.See also Pearce, J. E., and Jackson, A. T. jJennings, Jesse D. See Setzler, Frank M., and Jennings, Jesse D. IKellet, j. Charles. ;1947 a. The Lehmann rock shelter : A stratified site of the Toyah, Uvalde,and Round Rock Foci. Texas Arch, and Paleont. Soc. Bull., vol. 18. '?Lubbock. I1947. b. The cultural affiliations and chronological position of the Clear jFork Focus. Amer. Antiq. vol. 13, No. 2. |Kidder, Alfred Vincent. |1932. The artifacts of Pecos. Pap. Southwestern Exped., No. 6, Phillips |Acad., Yale Univ. Press. New Haven, JKneberg, Madeline. See Lewis, Thomas M. N., and Knebebg, Madeline. iKeieger, Alex D. 11944. The typological concept. Amer. Antiq., vol. 9, No. 3. '1946. Culture complexes and chronology in northern Texas. Univ. TexasPubl. No. 4640. Austin. 11947. See Sellards, E. H. ; Evans, Glen L. ; and Meade, Grayson E. 1Lewis, Thomas M. N., and Kneberg, Madeline. |1946. Hiwassee Island. Univ. Tennessee Press. Knoxville. j1947. The archaic horizon of western Tennessee. Univ. Tennessee Record,Extension Ser., vol. 23, No. 4. Knoxville.LiBBY, W. F. See Arnold, J. R., and Lieby, W. F.Maerz, a., and Paul, M. Rea.1930. A dictionary of color. New York.Martin, George Castor.1929. Notes on some Texas coast campsites and other remains. Texas Arch,and Paleont. Soc. Bull., vol. 1. Abilene.1931. Texas coastal pottery. Texas Arch, and Paleont. Soc. Bull., vol. 3.Abilene.Martin, Paul S. ; Qthmby, George I., Jr. ; and Collier, Donald.1947. Indians before Columbus. Univ. Chicago Press. Chicago.Meade, Grayson E. See Sellards, E. H. ; Evans, Glen L. and Meade, Gray-son E.Moore, Clarence B.1902. Certain aboriginal remains of the northwest Florida coast. Pt. 2.Reprint from Journ. Acad. Nat. Sci. Philadelphia, vol. 12. Phila-delphia.Moorehead, Warren King.1981. Archaeology of the Arkansas River Valley. Phillips Acad., Yale Univ.Press. New Haven.MoRFi, Fray Juan Agustine.1935. History of Texas, 1673-1779. Trans, and annotated by Carlos E.Castaneda. Albuquerque.Paul, M. Rea. See Maerz, A. pip.?ra/]'""' ADDICKS BASIN?WHEAT 251Peaece, J. E.1932 a. The present status of Texas archeology. Texas Arch, and Paleont.Soc. Bull., vol. 4. Abilene.1932 b. The archeology of east Texas. Amer. Anthrop., vol. 34, No. 4.Peaece, J. E., and Jackson, A. T.1933. A prehistoric rock shelter in Val Verde County, Texas. Univ. TexasBull. No. 3327, Anthrop. Pap., vol. 1, No. 3. Austin.PiCHAEDO, JOSE AnTONIO.1931. Treatise on the limits of Louisiana and Texas. Edited by Charles W.Hackett. Vol. 1. Austin.Plummer, F. B.1932. Cenozoic systems in Texas. In The Geology of Texas, vol. 1, Stratig-raphy. Univ. Texas Bull. No. 3232. Austin.PoTEET, Sybil.1938. The occurrence and distribution of beveled knives. Texas Arch, andPaleont. Soc. Bull., vol. 10. Abilene.PoTTEE, Wendell H.1930. Ornamentation on the pottery of the Texas coastal tribes. TexasArch, and Paleont. Soc. Bull., vol. 2. Abilene.QuiMBY, George I., Jr. See Fokd, J. A., and Quimby, George I., Jr. ; see also Mae-tin, Paul S. ; Qtjimby, Geoege I., Jr. ; and Coluer, Donald.Ray, Cyrus N.1938. The Clear Fork Culture Complex. Texas Arch, and Paleont. Soc.Bull., vol. 10. Abilene.See also Bryan, Kirk, and Ray, Cyrus N.Salisbury, Rollin D. See Chamberlain, Thomas C, and Salisbury, Rollin D.Sayles, E. B.1935. An archaeological survey of Texas. Medallion Pap., No. 17. Globe,Ariz.ScHULTz, C. Bertrand, and Frankforter, W. D.1948. Preliminary report on the Lime Creek sites: New evidence of earlyman in southwestern Nebraska. Bull. Univ. Nebraska State Mus.Sears, William H.1948. What is the Archaic? Amer. Antiq., vol. 14, No. 2. Menasha, Wis.Sellards, E. H. ; Evans, Glen L. ; and Meade, Grayson E.1947. Fossil bison and associated artifacts from Plainview, Texas (withdescription of artifacts by Alex D. Krieger). Bull. Geol. Soc. Amer.,vol. 58.Setzler, Frank M., and Jennings, Jesse D.1941. Peachtree Mound and village site, Cherokee County, North Carolina.Bur. Amer. Ethnol. Bull. 131.Shantz, H. L., and Zon, Raphael.1924. Natural vegetation, Section E, the physical basis of agriculture, pt. 1.Atlas Amer. Agric. Washington.Shippee, J. M.1948. Nebo Hill, a lithic complex in western Missouri. Amer. Antiq., vol.14, No. 1.Sutherland, W. J.1908. Physiography of the Gulf Coastal Plains. Journ. Geogr., vol. 6, No.11. New York.Swanton, John R.1911. Indian tribes of the Lower Mississippi Valley and adjacent coast ofthe Gulf of Mexico. Bur. Amer. Ethnol. Bull. 43. 252 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154SwANTON, John R.?Continued1928. Aboriginal culture of the Southeast. 42d Ann. Rep. Bur. Amer. Ethnol.,1924-25, pp. 673-726.1946. The Indians of the southeastern United States. Bur. Amer. Ethnol.Bull. 137.Watson, Virginia.1948. Projectile points from the Kissimmee region. Amer. Antiq., vol. 14,No. 2.Webb, C. H.1948. Evidences of Pre-Pottery cultures in Louisiana. Amer. Antiq., vol. 13,No. 3.Webb, William S,1938. An archaeological survey of the Norris Basin in eastern Tennessee.Bur. Amer. Ethnol. Bull. 118.1939. An archaeological survey of Wheeler Basin on the Tennessee Riverin northern Alabama. Bur. Amer. Ethnol. Bull. 122.1942. The C. and O. Mounds at Paintsville. Univ. Kentucky Rep. Anthrop.and Arch., vol. 5, No. 4.1946. Indian Knoll. Univ. Kentucky Rep. Anthrop. and Arch., vol. 4, No. 3,pt. 1. Lexington.Webb, William S., and DeJarnette, David L.1942. An archeological survey of Pickwick Basin in the adjacent portions ofthe States of Alabama, Mississippi, and Tennessee, Bur. Amer,Ethnol, Bull. 129,Wedel, Waldo R.1943. Archeological investigations in Platte and Clay Counties, Missouri.U. S. Nat. Mus., Bull. 183.Willey, Goedon R. See Ford, J. A., and Willey, Gordon R.ZoN, Raphael. See Shantz, H. L., and Zon, Raphael. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 29 Doering site and profile, a. View from the west bank of South Mayde Creek.at 1 East. b, Profile BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 30 Kobs site, a, View before excavation, b, View of excavation showing Burial 6 in foreground. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 31 Pottery: Restored vessels, Goose Creek ware. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 32 Pottery: Decorated sherds, Goose Creek Incised. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 33 Pottery: Decorated sherds, a-j. Goose Creek Incised, k, Tchefuncte Stamped./, m. Unidentified ware. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 34 1^ ^ c 64 ^ 4K ?w Projectile points, a-n, Perdiz Pointed Stem, o-j, Scaihorn Stemmed, t-x, EddyStemmed, j-, Length, 45 mm. (1.78 in.). BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 35 Projectile points, a-h. Alba Barbed, i, ;, Kobs Triangular. /, p. Provisional types.m-o, 1. k, 3. /, 4. Length of n, 70 mm. (2.76 in.). BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 35 Projectile points: Gary Stemmed type. Lengtli of i, 56 mm. (2.20 in.). BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 37 e Projectile points: Provisional Types, o, 5. d, e, 6. /-A, 7. i, j, 8. k, I, 9. m, n, 10.a-c, 11. p, q, 12. r, 13. /, 14. j, Wells Contracting Stem. Length of p, 11 mm.(2.83 in.). BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 38 Projectile points: Provisional Types, a, IS. b, 16. c, 17. d, 18. e, /, 19. g, h, 20.j, 21. n, 22. k, 23. m, 24. j, Pedernales Indented Base. /, Clovis Fluted '(?).Length of b, 67 mm. (2.64 in.). BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 39 Projectile points, unplaced chronologically: a, Copena. /, Plainview. m, n, Baird Bev-eled. Provisional Types: b,25. c,'26. d, 31. ^,32. A, 27. g, 28. i, j, 29. k, I,30. Length of a, 76 mm. (3.01 in.). BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 40 Knives, a, Type la. b. Type lb. c. Type Ic. d, e. Type Id. /, g. Type 2. h, i. Type3. Length of a, 109 mm. (4.25 in.). BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 41 Knives, a, b, Tvpe 4. c, Type 6a. d, Type 6b. e, Tvpe 5. Length of e, 108 mm.(4.22 in.). BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 42 Scrapers, a. End scrapers, large. /;, r, End scrapers, stemmed. (/, e. End scrapers, smalj-i. Side scrapers, small. Length of/, 76 mm. (3.01 in.). BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 43 Choppers, a, b. One end unmodified, r, e. Both ends modified, d, Celt. Length of d,137 mm. (5.37 in.). BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 44 I e 9 ^B aayj^ m Drills, gravers, and eccentric flint. Drill types: a, b, la. c, d, lb. e, /, Ic. _g, h, 2b. z, 3. ;', 2a. A^-o, Gravers, p, Eccentric flint. Length of i, 57 mm. (2.26 in.). BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 45 Ground stone, a, Ailatl weight, b-e, Sandstone abraders and saws. /, Handstone.g, Milling stone. Greatest diameter of/, 106 mm. (4.18 in.). BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 46 Miscellaneous minerals and paint pigments.^ Length'of d, 63 mm. (2.48 in. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 47 IPe Artifacts of bone, antler, and shell, a, Ulna spatula, b-e, cut bone awls. /, g. Workedantler tines. A, i. Antler projectile points. /, Bead (.?). k, I, Shell pendants. Lengthof a, 71 mm. (2.79 in.). SMITHSONIAN INSTITUTIONBureau of American EthnologyBulletin 154 River Basin Surveys Papers, No. 4The Addicks Dam SiteII. Indian Skeletal Remains from the Doering andKobs Sites, Addicks Reservoir, TexasBy MARSHALL T. NEWMAN 253 953842?53 18 CONTENTS PAGEIntroduction 527Description 258Analysis 264Literature cited 266FIGURES24. Stereograph contour of the skull from Burial 2 25825. Stereograph contour of the skull from Burial 4 26026. Stereograph contour of the skull from Burial 5 26227. Stereograph contour of the skull from Burial 6 26228. Stereograph contour of the skull from Burial 7 263255 INDIAN SKELETAL REMAINS FROM THE DOER-ING AND KOBS SITES, ADDICKS RESERVOIR,TEXAS By Marshalx, T. NewmanINTRODUCTIONThe purpose of this report is to describe and analyze the skeletalremains from burials 1, 2, 3, and 4 (Doering site) and 5, 6, 7 (Kobssite). Due to their fragmentary nature, few measurements could betaken, and much of the description is in terms of morphological obser-vations. Unfortunately, these observations are more subjectivelydetermined than are the measurements, and hence are less useful incomparisons with the work of others. For this reason, most of thecomparative analysis has to be based upon a few actual or estimatedmeasurements and upon the stereographic drawings of the skullvaults (figs. 24-48). These drawings are approximately y^ life size.The comparative data from the Texas coast has been taken froman unpublished State-wide study by Dr. Marcus S. Goldstein. Inaddition to permitting the writer to use his study, Dr. Goldstein wasalso kind enough to examine the Doering-Kobs material. His com-ments, based upon first-hand experience with a great deal of Texasskeletal material, are most valuable and have been incorporated in thepresent report.Both Dr. Goldstein and the writer are reluctant to make any sweep-ing interpretations of the Doering-Kobs material. There simply isnot enough of it, and even after careful restoration it is still frag-mentary. But it possesses definite value since it is archeologicallydocumented material.Wliile burial 1 has been tentatively assigned to the middle phaseof occupation at the Doering site, the other burials (2-7) were prob-ably made in intrusive pits from the thin midden surface of both sites,most likely during the late phases of occupation (see pp. 238-239).The absence of European trade materials suggests the sites werenot occupied in historic times. From an archeological standpoint,then, the remains of burials 2-7 can be considered as one series. In a257 258 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154 4Figure 24.?Stereograph contour of the skull from Burial 2. (Approximately }i.)properly tentative way, Wheat has linked the Doering and Kobs siteswith the Akokisa band of Atacapan-speakers, who occupied the areaat the time of the earliest French and Spanish explorations (seepp. 245-246) . Thus, for present purposes, the series from burial 2-7is probably late or protohistoric, and possibly Atacapan.The measurements and observations were made according to tech-niques previously discussed (Newman, 1947, appendix A, B). Thecontour drawings were made with the Schwarz stereograph. Nophotographs of skulls are included because of their fragmentarynature. DESCRIPTIONBurial 1 {Doering site).?Unrestorable skull fragments, almostintact lower jaw, incomplete long bones, fragmentary pelvis and ribsof a middle-aged (36-55 years) male. One skull fragment shows a"pinched" occiput, suggesting a long-headed individual. No occipitaldeformation is apparent. Strong areas for muscle attachment on theskull are indicated by large mastoid processes and a large mound-type occipital torus. The lower jaw is massive, with a mediobilateralcliin of medium projection, and strongly everted gonial angles. Noteeth were lost during life, but heavy (fourth degree) wear into the Riv. Bas. Sur.Pap. No. 4] ADDICKS BASIN?NEWMAN 259dentine exposed the pulp cavities of three of the molars. Apicalabscesses resulted, apparently involving the antrum in one case. Thewear is diagonal, heaviest on the lingual surfaces of the upper teethand the buccal surfaces of the lower teeth.Since only the condyles and part of the most distal shaft are missingfrom the right femur, an estimate of the dimensions of these missingparts was based upon a comparable femur from another collection.This gives an estimated maximum length of 430 mm. and suggestsa short stature in the neighborhood of 162-163 cm.Burial 2 {Doering site).?Partly restorable undeformed skull,part of right maxilla, most of lower jaw, and a few very small longbone and pelvic fragments of a middle-aged individual?probablyfemale. The sphenoid-shaped vault and "pinched" occiput suggesta long-headed individual. Cranial deformation is absent. Otherskull characteristics are: small, divided-type browridges; low fore-head of medium slope; medium development of median frontal andsagittal crests; small parietal bosses; medium temporal fullness;small mastoid processes ; medium lambda position ; and a ridge-formoccipital torus of medium size. The lower jaw is small, with narrowbilateral chin form, and only slight chin projection. The eversionof the gonial angles is medium. The lower right molar was lost dur-ing life through exposure of its pulp cavity by wear, with an apicalabscess resulting. The pulp cavity of the opposing upper molar wasalso exposed, but caused no apparent abscess. All the first molarspresent show excessive (fourth degree) wear, especially on the lingualsurfaces of the upper and the buccal surfaces of the lower teeth, wherethe entire crown and neck structure has been worn away. The secondmolars are less worn, although no enamel remains on their occlusalsurfaces. The third molars show even less wear, with the dentineonly visible in spots. No caries is present except in the cases ofexposure already noted.Table 1. ? Cranial measurements and indices Measurements (mm. ) and Indices No. 1male No. 2female? No. 4male No. 5female? No. 6female No. 7maleQlabello-occipital lengthMaximum breadthBasion-bregma heightAuricular heightMinimum frontal diameterBreadth of ascending ramusLength-breadth indexMean height indexLength-auricular height index-Module - 184'(150+) (188+)?128(130)(119) 172122(123) 18613334 32275 38(81.5)65.7 30268(82.3)(63.3)(148.7) 3070.9(83.7) 114903371.(139.0) 6L3 ' Figures In parentheses are close approximations. ' Estimates.Burials {Doering site).?Partly restorable skull vault, facial frag-ments, part of lower jaw, fragments of femoral and tibial shafts and 260 BUREAU OF AMERICAN ETHNIOLOGY [Bull. 154pelvis, bodies of three cervical vertebrae of a young adult (21-35years), probably female. What there is of the vault suggests amoderately long-headed individual. There is slight flattening on theocciput just above the inion, which may not be artificial. Other skullobservations : small, divided-type browridges with a medium glabella;low forehead of medium slope; small frontal bosses with no mediancrest in-between; medium mastoid processes; medium occipital form;medium-sized ridge-type occipital torus. The lower jaw is mediumin size, with a narrow bilateral chin. Gonial eversion is pronounced.All teeth, except a heavily worn upper incisor, were either lost orbroken off after death.The lower jaw exhibits a hook-shaped excrescence of bone just belowthe left mental foramen, which is probably of traumatic origin. Thethree bodies of cervical vertebrae show considerable crushing andlipping, probably arthritic. Figure 25.?Stereograph contour of the skull from Burial 4. (Approximately J^.)Burial 4 {Doering s^^5e).?Partly restorable skull, with parts of theface, most of the lower jaw, large portions of both femora, tibiae,humeri, pelvis, and part of the radii, fibulae, and sacrum of a male inadvanced middle-age. The skull is round and rather high-vaulted,with the length-breadth index estimated at 81.5, and the length-auricular height index 65.8. A trace of occipital flattening is present. i pip.Na'ff'^" ADDICKS BASIN?NEWMAN 261The strong areas for muscle attachment on the skull and long bonesbespeak a very rugged individual. An estimated maximum lengthfor the left femur of 455 mm. suggests a stature in the neighborhoodof 166 cm. Skull observations: Medium-sized divided-type brow-ridges; medium glabella; low forehead with pronounced slope; pro-nounced postorbital constriction; pronounced median frontal crest;medium sagittal elevation and parietal bosses ; large mastoid processes ; medium occipital curve; broad occipital form; large ridge-formoccipital torus; deep glenoid fossae; massive malars and lower jaw;mediobilateral chin form with medium chin projection; pronouncedeversion of gonial angles. Tooth loss cannot be determined. Wearon the three remaining lower molars is fourth degree. No mandibularcaries is present, but two apical abscesses resulting from pulp ex-posure are evident. On the mandible the left first premolar is ininverted position, with part of its crown showing through the mandi-bular bone on the lingual surface. The corresponding tooth on theright side erupted in normal orientation, but is misplaced toward thelingual side.A bony fusion of the left fibula and tibia took place about mid-shaft, and may be traumatic in origin. The body of the fifth lumbarvertebra shows heavy (arthritic?) lipping, as does the correspondingarticular surface on the sacrum.Burial 5 {Eobs site).?Part of the skull vault and base, showinganterior crushing; most of lower jaw; shafts of both femora andtibiae; left humerus; part of left pelvis of a middle-aged individual,probably female. The vault dimensions and the stereographic con-tour indicate a long-headed skull, with an estimated length-breadthindex over 68 from the measurements, and over 71 from the contour.Vault height appears low, with an estimated mean height index be-tween 82 and 83. Flattening occurs in the obelionic region, morelikely natural than artificial. Other cranial observations: pro-nounced sagittal elevation ; small parietal bosses ; slight temporal full-ness ; medium mastoid processes ; pronounced occipital curve and lowposition of lambda; pinched occipital form; ridge-formed occiptaltorus of medium size. The lower jaw is medium in size, with medio-bilateral chin form and pronounced eversion of the gonial angles. Noteeth are present for observation. The long bones are light andgracile.Burial 6 {Kohs site).?Skull lacking most of left parietal, but withface and lower jaw ; fragments of femoral, tibial, fibular and humeralshafts of an apparently young adult female. The skull is small,gracile, and long-vaulted, with a length-breadth index of 70.9. Itis not very high-vaulted, and has an estimated mean height index of83 to 84. Cranial deformation is absent. Other cranial data : small,divided-type browridges ; small glabella ; low forehead of pronounced 262 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154 Figure 26.?Stereograph contour of the skull from Burial 5. (Approximately 3^.) Figure 27.?Stereograph contour of the skull from Burial 6. (Approxinaately J^.) pip.?fo^lr' ADDICKS BASm?^NEWMAN 263slope with a large median frontal crest and small bosses ; pronouncedpostorbital constriction and slight temporal fullness ; medium mastoidprocesses; pronounced occipital curve, with low lambda position;pinched occiput; small mound-type occipital torus. Orbits are rec-tangular in form and medium in inclination; suborbital fossae aresmall ; anterior malar projection pronounced ; nasion depression slight;alveolar prognathism pronounced. The lower jaw is small, with amediobilateral chin of slight projection, and medium eversion ofgonial angles. No teeth were lost during life; no caries is present,and the wear is second degree (dentine visible) , Figure 28.?Stereograph contour of the skull from Burial 7. (Approximately J^.)Burial 7 (Kohs site).?Skull with fragments of face, almost com-plete lower jaw; complete right femur; fragments of left femur, bothtibiae, humeri, radii, ulnae, pelvis ; hand and foot bones ; and a few ribsof a middle-aged male individual. The skull is long-headed (71.5) andrather low-vaulted (length-auricular height index 61.3). There issome (natural ?) lambdoid flattening. Other cranial data: medium-sized divided browridges; medium glabella; low forehead of pro-nounced slope; pronounced postorbital constriction; small frontalbosses; medium median frontal crest and pronounced sagittal eleva-tion; over-medium parietal bosses; slight temporal fullness; largemastoid processes; pronounced occipital form; medium ridge-form 264 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154torus. The lower jaw is large with wide bilateral chin form of mediumprojection, and pronounced eversion of gonial angles. No teeth werelost during life; the diagonal tooth wear is fourth degree, which inthree cases exposed the pulp cavities and resulted in apical abscesses.The long bones are long and slender. The measurable right femurhas a maximum length of 482 mm., suggesting a stature of about172 cm. ANALYSISRajoial position.?The fragmentary remains from Burial 1 (Doeringsite), which is the sole candidate for a middle occupation phase posi-tion, suggest long-headedness and short stature. This information ispatently insufficient to permit any inferences concerning the middlej)hase population of the Addicks Eeservoir area.Burials 2-7, probably representing the late phase of occupation atthe Doering and the Kobs sites, afford a better but still inadequatesample of the protohistoric population. Pooling of these remains maybe justified from an archeological standpoint, since the late phasecultural remains from the two sites appear identical (see p. 257).Yet the physical differences between the skeletal remains from thetwo sites seem to call such pooling to question. In head form, forexample, the three Kobs site skulls (Burials 5, 6, 7) are on the borderof extreme long-headedness (c. 68, 70.9, 71.5 respectively). In thisrespect, they are similar to most precontact Texas crania from all butthe eastern part of the state (Goldstein ms.). For the Doering site,however, the skull from Burial 2 is barely long-headed (c. 75), andthe one from Burial 4 is definitely round-headed, with a closely ap-proximated index of 81.5.Morphological contrast between the Kobs site longheads and Burial4 is afforded by several observations as well. The longheads havepronounced sagittal elevations and occipital curves, while Burial 4 ismedium in both respects. The longheads also have pinched occiputsin contrast to the broad occiput of Burial 4. Any distinction in thelong bones is difficult to appraise. The greater thickness and rugged-ness of the shafts shown by Burial 4, as compared with the only Kobssite male (No. 7), may be only a function of shorter stature in theformer.The question raised by these physical differences may now be morefully stated : Are these differences of sufficient magnitude to make itunlikely that we are dealing with two samples of the same population ?A final answer cannot be expected, but some indications are given byan inspection of Goldstein's figures for two other south Texas series.The first comes from the Caplen Mound on Galveston Bay, a historicsite identified with the Atacapa.^ The second series is from the Oso KNa^T"^' ADDICKS BASIN?NEWMAN 265Mound, Nueces County, which is probably Karankawa.^ Limitingthe comparison to the length-breadth index, the Caplen Mound rangefor 13 male and female skulls is 73.4^83.8. The three Kobs site skullsfall below this range, but fit into the Oso range of approximately66-79.1 for 25 skulls of both sexes. The Doering site skull fromBurial 4, on the other hand, fits into the Caplen range. Accordingto Goldstein's observations, this skull would not be out of place inthe Caplen series.Wliether the Kobs site skulls would be out of place in the Caplenseries is the next question. To this, Goldstein has provided twoanswers. In the first place, he states that inland from the coast inthe historic Atacapa area, long-headed skulls are more frequent (per-sonal communication). Since the Addicks Keservoir area is some 60miles from the coast, as compared to the littoral location of the CaplenMound, increase in long-headedness might be expected in the former.Secondly, Goldstein's tables show that in all regions of Texas coveredby his study, the pre- or protohistoric series are longer-headed thanthose from contact sites (Goldstein, n. d.). This seems precisely thecontrast between the protohistoric Doering-Kobs series and the seriesfrom the historic Caplen Mound.In the light of these arguments, the Doering-Kobs sample couldhave been drawn from one population, provided it is assumed thatthe sample represents the extremes in head form. This view is essen-tially a conservative one, consistent with the premise that meager dataare more safely "lumped" than "split." The proviso to this view, how-ever, is by no means conservative. Although it is impossible to ap-praise accurately the chances that the head-form extremes of apopulation could be present in an unselected sample of seven skulls,it is likely that they would not be high. As an alternate view, theDoering skulls could be considered representative of the Atacapa(Caplen Mound), and the Kobs skulls the Karankawa (Oso) people.Some credence is given this view by Neumann's (n. d., p. 70) pro-nouncement that the Atacapa of Louisiana and the northern part ofthe Texas gulf coast were the westernmost of the Centralids typicalof the Southeastern States in late times. If this is so, the larger areaincluding Addicks Reservoir could have been a meeting ground ofthese long-headed and round-headed peoples. On this basis, the alter-nate view as stated above seems plausible. But it is hardly demon-strable, at least in the writer's opinion, by the meager data furnishedby seven fragmentary skulls lacking facial skeletons.From a physical standpoint there are several other notable features,which cannot for the most part be used in a comparative analysis.Table 1 indicates that the two long-headed Kobs site females (Nos. 5, * Also reported on by Woodbury, 1937. 266 BUREAU OF AMERICAN ETHNOLOGY [Bull. 1546) are not particularly high-vaulted, and in this respect most closelyresemble the females from west Texas around the Big Bend (Gold-stein, n. d.). The three female skeletons (Nos. 2, 5, 6) appear to bevery light-boned and gracile, except possibly for their masticatoryapparatus. Characteristic of the entire sample is a low forehead,with small frontal bosses, and medium to pronounced slope.Cranial defor'mation.?There is no definitely artificial cranial flat-tening in the Doering-Kobs series. The closest approach is the traceof occipital flattening on the skull of Burial 4. Burial 3 shows a slightamount of flattening just above the inion, which seems a natural ratherthan artificial phenomenon. Burials 5 and 7 show some apparentlynatural flattening above lambda.Pathology.?Manifestations of possibly traumatic origin are to beseen in the hooklike formation of bone on the lower jaw of Burial 3,and the midshaft fusion of the left tibia and fibula of Burial 4. Thereis no evidence of vault fractures.The cervical vertebrae of Burial 3 and the fifth lumbar of Burial4 show heavy lipping, probably of an arthritic nature.Heavy and usually diagonal wear on the teeth led to pulp exposureand consequent caries and apical abscesses in the case of Burials 1, 2,4, and 7. No teeth were present for observation in the cases of Burials3 and 6. Burial 6 was caries-free, and wear was only second degree(dentine visible). The heavy wear with occasional pulp exposure iswholly in line with Goldstein's (1948) observations.LITERATURE CITEDGoldstein, Mabcus S.1948. Dentition on Indian crania from Texas. Amer. Journ. PhysicalAntlirop., n .s. vol. 6, No. 1, pp. 63-84.n. d. Manuscript on Texas Indian skeletal material.Neumann, G. K.n. d. Racial differentiation in the American Indian. Ph. D. Thesis, Univ.Chicago, June 1950, 151 pp. MS.Newman, Marshall T.1947. Indian skeletal material from the Central Coast of Peru. Pap. Pea-body Mus., Harvard Univ., vol. 27, No. 4.Woodbury, G.1937. Notes on some skeletal remains of Texas. Univ. Texas Bull., No. 3734,Sept. 8, pp. 9-16. SMITHSONIAN INSTITUTIONBureau of American EthnologyBulletin 154 River Basin Surveys Papers, No. 5The Hodges Site I. Two Rock Shelters Near Tucumcari, New MexicoBy HERBERT W. DICK 267 CONTENTS PAGBIntroduction 271Excavation and stratigrapiiy 272Worked stone 274Projectile points 274Side scrapers 276Snub-nosed scrapers 276Rough flake knives 277Refined knives 277Drills 278Choppers 278Hammerstones 279Manos 279Metates 279Mortars 280Faunal material 280Animal bones 280Shells 280Ceramics 281Conclusions 282Literature cited 283ILLUSTRATIONSPLATES48. a, General view of the Hodges site. 6, Area A at the beginning ofexcavation 28449. a, Artifacts eroding out of the Number 2/3 fill under the rock fall in theshelter of Area A. h, Stratigraphy of section A-4, Area A 28450. Trench in the remaining fill of Area B 28451. Projectile points, drills, refined knives and snub-nosed scraper 28452. Scrapers and rough flake knives 28453. Manos, hammerstone, and choppers 28454. Fresh-water shells, Unimerus tetralasmus Say 284FIGURES29. Stratigraphy of the trench excavated in Area B 27230. Plan of Area A showing excavated portion and stratigraphy 273269 953842?53 19 TWO ROCK SHELTERS NEAR TUCUMCARI,NEW MEXICO By Herbert W. DickINTRODUCTIONThe excavation of the Hodges site was undertaken during August1947, because the site was being looted by workmen from constructionprojects nearby. The site actually was not in danger of inundationbut was being destroyed as completely as it would have been had thesite fallen within the pool area. Secondary factors were to aid thegeologist in determining the date of an alluvial deposit throughcultural material and to obtain and record additional archeologicalevidence for determining cultural types, the way of life, and thehistory in east central New Mexico.The geologic results are given in the second part of this dual reportby Sheldon Judson, of University of Wisconsin, whose problem is thatof determining the alluvial chronology of the region. The geologicstudy of the area was begun in 1941 by Judson.Hodges site consists of two shallow rock shelters 8 miles southeastof Tucumcari, N. Mex.^ (see Judson, fig. 32) on Plaza Larga Creek,an intermittent stream, which joins the Canadian River 21 miles tothe north. The Plaza Larga Creek originates in the breaks or north-ern edge of the Llano Estacado (Staked Plains) 31 miles southwestof the Hodges site. This creek takes the name "Tucumcari Creek" forthe lower half, after its confluence with Barranca Creek beforeentering the Canadian River.The immediate area surrounding the site is characterized by lowrolling hills which have little earth cover because of the erosionaleffects of the creek and its small tributaries (pi. 48, a) . The creek hascut two units of a Triassic sandstone to a depth of 20 to 30 feet. Thecontact zone of the two sandstones, red and gray, form a shelf. Inthe softer upper member, the red sandstone zone, shallow rock sheltershave been formed. ^The specific location of the Hodges site is Quay County, N%, NE14, sec. 31, T. 11 N.,R. 37 B.. New Mexico Principal Meridian. 271 272 BUREAU OP AMERICAN ETHNIOLOGY [Bull. 154EXCAVATION AND STRATIGRAPHYThe Hodges site was divided into two areas, A and B, which are twoshallow shelters situated on the west side of Plaza Larga Creek. AreaA is located some 200 yards north of Area B (pi. 48, &) . The excavation of the site can be classified as a salvage job. Despitethe elaborate treatment of the artifacts in this text the actual numberrecovered was small. The most notable fact concerning the artifactsis their diversity of type.Over three-fourths of Area B was removed by local archeologyenthusiasts before the writer's arrival. A single trench was excavatedin the remaining deposit of Area B (pi. 50). The stratigraphy ofthe trench (fig. 29) consisted of a top layer of sterile buff sand over- Surfcace -Buff sand - OterlLeLight ash lens -Rock fail -i uuff sand andash - Cultural laiftr ^Qnd lens 'Bedrock /yg> Figure 29.?Stratigraphy of the trench excavated in Area B.lying an ash lens which rested on a rock fall 6 to 9 inches thick.Directly under the rock fall was a deposit of buff sand liberally mixedwith ash, numerous bits of nacreous shells of fresh-water mussels, andartifacts. The deposit trenched was occupied up to the time of therock fall. The rock fall rests directly upon buff sand containing ashand artifacts. Subsequent occupation was of short duration as isindicated by the small ash lens, devoid of artifacts, directly above therock fall. All of the fill in the trench in Area B correlates withJudson's No. 2/3 Sand (pi. 49, a) . Kiv. Bas. Sur.Pap. No. 5] HODGES SITE?DICK 273About one half of the deposits of Area A (fig. 30) had been removedby erosion. Rock fall consisting of large blocks of sandstone filledthe shelter portion of Area A. The excavation of the area consistedof removing the major portion of the shallow deposits immediatelyin front of the shelter. A trench was excavated at right angles to the main sections to expose the deposits for a short distance underthe rock fall.Stratigraphy exposed under the rock fall in Area A indicated thatthe shelter was abandoned sometime before the rock fall took place.The rock rested directly upon 12 inches of sterile buff sand, underwhich was a layer of culture-bearing sand 8 inches in depth. 274 BUREAU OF AMERICAN ETHNDLOGY [BuU. 154The stratigraphy of Area A was of a shallow nature in front ofthe shelter proper (pi, 49, b). The deepest section measured 1 foot10 inches in depth: 3 inches of sterile buff sand (Judson's ModernSand) overlying 1 foot 7 inches of buff sand (Judson's No. 2/3 Sand)containing a liberal quantity of ash, numerous fragments of fresh-water mussel shells, and artifacts.The fill excavated in both areas was screened.WOEKED STONEPROJECTILE POINTSType 1 {pi. 51^ d^ e^ f).,5 examples.?This is the most common typeat the Hodges site. It is triangular with straight to convex sides;the base varies from straight to concave. Materials utilized werelight chalcedony, quartzite, and a mottled chert. All five of thepoints were found in rock shelter Area A in a homogeneous deposithaving a maximum depth of 2 feet 4 inches.Projectile points of this type have an exceedingly wide distributionin the southeastern periphery of the southwestern United States.They are reported by Holden from a site in Winkler County, Tex.,some 220 miles southeast of the Hodges site (Holden, 1938). Thecultures in the vicinity of Carlsbad, N. Mex. (150 miles south ofthe Hodges site) also show much resemblance to the lithic and generalcultural items found at Hodges site. Triangular points were foundin the Carlsbad region in midden circles composed of small fragmentsof limestone, showing evidence of hard firing, piled in circular heaps.The interstices of the heaps are filled with charcoal, ash, brokenpottery (Chupadero b/w and Plain Eed), bones, and fragmentaryfresh-water mussels (Mera, 1938) . The total material-culture contentof the areas above and in the San Jon area more closely resemblesthat of the Hodges site than that of all others studied.Far to the south, in the region of Corpus Christi Bay, Nueces County,Tex., triangular points are found in camp sites several acres in extent.These sites lacked potsherds and tanged points; triangular pointsalone were found (Martin, 1930). Sayles has designated triangularpoints as a characteristic type of the Jumano Phase (Sayles, 1935),marked by camp sites made up of sotol pits located in sand hills at themouths of small tributaries. It extends south of El Paso, in thevicinity of the Kio Grande toward the mouth of the Conchos Riverof Chihuahua, Mexico, and south along that stream.On the upper Red River drainage near Henrietta, Tex., in the areaof the historic Wichita Indians, Witte mentions one triangular projec-tile point found in a camp site (Witte, 1935). Triangular points alsoappear at a historic Wichita site termed "the Spanish Fort," locatedon the Red River in north central Texas (Witte, 1938). pip.No^S^]""^" HODGES SITE?DICK 275The writer found, in a survey along the lower Rio Puerco of theEast (in New Mexico), small unit sites exhibiting both Socorro b/wand Chupadero b/w pottery and small triangular points made ofobsidian. These were smaller and narrower than the ones found atthe Hodges site. Kidder found, in the Pueblo of Pecos (a glaze pot-tery site on the upper Pecos River in New Mexico), 398 triangularpoints out of a total of 1,190 projectile points and "knives." Hedivided the triangular points into three categories, depending onwhether they had convex, straight, or concave edges (Kidder, 1932).All of the triangular projectile points mentioned above are similarboth in size and in shape to the five found at the Hodges site. Be-cause of its wide distribution in several kinds of sites throughoutTexas, eastern New Mexico, and the main Pueblo area in centralNew Mexico, the type can be of little use as a guide object in deter-mining the specific group or phase of culture to which the sites belong.However, two significant facts do emerge from a study of the distribu-tion of the small triangular point: that it is usually found in con-junction with a pottery horizon, and that it is of a relatively late datein most sites.Type ^ {pi. 61^ h, j), 2 examples.?This type of projectile pointis finely chipped, triangular in shape, with narrow, corner notches andnarrow thin barbs ; the small expanding triangular bases are smallerin width than the shoulders. The barbs of the two specimens foundin Area B are three-fourths the length of the stem. The points werefound in the 4-foot level of Area B, a homogeneous ash-stained de-posit. Both were made of quartzite.A subtype (pi. 51, i) , 2 examples, is triangular in shape, with sideswidely notched so that the shoulders take on a flaring shape, and witha small, straight knoblike stem. It is made from a mottled chert notlocal to the immediate vicinity.Points very similar to type 2 and its subtype occur in camp sitesin the Carlsbad area and east of the Pecos and in large stone middens inthe Carlsbad area (Mera, 1938). In addition, distant variations ofthis type were found in most of the sites mentioned above in tracingthe distribution of type 1. However, these variations are character-ized by shorter, more flaring horizontal barbs, and straight peg-like stems with a more pointed base than is typical of the Hodges type2 projectile point.Type *X) a^ ^ 4ft 41 s> BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 54 Fresh-water shells, Unimerus letralasmus Say, SMITHSONIAN INSTITUTIONBureau of American EthnologyBulletin 154 River Basin Surveys Papers, No. 5The Hodges Site II. Geology of the Hodges Site, Quay County, New MexicoBy SHELDON JUDSON 285 953S42?53 20 CONTENTS PAGEIntroduction 289Physical setting 289The general Pleistocene sequence 292The sequence at the site 294The lower erosion surface 295The alluvial sequence 295No. 2 Fill 295No. 2/3 Sand 295No. 3 Fill 296Modern sand 296General considerations and conclusions 297Literature cited 301ILLUSTRATIONSFIGURES3 1 . Index map showing location of the Hodges and San Jon sites 29032. Block diagram to show the salient features of the Hodges site 29133. Generalized section to show the relation of the two erosion surfaces,the inner canyons, and the alluvial formations 29234. Bar graph to illustrate the relative duration of the various events inthe "Alluvial Chronology" 29435. Diagram to show the relation of the sand (No. 2/3 sand) containingthe culture horizon to the other deposits at the Hodges site 296287 GEOLOGY OF THE HODGES SITE, QUAY COUNTY,NEW MEXICO By Sheldon JxjdsonINTRODUCTIONThe archeology of the Hodges site occupies a critical position inthe "Alluvial Chronology" of eastern New Mexico. The study of thegeologic antiquity of the nearby San Jon, N. Mex., site (Koberts,1942) begun by the writer in the summer of 1941 demanded the de-velopment of this chronology.^ Moreover the Hodges site presentedideal conditions for cooperation between the archeologist and the geol-ogist. At this place occur bodies of sandy silts representing a knownperiod of wind activity of wide geographic extent and related to awell-known geologic sequence. These deposits contain cultural ma-terial and thus appeared susceptible to relatively precise dating. Inaddition, the cultural objects promised to throw some light on thepeoples living in this section of New Mexico at this particular time.The results of the study presented in this report and the companionpaper by Dick would seem to justify preliminary expectations.The site was first seen by the writer in September 1942, in the com-pany of the late Dr. Kirk Bryan, of Harvard University, and Dr.Franklin T. McCann, of Alabama Polytechnic Institute. It was notuntil August 1947, after a reexamination of the site by Bryan andJudson, that excavation was possible. The companion paper by Dickpresents the archeologic results of the investigation. Judson visitedthe site on two occasions during its excavation.PHYSICAL SETTINGThe site consists of two rock shelters in a sandstone cliff on the westbank of the Plaza Larga Creek, one of the tributaries of the Canadian 1 The study, of which this paper is a semi-independent part, was undertaken in con-junction with the excavation of the San Jon site. The continuing Interest, cooperationand assistance of Dr. Frank H. H. Roberts, Jr., Smithsonian Institution, has greatly facili-tated the work of the geologist In both field and office. Financial support was pro\idedby the Smithsonian. The late Dr. Kirk Bryan, Harvard University, supervised the fieldwork and has criticized the manuscript for this paper. 289 290 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154River in the so-called Canadian Valley. It lies in the NI/2, NE14,sec. 31, T. 11 N., R. 37 E., N. M. Pr. M. and B. and is 214 miles southof United States Highway 66, 1% miles east of New Mexico Highway18 and 8 miles south and east of Tucumcari (fig. 31).The Plaza Larga is a steep-sided, sandy-bottomed, intermittentstream which carries water only during and immediately after localrains. It is cut in comparatively soft clayey sand and silt of pre-viously deposited alluvium, but locally, as at the site, is borderedby rock cliffs. The introduction of irrigation water into the area ^already tends to increase the period of flow and to keep the stream { V. j^' Deaf Smith Co Tf^-iT i 1 " fA/ SfinJQn Site If .<-'h / -J'ct S ^ Curry Co. ??,? U.^ Figure 31.?Index map showing location of the Hodges and San Jon sites.bed moist for considerable lengths of time. Hence, grass, trees, andbushes will grow and for this reason the regime of the stream will bechanged and some aggradation of its channel will be induced. Thusthe irrigation project will very probably effect an artificial and localchange in the stream which will reverse the present trend.The Chinle formation, of Triassic age, forms the floor of the valleyplains of the Canadian River throughout this section of New Mexico.It is predominately brick-red in color, but locally green to buff. Itconsists of an upper member, the Redondo (Dobrovolny, Summer- " The Arch Hurley Conservancy District, centering around Tucumcari, will eventuallyconsist of 45,000 acres of Irrigated land. Water is being drawn from the lake behindConchas Dam, just below the confluence of the Conchas and Canadian Rivers, 30 milesnorthwest of Tucumcari. The dam was built before the war by the D. S. Army Engineersas a flood control project. Riv. Bas. Sur.Pap. No. 5] HODGES SITE?^JUDSON 291son, and Bates, 1946) characterized by thin-bedded, fissile shale, anda lower unnamed member, composed chiefly of massive, cross-beddedsandstone.The block diagram of figure 32 pictures the salient features of theHodges site. The two shelters are shallow caves formed in the sand-stone of the lower member of the Chinle formation. They are sepa-rated from each other by about 150 feet of near-vertical standstonecliffs, 25 to 30 feet in height. The floor of the northern shelter (Dick'sArea B) lies 10 feet above the modern grade of Plaza Larga, and thefloor of the southern shelter (Dick's Area A) is 18 feet above this localdatum. The cultural material is contained within a deposit of sandysilt laid down in the shelters as wind-borne dust during and after theiroccupation. Figure 32.?Block diagram to show the salient features of the Hodges site.In addition to these two shelters the culture-bearing sand extends 75feet southward from the southern shelter along a rock bench continu-ous with the shelter floor. Furthermore, artifacts are found mixedwith gravel along the high surface reaching back from the lip of thecliff in which the shelters are located.In places near the cliff edge, where the gravel has been strippedaway, milling pits have been sunk into the sandstone ledge. Theirposition presents an interesting correlation between a minor geologicfeature and human activity. The Triassic sandstone is criss-crossedby polygonal dessication cracks two to three feet in maximum dimen-sion. The milling pits have, almost without exception, been sunkalong these cracks or at their points of intersection. These weaknessesin the rock obviously afforded the easiest places for pit construction. 292 BUREAU OF AMERICAN ETHNOLOGY [BuU. 154THE GENERAL PLEISTOCENE ? SEQUENCESouth of the site the broad valley of the Canadian butts abruptlyagainst the northern escarpment of the Llano Estacado or SouthernHigh Plains. The valley, in places 50 miles wide, was cut during thePleistocene epoch by the Canadian River and its tributaries. The de-tailed record of valley development is fragmentary, however, and in-cludes only the most recent events in an otherwise long time interval.Two gravel-covered surfaces, cut chiefly in the Triassic rocks, formthe floor of most of the valley as shown in figure 33. The higher ofthese two surfaces, approximately 50 feet above the modem streamgrades, is best preserved along the headwater reaches of the tributariesto the Canadian. It represents a long period of plantation during 4 Late alluvial formations of the "Alluvial Chronology"3 Inner vadeys and canyonsZ Lov/er erosion Surface1 Upper erosion Surface [ j Sandstone and sViaie mostly of Triassic aqe Figure 33.?Generalized section to show the relation of the two erosion surfaces, the innercanyons, and the alluvial formations.which the Canadian River and its tributaries were stabilized at a levelmuch higher than the present. The gravel veneer on this bedrock sur-face was deposited by these streams during the final stages of erosion.The second and lower surface is approximately 30 feet above modernstream grade. It has been widely and well developed at the expenseof the higher surface. This surface marks a second planation whenthe local streams were graded to a lower level than that of the highersurface. The capping gravels represent the closing period of thisevent. Both surfaces are mantled by more recent wind-blown sandand by alluvium.Comparatively narrow inner valleys and canyons have been cut intothe lower of the two surfaces and form the modern bedrock grades. ? Pleistocene is here used to Include all time which has elapsed since the end of thePliocene. Such terms as "Recent" or "Post-glacial" are used only in an informal andlocal sense. (See Flint, 1947, p. 209.) pip.jfo^'sT''" HODGES SITE?^JUDSON 293The Canadian River itself flows in a bedrock gorge incised in placesto a depth of 200 feet below the general grade of these tributary val-leys. Figure 33 illustrates the relations of the modern bedrock chan-nels of the tributary of the Canadian to the remnants of the higherand lower surfaces.The formation of the erosion surfaces and the later stream incisionjust described were followed by various events of deposition anderosion which are of intimate concern to the archeologist. Althoughunspectacular when viewed against the immense scope of geologictime these brief events, mere incidents in earth history, coincide withthe human occupation of the valley from the day of the Paleo-Indianto the present. The changes from alluviation of the valleys to erosion,and then the reverse, form a sequence of events which undoubtedlyhad a profound effect on human activity. Cultural patterns musthave been influenced in ways as yet but dimly perceived and appre-ciated.The sequence of these events is deceptively simple. It is this verysimplicity, this monotonous, rhythmic repetition which dulls themind's retentiveness. Yet retention is important. The sequence iswidespread not only in eastern New Mexico but throughout the South-west. When man and his activity can be related to the sequence wehave an unparalleled stratigraphic tool to aid in the determinationof the relative and, in some cases, exact ages of the peoples concernedand the environment in which they lived.The sequence, the "Alluvial Chronology," is basically an alternationof stream alluviation with stream cutting. During certain periods inthe immediate geologic past the streams of the Southwest have filledtheir bedrock channels with silt, sand, and gravel. When the fillingwas complete the alluvial flood plains in some valleys were markedby shallow stream channels, whereas in others they were grassy andsmooth, marked only by quiet water holes, the charcos of Spanishdays. These intervals of alluviation are separated one from theother by periods of arroyo cutting, a process so characteristic of thearea today. Concurrently with the incision of channels strong windscarrying additional eolian material from the dry stream beds reacti-vated old sand dune areas.The sequence, discussed in general terms above, may be outlinedbriefly as suggested in the following tabulation. The tabulation inone sense is misleading. It suggests that the events have an approxi-mate equality in time, i. e., that each had the same duration as anyother. This is not true. Exact durations cannot be assigned to allthe incidents but in general the earlier the event the longer it lasted.Thus the bar graph in figure 34 indicates in a very crude fashion thesetime relationships. The higher the bar the greater the time involvedin each incident.953842?53 21 294 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154 :v^ Q ..o^ many places andNo. 2a Fill J given local names).No. 1/2 sand and dunes.No. 1 Fill (contains extinct animalsand is given various local names).THE SEQUENCE AT THE SITEBecause of the vagaries of deposition and preservation, the com-plete alhivial sequence is seldom displayed in any one place. Butthroughout a given drainage basin the sequence can be pieced togetherfrom scattered exposures of stratigraphically overlapping sections.Thus it is not surprising that the sequence at the site is incomplete.In this vicinity the No. 1 Fill, part of the No. 2 Fill, and the interven-ing sand deposits are missing, as is the highest erosion surface. Thesecan be found at other places within the drainage of the Plaza Larga,more particularly near the base of the escarpment of the Llano Esta- pip-Na'sT"^' HODGES SITE?^JUDSON 295cado or around outlying mesas, once integral parts of the Llano. Adescription of the sequence at the site is presented below.THE LOWER EROSION SURFACEThe higher erosion surface is not preserved in the vicinity of thesite but the lower surface is widely developed and here forms themajor part of the valley floor. Through it the Plaza Larga has cutits modern bedrock grade in a relatively narrow channel about 30 feetdeep. THE ALLUVIAL SEQUENCEWithin this bedrock channel a partial record of alluviation, arroyocutting, and eolian deposition is recorded. Part of the No. 2 Fill,the No. 3 Fill, intervening wind-borne material, and the modern wind-borne deposits are present. Thus only a part of the "Alluvial Chro-nology" need be discussed. NO. 2 FILLThis is a red sandy alluvium exhibiting a rudimentary jointingsystem. It is fairly compact but can be crumbled between the fingerswith some difficulty. Lime-filled tubules and films of lime along thejoint planes are characteristic. Toward its upper limit it contains adark humic zone. In the vicinity of the site the alluvium reachesa maximum thickness of 8 feet, and forms a small terrace along thestream. Elsewhere in the valley this formation is divisible into twounits on the basis of two humic zones, separated in places by erosionand concurrent wind-deposited sands. The No. 2 Fill here describedis believed to represent the upper of these two units, Fill No. 2 A offigure 34 and table 1. NO. 2/3 SANDA period of erosion or channel cutting similar to that of today fol-lowed the deposition of the No. 2 Fill. Wind activity accompaniedthis trenching and at favorable localities eolian deposits collected.This material is a reddish, well-sorted, fine-grained silt or sandy-siltresembling a loess in texture. Over 50 percent of the deposit hasa grade size less than 0.062 mm. It crumbles easily but nevertheless,in natural and artificial banks, stands with a vertical wall. It variesin thickness from a few inches to 7 feet. This is the deposit whichentombs the cultural material within the two shelters and also on therock bench continuous with the floor of the southern shelter. Theshelters trapped the sandy-silt as the wind carried it from the dryarroyo bottom. The sand can be traced south from the southern 296 BUREAU OF AMERICAN ETHNDLOGY [Bull. 154 Southern shelter(Dick's Area A") \ Side gully Present grade of Pla^a Larqa-Ve rt I ta I Scales in feet Mori J o n t a IModern -wind-borne sand and dustNo. 3 rill['!??' ??I No. 2/3 wind- borne sand and dust -Culture horfjonY//A NO. 2 nil I I Chinie Formation (Triossic aga^ Figure 35.?Diagram to show the relation of the sand (No. 2/3 sand) containing the culturehorizon to the other deposits at the Hodges site. shelter to a point where it is interbedded with alluvium and separatesthe Nos. 2 and 3 Fills as shown in figure 35.NO. 3 FILLFollowing the preceding erosion period the Plaza Larga began theaggradation of its channel and the No. 3 Fill was laid down. This isa gray, loose, friable alluvium containing little or no lime. In thisvicinity it does not exceed 2 feet in thickness. The great bulk ofit has been removed in the modern channel trenching but thin rem-nants of it mantle the stream banks and overlie the No. 2 Fill. Insome places the two fills are separated by a few inches of the No. 2/3Sand. MODERN SANDAbout 1900 the Plaza Larga began to cut its present channel. Inso doing it scoured out much of the No. 3 Fill until it again reached itsbedrock grade. With this cutting came another period of wind activityand concurrent eolian deposition. At the site the sand forms a discon-tinuous cover ranging from a thin film up to 6 inches. It is a gray,loose, sandy to loessic deposit. It is best developed within the sheltersoverlying the preceding wind-blown sand. Elsewhere in the areaactive sand dunes of considerable size and destructiveness mark this Pap.NasT'^' HODGES SITE?^JUDSON 297period of wind activity, and also the deep red scars of the modernarroyos testify to the effectiveness of stream erosion.GENERAL CONSIDERATIONS AND CONCLUSIONSSeveral conditions must be satisfied before the geologic method cansuccessfully be applied to an archeologic site. Bryan and Ray (1940)list these conditions as follows: (1) A well-defined culture-bearinghorizon related to a definite geologic event; (2) a local sequence intowhich this event may be fitted; (3) a general sequence of wide geo-graphic extent which includes this local sequence; and which (4)can be related to an absolute time scale.The Hodges site well illustrates the importance of satisfying theconditions outlined above. The open site above and behind the shel-ters defeats the geologist before he begins. The artifacts foundmixed with the gravels of the lower erosion surface, can hardly beconsidered contemporary with that relatively ancient period of graveldeposition. Aside from the gravels there is no geologic clue as totJie age of the artifacts, either relative or absolute. It seems appar-ent, however, from an examination of the site that the artifacts wereintruded into the gravels long after the gravels were laid down.Thus, because the artifacts cannot be tied to a local geologic eventthe first condition is not satisfied and the geologist can go no farther.The shelters are altogether different. Their stratigraphy providesan ideal point of departure for the geologic method. The culturalmaterial is entombed within a natural deposit representing a definitegeologic event, the blowing of fine dust out of the sandy bed of thePlaza Larga. Condition number one is satisfied. The second condi-tion is met because this dust can be traced southward where it liesalong the disconformity between the Nos. 2 and 3 Fills. Thus thesingle event is related to a local sequence of successive intervals oferosion and of alluviation of the stream channel. This local se-quence is a fragment of the larger sequence, the "Alluvial Chro-nology," which although not well displayed at the site can be es-tablished elsewhere in the Plaza Larga and in nearby streams. Thischronology is now established not only in eastern New Mexico butalso at many other localities in the Southwest. Requirement num-ber three is met. The final condition is answered by the archeology.Tlie cultural material provides a reasonably precise date for thelocal geologic event. The date is in accord with that arrived atfor similar events recorded through three Southwestern States.On the basis of the physical evidence, it is indisputable that atthe time of the occupation of the shelters the Plaza Larga was suf- 298 BUREAU OF AMERICAN ETHNOLOGY [BuU. 154fering from a period of channel trenching, wind activity, and eolianaccumulation of sand and dust. The conditions must have paralleledvery closely those of the present day. The Plaza Larga was then,as now, an intermittent stream. During most of the year the streambed was dry and winds funneled down its channel, picking up sandand dust and showering it on the occupants of the shelters.It is also evident that the Plaza Larga had a completely differentaspect between this erosion and the modern channel trenching. Thestream bed filled with alluvium (the No. 3 Fill), the old arroyo scarshealed, and water holes lined its now-eroded course, a conditionrecorded by the pioneers of the area. We further know that, imme-diately preceding the erosion period during which the shelters wereoccupied, the Plaza Larga was aggrading, and the stream had muchthe same appearance as did the grassy-bottomed draws of 50 yearsago.This fragmentary record of events keys into the "Alluvial Chro-nology" of Eastern New Mexico and the Southwest. This correla-tion is presented in table 1. The importance of the Hodges site liesin the fact that a given event in this chronology can be dated. Thewind-borne material representing the erosion interval between theNos. 2 and 3 Fills contains pottery and artifacts. Dick suggests thatthe material falls between A. D. 1300 and 1540. Thus the sandwhich contains this material and the arroyo cutting with whichit is correlative must share these dates. The dates cannot be consid-ered exact limits for this interval but probably approximate suchlimits. The arroyo cutting in the vicinity of the site is known tohave occurred throughout the Canadian Valley. The date deter-mined at the Hodges site may be applied to this event throughoutthe general area.The "Alluvial Chronology" of eastern New Mexico is similar inmost details to the late Pleistocene sequence throughout the South-west. The complete correlation of these sequences, as developed atvarious points in Texas, New Mexico, and Arizona, is not here pre-sented.^ Table 2, however, lists various localities at which are presenterosion intervals equivalent to the period of arroyo cutting duringwhich the shelters at the Hodges site were occupied. At several ofthese localities, as indicated in the table, an estimate of the time periodinvolved by this arroyo cutting and some statement concerning con-temporary human activity is possible.The actual date of this arroyo cutting undoubtedly varied fromplace to place throughout the Southwest, even as has the initial date * For a complete summary of these chronologies, the reader is referred to Bryan, 1941 ; Bryan and McCann, 1948 ; and Hack, 1942. Riv. Bas. Sur.Pap. No. 5] HODGES SITE?^JUDSON 299for the modern epicycle of erosion. The magnitude of this variation,however, was probably not great enough to account for the discrepan-cies among the various dates listed in table 2. The reason for thesediscrepancies appears, rather, to reflect the method of dating.An exact date for the interval at any locality is extremely difficult,if not impossible, to establish. Two conditions must be met beforean exact date can be determined. First, the cultural material mustbe related to both the beginning and to the end of the interval. Sec-ond, this material must be so diagnostic that the archeologist cannarrowly restrict it in time. The difficulties arising from these twoconditions are illustrated by the Hodges site. First, there is at thissite no evidence, either archeologic or geologic, that the occupationperiod in the shelters spans the entire period of the coincident arroyo.It probably does not. Therefore no date arrived at via the culturalobjects can represent the entire span of the arroyo. Second, evenwere the time interval bracketed by cultural horizons?and it isn't ? the material from the site cannot be dated as precisely as might bewished. Dick states that the material has a range from A. D. 1300to 1540 and may even range somewhat before A. D. 1300. But he canadduce no evidence that the occupation is actually defined by thisspan. On the contrary, although the pottery and artifacts have thesemaximum dates, they more probably represent a period of timeshorter by some unknown amount. Dick suggests that the actual timerange represented by the material lies closer to A. D. 1300 thanto A. D. 1540.Table 1. ? Correlation of the sequence at the Hodges site with that developed elsewherein the Canadian Valley and at the San Jon site along the northern edge of the southernHigh Plains Event 300 BUREAU OF AMERICAN ETHNOLOGY [BuU. 154Table 2. ? Correlation of the erosion intervals between the No. 2 and No. S fills,which is represented at the Hodges site by wind-borne silts, with similar events inTexas, New Mexico, and Arizona Locality and investigator pip.Na'sT'^' HODGES SITE JUDSON 301partial relation between arroyo cutting and the shifts in Puebloanpopulations which date from slightly prior to A. D. 1250 through1400. Furthermore, the coincidence between this arroyo cutting withthe drought of A. D. 1276 and 1299 as recorded in the tree rings of theSouthwest (Douglass, 1935, and Schulman, 1938) has been pointed outby several investigators (Bryan, 1941; Sayles and Antevs, 1941; andHack, 1942). Thus a growing body of geologic, archeologic andbotanical evidence gives ever-increasing support to the inference thata period of arroyo cutting existed throughout the Southwest betweenA. D. 1200 and 1400 and centered around A. D. 1300.The Hodges site enjoj^s an importance out of all proportion to itsantiquity or archeologic plunder. Its stratigraphy and archeologycombine to establish it as another link in the "Alluvial Chronology,"the chronology which promises to add much to our understanding ofthe complex history of the Late Pleistocene in the Southwest.LITERATURE CITEDAlbrttton, Claude C, Je. See Huffington, Roy M., and Aubritton, OlaxideC, JB.Albritton, Claude C, Jr., and Bryan, Kirk.1939. Quarteruary stratigraphy in the Davis Mountains, Trans-Pecos Texas.Bull. Geol. Soc. Amer., vol. 50, pp. 142.3-1474.Antevs, Ernst. See Sayles, E. B., and Antevs, Ernst.Bates, R. L. See Dobrovolny, Ernst ; Summerson, C. H. ; and Bates, R. L.Bryan, Kirk.1941. Pre-Columbian agriculture in the Southwest, as conditioned by periodsof alluviation. Ann. Assoc. Amer. Geogr., vol. 31, pp. 219-242.See also Albritton, Claude C, Jr., and Bryan, Kirk.Bryan, Kirk, and McCann, F. T.1943. Sand dunes and alluvium near Grants, New Mexico. Amer. Antiq.,vol. 8, pp. 281-295.Bryan, Kirk, and Ray, Louis L.1940. Geologic antiquity of the Lindenmeir site in Colorado, SmithsonianMisc. Coll., vol. 99, No. 2.Campbell, T. N. See Kelley, J. Charles ; Campbell, T. N. ; and Lehmer,Donald, J.Dobrovolny, Ernst; Summerson, C. H. ; and Bates, R. L.1946. Geology of northwestern Quay County, New Mexico. U. S. G. S. Oiland Gas investigations, preliminary map 62.Douglass, A. E.1935. Dating Pueblo Bonito and other ruins of the Southwest. Nat. Geogr.Soc. contributed tech. pap., Pueblo Bonito series, No. 1.Flint, R. F.1947. Glacial geology and the Pleistocene epoch. New York.Hack, J. T.1942. The changing physical environment of the Hopi Indians of Arizona.Reports of the Awatovi expedition, Peabody Museum, Harvard Uni-versity. Pap. Peabody Mus. Amer. Arch, and Ethnol., Har-vard Univ., vol. 35, No. 1. 953842?53- 21 302 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154Hack, J. T.?Continued1945. Recent geology of the Tsegi Canyon. In Beals, R. L, Archaeologicalstudies in northeast Arizona, pp. 151-158. Univ. Calif. Publ. Amer.Arch, and Ethnol. Berkeley.Htjffington, Roy M., and Albritton, Claude C, Jk.1941. Quarternary sands on the Southern High Plains of western Texas.Amer. Journ. Sci., vol. 239, pp. 325-338.Kelley, J. Charles ; Campbell, T. N. ; and Lehmek, Donauj J.1940. The association of archeological materials with geological deposits inthe Big Bend region of Texas. Sul Ross State Teachers College,Bull., vol. 21, No. 3, Arch, issue and West Texas Hist, and Sci. Soc,Publ. No. 10.Lehmeb, Donald J. See Kelley, J. Charles ; Campbell, T. N. ; and Lehmee,Donald J.McCann, F. T. See Bryan, Kirk, and McCann, F. T.Uay, Louis L. See Bryan, Kirk, and Ray, Lotjis L.Roberts, Frank H. H., Jb.1942. Archeological and geological investigations in the San Jon district,eastern New Mexico. Smithsonian Misc. Coll., vol. 103, No. 2.Sayles, E. B., and Antevs, Ernst.1941. The Cochise culture. Medallion Pap., No. 29. Gila Pueblo, Globe,Ariz.ScHULMAN, Edmund.1938. Nineteen centuries of rainfall history in the Southwest. Amer. Meteor.Soc, Bull., vol. 19, pp. 211-215.SUMMEBSON, C. H. See Dobbovolny, Ernst ; Summebson, C. H. ; and Bates, R. L. SMITHSONIAN INSTITUTIONBureau of American EthnologyBulletin 154 River Basin Surveys Papers, No. 6The Rembert Mounds, Elbert County, GeorgiaBy JOSEPH R. CALDWELL 303 CONTENTS PAGEIntroduction 307The large mound 309The small mound 314The village site 314Pottery 315Lamar Complicated Stamped 316Lamar Plain 316Lamar Bold Incised 316Other types of pottery presumably made during the Lamar period atthe Rembert site 316Earlier pottery types 317Artifacts other than potsherds 317Trait list 318Conclusions 319Literature cited 320ILLUSTRATIONSPLATES55. a, Aerial view of the Rembert mounds' location and the accompanyingvillage site. The outlined area indicates the mounds, and the loca-tion of the test pits is shown by the black dots in the field below.b, Pottery pipes, stone disks, projectile point, and shell beads 32056. Potsherd types 320FIGURES36. C. C. Jones' map of the Rembert mounds 30837. Map showing test pits and wash-outs in the vicinity of the largemound remnant 31138. Profile of the large mound, adapted from Thomas 31239. Section through large mound, adapted from Thomas 31240. Profiles of test pits in the large mound and vicinity 313305 THE REMBERT MOUNDS, ELBERT COUNTY,GEORGIA By Joseph R. CaldwellINTRODUCTIONDuring an archeological survey of the Clark Hill Reservoir inGeorgia and South Carolina, by Carl F. Miller and the writer fromJanuary 12 to June 1, 1948, a number of test excavations were madeat the site of the former Rembert mounds. This important group,almost totally destroyed since it was first described by William Bar-tram in 1773 (Van Doren, 1928, pp. 265-66), has never adequatelybeen treated in a scientific report, and indeed is scarcely mentioned inany recent publication dealing with southeastern archeology. Thewaters of the new Clark Hill Reservoir will cover the place sometimein 1952, and it is possible that no more digging ever will be done.The locality of the former mounds is in Elbert County, Ga., on thewestern side of the Savannah River 3 miles above its confluence withthe Broad River and 50 miles above Augusta. There, the river turnseastward on its southeasterly course, curving around broad bottomlands. The aboriginal remains are situated close to the river at thenorthwestern end of the bend (pi. 55, a). The owner of the landis Webb Tatum of Elberton, Ga. His son, Webb Tatum, Jr., who wasmanaging the property at the time of our visit, courteously permittedus to make our investigations.When Bartram described the works in 1773, they comprised animposing group of one large and several smaller mounds standingadjacent to some extensive structures which he called tetragon ter-races, Bartram was impressed by the size of the main mound, thelargest he had ever seen (Bartram, N. D.). Seventy-five years later,the site was visited by George White who noticed that the smallermounds had been nearly destroyed (White, 1849, pp. 229-230) . In1878, Charles C. Jones, Jr., reported that only traces of the smallermounds yet remained (fig. 36) and that the tetragon terraces were nomore that gentle elevations (Jones, 1878, pp. 284-285) . John P. Roganmade an archeological investigation at the site in 1886, under the307 308 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154 Figure 36.?C. C. Jones' map of the Rembert mounds.general direction of Cyrus Thomas. Thomas reported that only thelargest mound and one of the smaller were still standing at that time(Thomas, 1894, pp. 315-317) .^When the writer arrived in 1948, it was learned that the largemound had been almost completely demolished during a particularlydestructive flood in 1908. The remaining basal portions of the mainmound were pointed out by Dude Dubose, an aged man who hadplayed about there when a child, and his identification of the ruin wasconfirmed by another elderly man introduced to us only as the "Preacher." An unsuccessful search was also made for the smallmound which had still been visible in 1886.During 3 weeks' work, the survey party excavated five test pits inand about the large mound and dug six pits in the adjacent villagearea. Our purpose was to see if any important structural featuresmight be present and whether there would be a great loss of materialif the site were allowed to go under water without further excavation.These pits provided cross sections of strata and brought to light aconsiderable number of faunal remains, fragments of aboriginal pot-tery, and a few other artifacts.The decorated potsherds most frequently occurring in the test pitscould readily be assigned to the widespread type named Lamar Com-plicated Stamped (Southeastern Archeological Conference, 1939) * Most of the material recovered during Thomas' excavation program was subsequentlyplaced In the United States National Museum, Strangely, there is nothing there from theRembert site. pIp-nTg^]"^' REMBERT mounds CALDWELL 309which has been found throughout Georgia and in parts of South Caro-lina and eastern Tennessee. This is the marker type for the Lamargroup of cultures, which in Georgia is correlated with Muskogean-speaking peoples, some of which later were known as Creeks. Thepottery style is currently dated between A. D. 1450 and 1650. TheEembert mounds may not have been deserted for more than 200 yearsbefore Bartram's visit.THE LARGE MOUNDAll that remained of the large mound in 1948 was an irregularremnant hidden in a fringe of scrub timber bordering the river(pi. 55, a). The remains reached to within 130 feet of the water'sedge forming an elevation about 4 feet above the surrounding landand about 118 by 35 feet in extent. The rampaging of the river waseverywhere apparent : washouts noted by Jones and Thomas were stillpronounced, and another, reaching between them around the remainsof the mound, was plainly that which had caused its destruction.In 1773 the appearance of the large mound had been described byBartram as follows : The great mount is in the form of a cone, about 40 or 50 feet high, and thecircumference of its base 200 or 300 yards, entirely composed of the rich loamyearth of the low grounds : the top or apex is flat : a spiral track or path leadingfrom the ground up to the top is still visible, where now grows a large, beautifulspreading red cedar (Juniperus Americana) : there appear four niches excavatedout of the sides of the hill, at different heights from the base, fronting tlie fourcardinal points; these niches or sentry boxes are entered into from the windingpath, and seem to have been meant for resting places or lookouts.Jones in 1878 (fig. 36) wrote about the mound in this fashion:Overleaping the river bank, the turbid waters had carved deep pathways inthe surface of the valley on both sides of the "great mount." There it remained,however, totally unaffected by these unusual currents. It had evidently sufferedno perceptible diminution in its recorded dimensions. The Savannah Riverstill pursued its long established channel, but "the four niches or sentry boxes,"if they formerly existed, were entirely gone, and of "the spiral path or trackleading from the ground up to the top" we could discover no trace. On thesouth a roadway, about 15 feet wide and commencing at a point some distancefrom the base of the mound, leads with a regular grade to the top. This mani-festly furnished the customary means of ascent, as the sides are too precipitousfor convenient climbing. This feature seems to have escaped Mr. Bartram'sobservation. . . .The material employed in erecting this large tumulus differs from the soilof the surrounding bottom. It is a dark-colored tenacious clay while the surfaceof the valley is covered with a micaceous loam readily dissolving into an almostimpalpable powder. Nearby are no traces of pits or excavations. Nor arethere any indications that any earth was scraped up around the base. Thesefacts afforded confirmation of the statement made by the present owner ofthe plantation upon which these tumuli are located, that the big mound hadbeen built with clay brought from the Carolina side of the Savannah River. 310 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154There clay abounds ; and we were informed that in the side of tlie hill imme-diately opposite, the excavations may still be seen whence the tough materialwas obtained for heaping up this mound.Thomas, 1894 (figs. 38, 39) described Eogan's excavations in thelarge mound : This, which is much the larger of the two, stands 130 feet from the riverbank, and is, exclusive of the ramp or projection, an exact circle 151 feet indiameter, nearly flat on top, and 30 feet high at the highest point (north side),but only 27 feet near the south side. The diameter of the top is about 70 feet.The plan of the ramp or rather extension, as it seems to be, is shown, . . .The vertical outline of the mound, with a section of the shaft, is presented . . .[see fig. 39]. The right or southern end of this shows the slope of the extension.This has an average width on top of 20 feet.The mound is covered with trees such as sugarberry, walnut, hickory andoak. One sugarberry is 6 feet in circumference (at stump height) ; a walnut,5 feet; a hickory, 314 feet; and an oak 10 feet. The shaft was carried down tothe bottom. The first foot was of soil [fig. 39, A], then 7 feet of dark sandyloam [B], next ll^ feet of thoroughly burned yellowish clay and sand [(7],with a large percentage of ashes. This layer had the appearance of having beenput down and packed while wet and then burned ; it was so hard that it wasdiflicult to break it. Next 3 feet of black eartli, also packed [D] ; then SYzfeet of pure sand [E] ; and last, resting on the original surface, 6 feet of hardbluish muck [F]. All these layers, except the bottom one, had charcoal, mica,fragments of pottery, and animal bones scattered through them, but the lastwere so far decomposed that none of them could be saved.As fragments of pottery and animal bones were found in spots, together withashes and other indications of fire, it is probable that these were Are beds wherecooking had been done. All that portion of the shaft below the layer of burnedclay was so very dry that when turned up it would crumble to dust. It ispossible that the bottom layer of blue "muck" is partly the original soil, as itis so much like the surrounding soil, and that a part of the surrounding surfacehas been washed away since the mound was built.Our authorities do agree that the mound was circular with a flattop, and both Jones and Thomas describe and illustrate a ramp pro-jecting from the southern side. Bartram's failure to mention thisramp, which could not have been less conspicuous in his day, castsdoubt upon his observation of a spiral path with four niches orsentry boxes. Jones found no such features, and questioned whetherthey ever had been present. A spiral path would seemingly have beeninterrupted by the projecting ramp. No such niches as Bartramnoticed have ever been reported in the Southeast, but mounds withtwo ramps, one projecting and the other ascending along one side,have been found in Georgia, for example, on the large Tumlinmound near Cartersville in the Etowah Valley (Thomas, 1894, pi, 16)and on one of the superimposed mounds at Irene on the Georgia coast(Caldwell and McCann, 1941, fig. 9).Jones' suggestion that the mound had been built of clay broughtacross the Savannah Kiver seems not only incredible, but senseless on Riv. Bas. Sur.Pap. No. 6] REMBERT MOUNDS?CALDWELL 311the part of the builders. Bartram's opinion was that the mound was "entirely composed of the rich loamy earth of the low grounds," andThomas, discussing the washouts flanking the mound, says that "thereare reasons for believing that at least a portion of the earth used inthe construction of the mounds was obtained here, leaving depres-sions, and that, during high water, when the land was overflowed,as is frequently the case, channels were washed out from them to theriver." None of the mound layers encountered by Rogan and Thomascorresponded to the type of clay described by Jones except possiblythe extreme bottom layer which Jones could not have seen. The realcoup de grace to Jones' theory, however, is administered by the multi-tude of potsherds, animal bones, etc., which Thomas found scatteredthrough the mound and which must have been scraped up with thesoil from the adjacent village site.Our own test pits in the mound, Nos. 2 and 6 (fig. 40) showedthat a considerable portion of the mound base had not been disturbedby the flood of 1908. Our strata can be correlated with Thomas'account of the mound structure when we consider that his upper levelshave been washed away. All the laj^ers we noticed in pits 2 and 6 oc-curred in Thomas' second lowest level, that which he called 8i/^ feetof pure sand. Thomas wrote that this layer was the lowest stratumcontaining potsherds, thus it must have extended down to our pre- FiGURE 37.?Map showing test pits and wash-outs in the vicinity of the large moundremnant. Adapted from a topographic map by Kelly Mims, U. S. Engineers. 312 BUREAU OF AMERICAN ETHnSTOLOGY [Bull. 154mound occupation layer. The premound occupation zone is shown infigure 40 on the profile of pit 6. It was also found in pit 2, but did notappear on the west wall which is illustrated. This layer was composedof a 2-inch thick deposit of dark gray sand containing numerous pot-sherds and animal bones. Augur tests showed that the undisturbedsand below became gradually infused with clay as it extended down-ward. All the mound layers we found in the two test pits slopeddownward toward a central point which presumably was the centerof the mound. The significance of this cannot be determined with-out additional excavation. WASHOUT LARGE MOUND WASHOUTFigure 38.?Profile of large mound, adapted from Thomas.Thomas' lowest level was : . . . resting on the original surface, 6 feet of hard bluish muck. . . .It is possible that the bottom layer of blue "muck" is partly the original soil, asit is so much like the surrounding soil. . . .Our own findings corroborated Thomas' opinion that his lowest layerwas natural and not part of the mound structure. It appeared to beundisturbed in our pits, contained no cultural material, and did indeedresemble the soil in the bottoms surrounding the mound. Thomas'/ By RlT. Bas. Sur.Pap. No. 6] REMBERT MOUNDS CALDWELL 313have been the point toward which the deposits in pits 2 and 6 slopeddownward, then the position of pits 3 and 5, respectively 60 and 80feet from the supposed center, is approximately the radius of themound, the diameter of which, according to Thomas, was 151 feet.The three upper layers in pit 3 (fig. 40) were water deposited andmay have developed since the abandonment of the site by theIndians. Layer 4 was composed of dark organically stained sandcontaining a considerable amount of clay in massive deposits. Allthe cultural material found in the pit came from this layer, which was ^*MO aecoMiM: wnjsco w/th sil t fi/TJ SOUrH liVAti. Figure 40.?Profiles of test pits in the large mound and vicinity.full of potsherds, animal bones, and irregular rocks and pebbles. Thebones looked extremely fresh when first brought to light and not atall decomposed. The bulk of the deposit sloped downward towardthe west. Unfortunately, we did not have time to enlarge thisexcavation.Pits 4 and 5 were dug 23 feet to the south and 24 feet to the westof pit 3 respectively. Pit 5 also showed the outward slope noticedin pit 3, probably the feather edge of the mound. Pit 4 containedthe same strata as pits 3 and 5, but did not show any slope. 314 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154THE SMALL MOUNDIn Thomas' time only one of the smaller mounds was still noticeable.He described it as standing : . . . about 40 feet west of the base of No. 1. It is oblong in form, 58 feet longnorth and south, 41 feet wide, and 6 feet high. A large shaft had been sunk inthe middle by some previous explorer, hence investigations were confined tothe eastern and western sides, which presented one or two peculiarities. Withthe exception of the top layer of soil, one foot thick, the remainder of the eastside consisted of river sand, with particles of charcoal and vegetable mattermixed through it, while on the west it was composed of small masses of redclay and dark earth. In this, at a depth of 2^^ feet were the bones of a singleadult skeleton. These were packed together in a space 2 feet square and 18inches deep ; the skull was placed face down and all the other bones piled aboutit. Immediately over the bones was a layer of red clay 2 inches thick, burnedhard. Resting on this layer were the remains of a pretty thoroughly burnedfire. A few fragments of pottery and a small clay pipe were found.We presume that the other small mounds seen by Bartram andJones were burial mounds. White (1849, p. 230) wrote that:Captain Rembert has excavated the smaller mounds and found human skele-tons, jars, pipes, beads, breast-plates, stone hammers, hatchets, arrowheads,etc. . . . THE VILLAGE SITEThere is no doubt that the Rembert mounds stood upon an extensivevillage site. Bartram stated that there were : , . . some very large tetragon terraces on each side [of the mound group], nearone hundred yards in length, and their surface four, six, eight, and ten feet abovethe ground on which they stand.*We have already noted that when Jones visited the site : The tetragon terraces had lost their distinctive outlines, and were little morethan gentle elevations ; their surface littered with sherds of pottery, flint chips,and occasionally fragments of human bones.In 1948 no surface materials were visible in the area adjacent to themound except for a few sherds in the old washout to the west. Sixtest pits were dug in the general village area (pits 7-12) and culturalmaterial appeared in four of them. The locations of these pits areshown in plate 55, a. Each one was only about 2 feet square andit seemed unnecessary to mark them for future reference. The cul- * Bartram's term, "tetragon terrace" which he also calls a "four square" terrace, Istoday obscure. Elsewhere In his writings he has said "the tetragon terraces seem to bethe foundation of a fortress" (Van Doren, 1928, p. 407), and "Cherokee mounts are alwaysaccompanied by the vast tetragon terraces placed at one side or the other" (Bartram MS.,p. 83). He states that tetragon terraces are characteristic of the region of northernGeorgia and adjacent States which were formerly occupied by the Cherokees, but are notfound In the lower Creek country south and west of the Altamaha River. (See VanDoren, 1928, pp. 406-407.) pip.N^o^'eT"^' REMBERT MOUNDS?CALDWELL 315tural deposits encountered in the pits varied from 0.7 to 1.4 feet intliickness' and in pit 7 obtained a thickness of 3.3 feet. Most of thematerial found consisted of potsherds, but pit 12 yielded animal bonesand ash as well. Exploratory trenches in this area would be advisableif additional work should be done at the site.POTTERYThe numerous potsherds found by us at the Rembert site wereexamined while they were coming out of the test pits but there wasno evidence of stratigraphic change in the pottery types. The largemound and village area were used for the most part during onemajor period, although random sherds datable to earlier times indicatesporadic occupations by previous peoples. The bulk of the decoratedsherds from all the test pits belong to the type of pottery calledLamar Complicated Stamped. This is a common style at late sitesin most of Georgia and the adjoining parts of South Carolina andeastern Tennessee. Fairbanks has suggested the dates 1540-1650 forthe type as it occurs in central Georgia (Fairbanks, 1946, p. 103), theregion where it was first described ( Southeastern Archeological Con-ference, 1939) . A terminal date of 1600 was suggested for the variantwhich was found at the Irene site on the Georgia coast (Caldwell andMcCann, 1941, pp. 46-47). A recent statement is that:The Lamar site at Macon and the Irene site on the coast are among the latestprehistoric sites in their respective areas. No historic materials have beenfound at either, but Waring has reported Spanish wheel-made pottery fromsome coastal Lamar sites, and historic chinaware and nails were found in theupper levels at Hollywood [near Augusta on the Savannah River]. A numberof historic Cherokee sites in eastern Tennessee and western North Carolinahave a Lamar-like pottery which is considerably later than in Georgia. It islikely that many and perhaps all Lamar-like sites are post DeSoto [Caldwell,n. d.].A. R. Kelly and Gordon R. Willey are now inclined to a somewhatearlier begimiing date for Lamar Complicated Stamped (personalcommunication from Gordon R. Willey) but no earlier than 1450.We are on fairly sure ground if we place the occupation at the Rembertsite between this date and 1650.The most abundant single type of pottery at the Rembert site wasLamar Plain. A variant of this has been described on the Georgiacoast (Caldwell and McCann, 1941, pp. 48-49) but not in centralGeorgia where it also occurs. At the Rembert site as elsewhere, asso-ciation with the stamped type and correspondence in many ceramicfeatures other than surface finish indicate that Lamar Plain wasmade at the same time as Lamar Complicated Stamped. This is alsotrue of the type Lamar Bold Incised (News Letter, SoutheasternArcheological Conference, 1939). The corresponding type on the 316 BUREAU OF AMERICAN ETHNOLOGY [Bull. 154Georgia coast is Irene Incised (Caldwell and McCann, 1941, pp.47-48). LAMAR COMPLICATED STAMPED (pi. 56, A-F)This type of pottery at the Eembert site was made of gritty localclay by the coiling method usual in the southeast, and further tem-pered with relatively large particles of quartz. The exterior surfacesof the vessels were fired to colors ranging from dark gray throughshades of brown to red and buff. The firing was not well controlled.A single color is often seen to prevail over a large vessel fragment,but it is not unusual to find a sherd showing two or more shades ofcolor. The interior surfaces were frequently darker than the exterior,generally well smoothed, often burnished.The entire exterior surfaces of the vessels were stamped with im-pressions of carved paddles, possibly of pottery or wood. A ratherelaborate fylfot cross was the principal design motif (pi. 6&,A-D) hereas in central Georgia and on the coast, but other designs are found aswell (pi. 56, E, F) in all three areas. Most vessel rims at the Eem-bert site have an applique strip below the lip, which was decorated byincising, notching, or pinching. No complete vessels were found, butthe sherds indicate that the customary Lamar form of jar with roundedbottom, rounded sides, and outcurving rim was present. The sidesseem to be somewhat straighter and the rim less flaring than in centralGeorgia or on the coast.LAMAB PLAIN (PLATE 56, G-L)The unusually high proportion of plain pottery at the Eembert siteis itself a noteworthy fact. The sherds conform well to the coastaltype in paste and surface finish, but there are proportionately morejars and fewer bowls. The former have the straighter sides and lesspronounced rim flare which we noted to be characteristic of thestamped pottery. Similar applique rim treatment is frequentlyfound. LAMAR BOLD INCISED (PLATE 56, M)Only three sherds of this type were found, and since a much higherproportion is usual at Lamar sites elsewhere, we consider the scarcityof Lamar Bold Incised to be a distinctive feature of the local potterycomplex.OTHER TYPES OF POTTERY PRESUMABLY MADE DURING THE LAMARPERIOD AT THE REMBERT SITEA few other sherds are believed to have been made at the same timeas the Lamar types because they conform to these in most respects pip-No^'eT"^' REMBERT MOUNDS?CALDWELL 317except surface finish. Tliey were too few, however, to be assignedtype names. Plate 56, N is one of a group which shows a curiouslyrough and uneven surface. This might be intentional roughening orpossibly the circmnstance of the vessel having been fired without firsthaving been smoothed. Other sherds show a lightly brushed orcombed surface (pi. 56, ^), still fewer are cord roughened (pi. 56, S),and two sherds look as though they might have been decorated witha corncob, although it is unlikely that this was the method used.EARLIER- POTTERY TYPESSome sherds resembling types which in other areas have been foundstratigraphically below Lamar pottery show that the Rembert site wasoccasionally occupied during earlier times. A few may be SavannahCheck Stamped (not shown) (Caldwell and McCann, 1941, pp. 4ArA5) .Others had the distinctive bar and triangle and bar and circle designswhich are called Etowah Stamped (pi. 56, O, P) . This type has beenfound abundantly at the Tumlin mounds near Cartersville, Ga., andat other sites on the Etowah River (Wauchope, 1948, pi. 18, B, toprow). Still earlier pottery in the Etowah area is like the checkstamped sherd shown in plate 56, F, the simple stamped sherd, 56, U^and the sherd decorated by a cord-wrapped stick, 56, T. Not shownis a specimen of the type Woodstock Diamond Stamped (Wauchope,1948, pi. 18, A, top row), and an unidentified net-marked sherd.ARTIFACTS OTHER THAN POTSHERDSSuch objects were not especially frequent, but nearly all specimenswere characteristic of other Lamar sites in Georgia.Sherd disks {pi. 55^ b, E) . ?Seven disks made of sherds were found.Four of them were made from Lamar Complicated Stamped sherds,two were plain, and the decoration of the other was indistinct.Pottery pipes {pi. 55, 5, Z>, H).?There was one complete potteryelbow pipe and one pipe bowl fragment. No tempering material wasvisible in either. Both had burnished exterior surfaces.Fired clay wall plaster {pi. 55, h,A) .?Three small fragments show-ing wattle impressions indicated the former presence of wattle anddaub buildings.Bone tube {not illustrated).?A fragment of a cut bone tube (prob-ably turkey tibio-tarsus) 41/^ inches long and ^ inch in diameterwas found. A roughly triangular perforation in one side three-fifthsof the distance from the cut end may be accidental, for the specimenis in bad condition.Stone disk {pi. 55, b, F).?There was one small greenstone disk,somewhat chipped. Thickness about % inch. 318 BUREAU OF AMERICAN ETHNOLOGY [Bull. 354Stone projectile points.?In this region at present, only a few typesof stone points can be assigned to chronological periods. One of thetwo points found at Rembert's (/), made of quartz, can be duplicatedat other Lamar sites. The remaining specimen, a fairly large coarselyflaked stemmed point of weathered chert, resembles examples fromStalling's Island (see Claflin, 1931, for various examples of earlypoints) and is probably an old type.Stone chips {not illustrated).?Several chips and flakes of quartzand chert were recovered in the excavations, but such materials aremuch less frequent at Rembert's than at most sites in the vicinity.Shell {pi. 65, &, B, C) . ?One massive bead made of the central colu-mella of conch, and two small disk beads probably also conch, werefound. There was also one cut fragment of conch shell. Such shellsmust have been brought from the ocean, probably from the Georgia-South Carolina coast.Other materials.?There were two small unworked fragments ofsheet mica, obtainable locally, and a piece of bitumen, probably fromthe coast. TRAIT LISTGeneral : A fair-sized town with a group of burial mounds around a very large sub-structure mound with associated tetragon terraces.Mound building and burial : Large mound circular with flat top and a projecting ramp from the summit,but probably not with a spiral ramp and probably not with niches orsentry boxes at the cardinal points.The mound constructed of layers of sand and clay probably obtained fromthe ground adjacent and the lower building deposits sloped inward to-ward the center of the mound.The smaller mounds contained burials with grave offerings which accordingto White were : Jars, pipes, beads, breastplates, stone hammers, hatchets,arrowheads, etc.An oblong mound (dug by Rogan) had a burial of the bundle type face downand other bones on skull and over it a layer of red clay burned hardwith fragments of pottery and a clay pipe perhaps associated.Pottery : OccurrenceLamar Complicated Stamped Frequent.Lamar Plain Frequent.Lamar Bold Incised Few.Minority Lamar (?) types Few.Etowah Stamped Moderate.Savannah Checkstamped Few.(Early) Checkstamped Few.(Early) Simple Stamped Few.(Early) Cordwrapped Stick Decorated Few.Woodstock Diamond Stamped Few. pip.Na"6^]"''' REMBERT MOUNDS?CALDWELL 319Other artifacts:Sherd disks 7Pottery elbow pipes 2(Fired) clay wall plaster 3Bone tube 1Stone disk (small) 1Stone projectile point, triangular, slightly concave base, quartz 1Stone projectile point (probably pre-Lamar), large, stemmed, coarsechipping, weathered chert 1Stone chips, quartz, and weathered chert XShell bead, massive conch 1Shell bead, small disk, conch ( ?) 2Shell cut fragment, conch 1Mica, small unworked fragments 2Bitumen, small piece 1Mammalian remains : ' Deer (Odocoileus virginianus) :2 distal ends of humeri.1 Distal end of ulna.2 fragments of ulnae.1 distal epiphysis of femur.7 fragments of radii.6 fragments of ribs.1 fragment of scapula.5 vertebrae.1 fragment of pelvis,8 fragments of mandibles. .3 fragments of maxillae.1 fragment of antler.Raccoon (Procyon lotor) :1 mandible. CONCLUSIONSThe recent investigations at the Kembert site provided certain newdata which place this once important mound group into the currentpicture of southeastern archeology. The site was once the place of afair-sized town, apparently with permanent buildings of wattles wovenbetween uprights and plastered with clay. The most striking featuremust have been the imposing circular and flat-topped substructuremound, with a projecting ramp leading up to the summit, the largestmound Bartram had ever seen. Around this stood a group of burialtumuli and the extensive platforms which Bartram called tetragonterraces.The mounds and associated village were built and utilized duringthe Lamar period in Georgia, at some time between A. D. 1450-1650,probably by a Muskogean-speaking people.Local variations in the Lamar pottery at the Rembert site arestraighter sides and less rim flare on jars than are found at Lamar ? Identified by Henry W. Setzer, Division of Mammals, United States National Museum.. 320 BUREAU OF AMERICAN ETHNOLOGY [BuU. 154 sites in central Georgia or on the coast. There is also a much higherproportion of undecorated pottery and much less pottery decoratedby incising. We could discover no stratigraphic variation in the pot-tery of the Lamar types occurring at the site, although random sherdsindicated that the place had occasionally been occupied by earlierpeoples.The relatively few artifacts other than pottery conform in themain to the kinds of artifacts found at other Lamar sites. The em-ployment of bitumen and the use of conch shells for beads showconnections with the Coast.It is extremely doubtful if the large mound ever had the spiral rampwith niches at cardinal points which Bartram described. We are alsoinclined to discredit Jones' theory that the large mound was con-structed of clay brought from the other side of the Savannah River.At this writing much of the basal portion of the large mound isstill intact. In view of the neglect of this site in the past, it has beenrecommended that an excavation program be undertaken before 1952when the place will be covered by water.LITERATURE CITEDBabtram, William.n. d. Manuscript copy in the files of the Division of Archeology, UnitedStates National Museum, Washington. This was the MS. used bySquier.See also Van Doren, 1928.Caldwell, Joseph R.n. d. The Archeology of eastern Georgia and South Carolina.Caldwell, Joseph R., and McCann, Catherine.1941. Irene Mound Site, Chatham County, Ga. Univ. Georgia Press.Athens.Claflin, William H., Jr.1931. The Stalllngs Island Mound, Columbia County, Georgia. Pap. PeabodyMus. Amer. Arch., Harvard Univ., vol. 14. Cambridge.Fairbanks, Charles H.1946. The Macon earthlodge. Amer. Antiq., vol. 12, pp. 94-108.Jones, Charles C.1878. Ancient tumuli on the Savannah River, visited by William Bartram,in 1776. Ann. Rep. Smithsonian Inst, for 1877, pp. 283-286.Southeastern Archeological Conference.1939. Newsletter (mimeographed). Univ. Kentucky. Lexington.Thomas, Cyrus.1894. Report on the mound explorations of the Bureau of Ethnology. 12thAnn. Rep., Bur. Ethnol., 1890-91, pp. 3-730.Van Doren, Mark.1928. The travels of William Bartram, ed. by Mark Van Doren. New York.Watjchope, Robert.1948. The ceramic sequence in the Etowah Drainage, Northwest Georgia.Amer. Antiq., vol. 13, pp. 201-209.White, George.1849. Statistics of the State of Georgia. W. T. Williams Co., Savannah. i ^ t Z ^ -^^ o BUREAU OF AMERICAN ETHNOLOGY BULLETIN 154 PLATE 56 Q .^ap?a?'-?s^ ' ^ jt ^^^*^jiK v^A tm ?#**"JTt p v^^^G 'rffrrj 'UUlK Potsherd types. APPENDIXLIST OF RIVER BASIN SURVEYS REPORTS PUBLISHED IN OTHERSERIESBliss, Wesley L.1948. Palachacolas Town, Hampton County, South Carolina. Journ. Wash-ington Acad. Sci., vol. 38, No. 10, October 15.1949. Early Man in the Northwestern Plains. Proc. Fifth Plains Conferencefor Archeology, Note Book No. 1, Lab. Anthrop., Univ. Nebraska.1950. Birdshead cave, a stratified site in the Wind River Basin, Wyoming.Amer. Antiq., vol. 15, No. 3, January.Caldwell, Joseph R.1950. A preliminary report on excavations in the Allatoona Reservoir.Early Georgia, vol. 1, No. 1.Champe, John L.1949. White Cat Village. Amer. Antiq., vol. 14, No. 4, April. (Dr. Champe'sreport concerns excavations by the Laboratory of Anthropology ofthe University of Nebraska at the Harlan County Reservoir.)Cooper, Paul L.1949. Recent investigations in Fort Randall and Oahe Reservoirs, SouthDakota. Amer. Antiq., vol. 14, No. 4, April.Hewes, Gordon W.1949. Burial mounds in the Baldhill Area, North Dakota. Amer. Antiq., vol.14, No, 4, April. (Excavations by University of North Dakota andthe North Dakota Historical Society.)Holder, Preston, and Wike, Joyce.1949. The Frontier culture complex, a preliminary report on a prehistorichunter's camp in Southwestern Nebraska. Amer. Antiq., vol. 14,No. 4, April. (Excavations by the University of Nebraska StateMuseum.)Hughes, Jack T.1949. Investigations in Western South Dakota and Northeastern Wyo-ming. Amer. Antiq., vol. 14, No. 4, April.KrvETT, Marvin F.1949. Archeological investigations in Medicine Creek Reservoir, Nebraska.Amer. Antiq., vol. 14, No. 4, April.Miller, Carl F.1948. Early cultural manifestations exposed by the archeological surveyof the Buggs Island Reservoir in southern Virginia and northernNorth Carolina. Journ. Washington Acad. Sci., vol. 38, No. 12,December.1949. Appraisal of the archeological resources of the Buggs Island Reservoirin southern Virginia and northern North Carolina. Quart. Bull.,Arch. Soc. Virginia, vol. 4, No. 1, September.1949. The Lake Spring Site, Columbia County, Ga. Amer. Antiq., vol. 15,No. 1, July.1949. Early cultural manifestations exposed by the archaeological survey ofthe Buggs Island Reservoir in southern Virginia and northern NorthCarolina. Quart. Bull, Arch. Soc. Virginia, vol. 4, No. 2, December.1950. Early cultural horizons in the Southeastern United States. Amer.Antiq., vol. 15, No. 4, April. (A general article but containing datacollected during survey work.)953842?53 23 321 322 BUREAU OF AMERICAN ETHNOLOGY [Ball. 154Osborne, Douglas.1950. An archeological survey of the Benham Falls Reservoir, Oregon.Ainer. Autiq., vol. 16, No. 2, October.Seaks, William H.1950. Preliminary report on the excavation of an Etowah Valley site.Amer. Antiq., vol. 16, No. 2, October. (This report is on a projectof the University of Georgia at the Allatoona Reservoir. The Uni-versity assumed responsibility for that particular site.)Smith, Caklyle S.1949. Archeological investigations in Ellsworth and Rice Counties, Kans.Amer. Antiq., vol. 14, No. 4, April. (Dr. Smith's report concernsexcavations by the Museum of Natural History of the Universityof Kansas at Kanopolis Reservoir.)Solecki, Ralph S.1949. An archeological survey of two river basins in West Virginia. WestVirginia Hist., vol. 10, Nos. 3 and 4.Stephenson, Robert L.1947. Archeological survey of Whitney Basin. Bull. Texas Arch, andPaleont. Soc, vol. 18.1948. Archeological survey of McGee Bend Reservoir. Bull. Texas Arch.and Paleont. Soc, vol. 19.1949. Archeological survey of the Lavon and Garza-Little Elm Reservoirs.Bull. Texas Arch, and Paleont. Soc, vol. 20.1949. A note on some large pits in certain sites near Dallas, Texas, Amer.Antiq., vol. 15, No. 1, July.Wedel, Waldo R.1947. Prehistory and the Missouri Valley Development Program : Summaryreport on the Missouri River Basin Archeological Survey in 1946.Smithsonian Misc. Coll., vol. 107, No. 6, April.1948. Prehistory and the Missouri Valley Development Program: Summaryreport on the Missouri River Basin Archeological Survey in 1947.Smithsonian Misc. Coll., vol. Ill, No. 2, November.1949. Some provisional correlations in Missouri Basin Archeology. Amer.Antiq., vol. 14, No. 4, pt. 1, April.Wheat, Joe Ben1947. Archeological survey of the Addicks Basin: A preliminary report.Bull. Texas Arch, and Paleont. Soc, vol. 18. INDEX Abilene district, Texas, 220Abraders, sandstone, 37, 41, 42, 201, 227,229, 242Addicks Dam Basin, Archeological Sur-vey of (Wheat), 143-252Addicks Dam Basin, Texas, 151, 152,154, 161, 163 (map), 196, 238, 243(graph), 247-248 (table), 265discussion, 192-196environment, 152-157lithic complex, 196-233site 42/66A6-6, 166site 42/66A6-7, 166-167site 42/66A&-8, 1G7summary and conclusions, 238-246Addicks "Mound"' (42/66A6-1), 164(map) , 167, 184, 229, 230, 233, 236,237, 238, 240, 242, 243Agaronia testacea, 136Agate Basin, Wyo., 23, 51Agate, moss, 276Aqiatrodon mokasen, 157Akokisa Indians, 158, 159, 160, 161, 162,234, 245, 246, 258Alabama Polytechnic Institute, 289Albuquerque, N. Mex., 276Alfalfa, 89AUen site (25FT50), Medicine CreekReservoir, 49, 50, 51Alligators, 235, 237"Alluvial Chronology," 293, 294(graph), 297, 298,301Alma, Nebr., 40, 92Alma Ossuary (25HN2), 111, 129-130(table), 131, 185Amherst, Nebr., 43Amherst Reservoir, Nebr., 132Amphibians, 42An Archeological Survey of the Ad-dicks Dam Basin, Southeast Tex-as (Wheat), 143-252Andropogon fuscatus, 156Andropogon glomeratus, 156Andropogon haroides, 156Angostura, S. Dak., 3, 9, 11, 12, 29, 31,33, 49, 51, 52, 56, 68Angostura Reservoir site, S. Dak., 20,21-24, 31, 32, 51, 54, 68, 69, 71,74-80, 98, 99, 100Area A, 77, 78Area B, 77, 78Area C, 77, 78description of, 21, 74site 39FA10, 75site 39FA23, 76site 39FA30, 76, 100 Angostura Reservoir site?Continuedsite 39FA35, 78site 39FA38, 78site 39FA42, 78site 39FA45, 78site 39FA48, 78site 39FA56, 75site 39FA61, 77site 39FA65, 50, 77, 98site 39FA68, 75, 76site 39FA78, 78site 39FA79, 78Animals, furbearing, 79game, 79hunted, 235-236 (table)Antelope, 42, 87, 110, 234, 235, 236, 237,280Antelope County, Nebr., 43Anthony, John W., 149Anthropomorphic figures, 78Antilocapra sp., 235, 236Arapaho Indians, 23, 37, 84Arch Hurley Conservancy District, 290Arikara Indians, 55, 58, 83Caddoan-speaking, 26Arizona Bureau of Mines, 149Arizona State Museum, 150Arkansas River, 54, 111Arrowheads, 314Arrow points, 198triangular, 22Arroyo cutting, 298, 300, 301Artifacts, 40, 42, 45, 48, 55, 75, 77, 93,111, 118, 130, 274, 291, 297, 317-318, 320antler. 111, 118, 119-121, 231(table), 232, 233bone, 42, 83, 95, 111, 112, 114, 116,118, 119-121, 127, 196, 230(table), 231-232, 243clay, 190flint, 80, 95, 171, 227ground stone, 240iron, 95Plainview type, 77pottery, 25, 41, 95, 184-196shell, 25, 83, 111, 112, 114, 118, 123,127, 129 (table), 196, 231 (table),233stone, 30, 72, 75, 82, 88, 89, 95, 99,111, 114, 118, 121-123, 127, 129(table), 174, 196, 198-201 (table),220-228, 282stratigraphy of, 233Artiodactyls, fossil, 32, 43 323 324 BUREAU OF AMERICAN ETHNOLOGYAsh, 1, 9, 26, 38, 84, 109, 110blue-foliaged {Fraxinus amerircana), 110Ash Creek, Kans., 37Ash Hollow Cave site, 131Ash Hollow fauna (Middle Pliocene),43Ashland, Nebr., 43Ash lenses, 36Aspen, 72, 86Asphalt, use of, 196Atacapa Indians, 264, 265Atacapan-speaking Indians, 258Atlatl weights, stone, 201, 227, 229, 230,240, 241Attakapa Indians, 160, 161, 196, 245,246Augusta, Ga., 307Austin, Tex., 276Avoyel Indians, 162, 246Awls, 127bone, 15, 22, 41, 42, 73, 120, 231,233, 243horse-bone, 73splinter, 37Axhead, iron, 92Badgers, 235-236 (table), 237Bad River, S. Dak., 74, 80Baldhill, N. Dak., 10, 47, 49Baldhill Creek, N. Dak., 47Baldhill Reservoir site, investigations,47-48, 53, 67Barker Dam, Tex., 151, 163 (map), 166Barnes, Donald K., 68Barranca Creek, N. Mex., 271Barreyro, Alvarez, 159Barrios, Governor, 160Bartram, William, Jr., 307, 309, 310, 311,314, 319, 320Basketmaker people, 277Basketry fragments, 83Batson, A. A., 13Battle Mountain, S. Dak., 23Bauxar, J. J., 8, 12, 19, 67, 73Bazterra, Orobio, 161, 162Beads, 128, 314antler, 233bone, 128clay, 48cylindrical copper, 48dentalium shell, 80disk, 15, 123glass trade, 30, 83making of, 123-124shell, 126, 131, 136, 318, 320shell disk, 52, 112, 113, 114, 115,116, 117, 118, 123, 127, 130, 136,318tubular bone, 15, 22, 41, 52, 120, 136tubular shell, 25wooden, 234Beans, 14, 16, 57, 110 Bear, black (Ursus americanus) , 157,162, 234Bear Creek, Tex., 151, 155, 163, 174Beaumont formation. Pleistocene age,154Beaver, 41, 110Beaver City, Nebr., 10Beaver Creek, Boone County, Nebr., 45Beaver Creek, Wyo., 29, 30Beaver skull, fossil, 94Belle Fourche-Cheyenne drainage, 85Belle Fourche River, Wyo., 30, 31Bellisle, Simars de, 159, 160, 161, 162,234, 237Beranger, Jean, 159, 161Berries, wild, 79Bidai Indians, 160, 161, 162Big Horn Mountains, Mont, and Wyo.,84Big Horn River, Wyo., 31, 84Bioto Province, Tex., 156-157Bird bones, use of, 120Birds, 42, 157, 235, 237, 280fossil, 43Birdshead Cave, Boysen, Wyo., 10Bismarck, N. Dak., 26, 27Bison, 16, 17, 42, 50, 110, 162, 234, 235-236 (table), 237, 277extinct, 99fossil, 93hunters, 79kills, 20, 23, 51skeletons, 48skulls, 94Bison antiquiis, 50sp., 235, 236taylori, 50Bitter Creek, Moorhead Reservoir area,85Bitternut {Sickoria cordiformis) , 156Bitumen, 318, 320Bivalves, freshwater, 112, 124, 125Bixby Reservoir area, S. Dak., 68, 70,71, 74, 80investigations, 80site 39PE10, 80Black Hills, S. Dak. and Wyo., 21, 23,24, 26, 29, 30, 52, 55, 75, 79, 80, 81Black, W. M. (Marshall Black), 149,165, 184, 229, 234, 236, 237, 238Blades, 51, 85, 94, 174chipped stone, 30, 31, 42Blanepain, Joseph, 160Bliss, Wesley L., 8, 12, 67Blue River drainage, 127Bluff Creek, Kans., 35Bobcat (Lynx riifus texanns) , 157Bolton, Herbert E., 161Bone industries, 57, 119-121Bones, animal, 36, 41, 42, 51, 69, 76, 77,78, SO, 82, 86, 87, 90, 93, 94, 99, 165,236-238, 280, 310, 313, 315 INDEX 325 Bones, animal?Continuedbear, 87bison, 19, 21, 41, 73, 87, 280deer, 117, 120, 128, 171, 231human, 112, 114, 115, 117, 127, 130,237, 314Bonny, Colo., 10Bos sp., 235Bottles, pottery, 196Bow and arrows, 17Bowls, wide-mouthed, 194, 239, 31bBox elder, 38, 72, 80, 84, 109Boysen Reservoir area. Wye, 31, 49, b?Boysen, Wyo., 3, 10, 11, 12, 33Brass, surface, 41Brays Bayou, Tex., 155Brazos River, Tex., 154, 161Breast-plates, 314Broad River, Ga., 307Broom sedge {Andropogon glomeratus) , 158Bryan, Kirk, 289Buffalo, 110 ^^ ^^^ ^__Buffalo Bayou, Tex., 151, 154, 155, 15b,158, 161, 163, 164, 166, 238Buffalo Creek, Nebr., 10, 43Buffalo Pasture (39ST6) , "S. Dak., 28, 29Buildings, wattle and daub, 317, 319Bureau of American Ethnology, 27, b5Bureau of Reclamation, 2, 3, 4, 12, 13,43, 65, 66, 67, 71, 72, 73, 74, 80, 81,82, 84, 88, 89, 90, 93, 94Burial cairns, 35, 36, 37mounds, 20, 24, 47, 48, 53, 319pits, 22, 47, 239, 242Burials, 15, 17, 18, 36, 73, 129 (table),172 (fig.) , 173, 180 (figs.) , 238, 242extended, 129 (table)flexed, 38. 77, 129 (table), 171, 178,181, 182, 239, 242primary. 111 ^??secondary, 25, 52, 116, 117, 127semiflexed, 116, 166, 171, 178, 181,238, 239, 242Burial si^e^!, 8^Doering, 258-261 ,,,??,,Harlan County, Nebr., (14PH4),39Kobs, 261-264Business Council of the Shoshones andArapahoes, 84Busycon columella, 136contrarium 125, 126sp., 128Cache pits, 40, 54Cactus, 81Caddo Indians, 162, 246Cairo, Nebr., 43Calcite crystals, 118 ,_^ ? . .Caldwell, Joseph R. (The RembertMounds, Elbert County, Georgia) , 303-320Camel, Pleistocene, 15, 94Campbell, Thomas N., 149, 197 Camp sites, 22Canadian River, N. Mex., 271, 289, 290,292, 293Canadian Valley, N. Mex., 290, 292, 300(table)Caney Creek, Tex., 154Canis, sp., 235, 236Cannibalism, 162, 237, 238Cannonball Reservoir area, N. Dak., b?,70, 71, 72-73investigations, 72-73Cannonball River, N. Dak., 72Canyon Ferry Reservoir site, Mont., 10,11, 12, 32, 33, 49, 90Caplen Mound, Tex., 234, 265Capoque Indians, 160cSlsbad, N. Mex., 274, 275, 277, 279,281, 282, 283Carnegie Museum, Pittsburgh, Pa., 32Carnivores, fossil, 32, 43, 94Cartersville, Ga., 310Carunculina texasensis, 23oCedar, red {Jitniperus americana) , dWCedar Bluff Reservoir site, Kans., 11,12, 33, 49, 88investigations, 88-89Celt,'chipped:?7!'5?i27, 200, 221, 225,229Central Plains sites, 53, 57, 58, 119, 121,122, 126, 128, 131, 137Ceramics, 281See also Pottery.Ceramic stratigraphy, 190-192Chalcedonic flakes, 121Chalcedony, 274, 275, 277Champe, John L., 38, 41, 44, 91, 107Charcoal, 36, 76, 77, 78, 82, 90, 111, 112,117, 120, 175, 238, 310Cherokee Indians, 314, 315Chert chips, 35, 318Chert cores, 30Chert, mottled, 274, 275, 277 ^? ^^ _ .Cheyenne Indians, 23, 24, 26, 37, 79, 111Cheyenne River, S. Dak., 20, 21, 27, 28,32,51,74,76,78Cheyenne River Basin, Wyo., and b.Dak., 10, 11, 24, 29, 52, 79, 81Chimney Butte, S. Dak., 82Chinle formation, 290, 2918Ser86;2li?bIe),225,229,240.243, 278, 283bone, 41chipped stone, 200Cimarron River, N. Mex., 277, 281, 283Valley, 277Cist, rectanguloid, 36Clams, 157, 165, 235Clark, Dean, 8, 68Clark Hill Reservoir, 307Clear Creek, Kans., 35, 36Clear Creek, Wyo., 85Clear Fork, Tex., 23, 154 326 BUREAU OF AMERICAN ETHNOLOGYClearwater Ci-eek, Antelope County,Nebr., 46Cleai'water, Nebr., 46Clearwater Reservoir, see RosedaleReservoir.Climate, 155-156Cocos Indians, 158, 160, 161Colorado River, Tex., 154, 160, 161Columbia University, 27Columbus, Nebr., 43Comanche Indians, 23, 55Committee for Recovery of Archeologi-cal Remains, 2, 66Conboy, Neal, 23Conchos River, Mexico, 274Condylarths, fossil, 32Cooper, Paul L., 8, 12, 19, 20, 24, 68, 70,73, 107Copperhead (AgJdstrodon moJcasen),157Corn, 16, 89, 109, 113, 115charred, 14Corn-growing people, 24Corps of Engineers, 2, 4, 20, 24, 34, 37,41, 47, 65, 66, 97, 149, 151Corpus Christi Bay, Tex., 274Cottonwood, 19, 26, 35, 38, 72, 80, 84, 86,109, 110Cottonwood Creek, Wyo., 31, 32Coyote, 42, 110Creeks, 309, 314Crotalus Jiorridus, 157Culbertson, Nebr., 10Cummings, Robert B., Jr., 8, 68, 107, 137Cushing, Nebr., 44, 49Cushing Reservoir site, investigations,44-45Dakota Sioux Indians, 26Dart points, 198Davis Creek Reservoir, Nebr., 4, 10, 68,70, 132, 133Davis, E. Mott, 91, 93Deadoses Indians, 160, 161Deer, 17, 41, 42, 87, 100, 234, 235-236(table), 237, 319antlers, 94, 119, 127Virginia (Odocoileus virginianustexanus), 157Deer Creek, Wyo., 31Dentalium shell, 80Department of the Army, 65Department of the Interior, 65Devil Dive Creek, Wyo., 86Devils Lake, N. Dak., 53Dick, Herbert W., 289, 299(Two Rock Shelters near Tucum-cari. New Mexico), 267-284Dickinson, N. Dak., 73Dickinson Reservoir site, N. Dak., 7071, 72investigations, 73-74Didelphis sp., 235Didelphodus sp., skull of, 32Didymictus sp., skull of, 32 Digging sticks, 117Digging tools, chipped-stone, 48Disks, chipped, 37copper-covered wooden. 111sherd, 317stone, 317Dismal River culture, 55, 57, 92, 93, 100.Ill, 132people, 22, 46site (25H021), 95sites, 39, 51, 55, 95, 96Doering sites, 190, 191, 194, 214, 215(graph), 218, 219, 229, 233, 236,237, 238, 240, 241, 243, 244, 245Site 42/66A6-2, 167, 168 (map), 169(map), 170 (map), 172 (fig.),178site 42/66A6-4, 174, 175 (map), 191(table)site 42/G6A6-5, 173-174skeletal remains from, 257-266Dogs, 87, 236, 237domestic, 56, 69Donohoe, John C, 8, 31Drainage system, Addicks Dam Basin.154Drills, 41, 42, 73, 77, 82, 86, 221 (table),225-226, 229, 230, 240, 242, 243,278, 283double-end, 226expanding base, 225-226stemmed, 226stone, 72, 200 (table)T-shaped, 37Dry Creek, S. Dak., 74, 78, 85Dubose, Dude, 308Dunn site (25FR2), 129-130 (table)Eagle Creek, Nebr., 53, 137Eden Valley Yuma horizon, 50Edgemont Reservoir site, Wyo.. 29-3085Edgemont, Wyo., 10, 31, 49, 56Edwards, Guy D., 2Edwards Plateau, Tex., 220Elberton, Ga., 307Elkhorn River, 43, 46Ellsworth, Kans., 34, 37Elm, 26, 35, 38, 109Elm Creek, Kans, 35, 36site 14EW12, 37El Orcoquisac village, Tex., 160El Paso, Mex., 274Equus complicatns, 235Ericson Reservoir site, Harlan County,Nebr., 21, 43, 46Erosion intervals, 300 (table)Ethnohistory, 157-162Etowah Valley, Ga., 310, 317European contacts, 58Fairbanks, Charles H., 315Federal water-control progi.\'im, 59, 65,71, 88 INDEX 327Field work and explorations, 11-33by cooperating agencies, 33-48in paleontology, 31-33Figurines, 242Fir, 86Fire basins, rock-lined, 22, 51beds, 310pits, stone-filled, 22, 40, 51, 238Fireplaces, 15, 17, 19, 22, 23, 30, 40, 50,51, 76, 77, 95, 238, 314circular, 77Fish, 157, 237fossil, 33, 43, 93Fishing, 17, 57, 234Flag Creek site 25HN4, 129-130 (table)Fleshers, 41Flint, 162, 276, 277chips, 36, 88, 93"Flint Hill," Hell Canyon, S. Dak., 23Folsom complex, 50, 56Fort Bennett, S. Dak., 8, 27Fort Pierre, S. Dak., 27Fort Randall Dam, S. Dak., 12, 20, 24,49, 53, 54, 67, 97Fort Eandall Reservoir, investigations,24-25site 39CH4, 53site 39CH9, 53Fort Sully village (39 SL4), S. Dak., 28Fort Yates, N. Dak., 27, 28Fossils, 81, 85Cretaceous, 93Lower Eocene, 32Miocene, 32Oligocene, 32, 81Pleistocene, 93Pliocene, 43, 93, 94Upper Cretaceous, 33Franke, Carol, 107Frankforter, W, D., 38, 91Franklin, Nebr., 38Fraser, Dorothy E., 8, 20Fraxinus americana, 110Freemont, John C, quotations from,110Fruits, wild, 79, 109Galenkamp, Charles, 149Galveston Bay, Tex., 153, 155, 158-161,184, 195early exploration of, 158-160historic Indians around, 160-161Galveston Bay Indians, culture of,161-164Galveston Island, Tex., 158Garay, Francisco de, 158Gar (fish), 235, 237Garrison, N. Dak., 10, 67Garrison Dam, N. Dak., 97Gatherers, 24Gathering, 234, 240Genoa, Nebr., 45Geology, 154 Geology of the Hodges Site, Quay Coun-ty, New Mexico (Judson), 285-302Olebula rotundata, 235Glendo, Wyo., 10Glen Elder Reservoir, Kans., 88, 89investigations, 89-90site 14ML1, 89site 14ML8, 89Glen Elder (town), Kans., 89Goldstein, Marcus S., 257, 264, 265, 266Gosper County, Nebr., 43Gouges, 23Graham Ossuary, Nebr., 126Graham site 25HN5, 111, 112, 118, 128,129-130 (table), 131, 135Grand River, S. Dak., 26, 74, 83Grapes, wild, 109Graphite, 75Grass, 35, 72, 73, 76, 80, 81, 85, 86, 87,109, 113marsh, 157switch (Paniciim virgatum), 156water (Paspahim sp.), 157Gravers, 82, 200, 221, 227, 229, 230, 240,243Great Bend culture, 37, 54Great Plains region, 43, 49, 52, 54, 77Greens Bayou, Tex., 155Grifl5n, James B., 149, 190, 194Griflath, D. W., 149Grinding implements, 22, 42Grinding slabs, sandstone, 37Grisbee site (42/66A3-1), 174, 182(map), 183 (fig.), 191 (table),219, 238, 244Guadalupe Mountains, N. Mex., 277Guide Rock, Webster County, Nebr.,39 137site (25WT3), 118, 129-130 (table)Gunnerson, D., 43Gunnerson, J. H., 43Guns, figures of, 78, 79Gypsum, 33Hackberry, 35Hall, Robert L., 8, 19, 20Hammers, stone, 314Hammerstone, 83, 86, 136, 228, 229, 243,279, 283Handstones, 201, 221, 228, 230, 240, 242Han Indians, 160Harlan County Dam, description of, 38Harlan County Reservoir site, Wyo., 10,21, 38-43, 49, 54, 55, 67, 68, 69,91, 92, 95, 100investigations, 92-93, 98, 114, 132site 25HN11, 92site 25HN37, 92, 95Harpe, Barnard de la, 159, 161Harvard University, 289Hasinai Indians, 158, 159Hatchets, 314Hat Creek, Nebr., 23 328 BUREAU OF AMERICAN ETHNOLOGYHaiiry, Emil W., 149Hay cultivation, 39Heart Butte, N. Dak., 3, 8, 11, 49Heart Butte Reservoir site, investiga-tions, 19-20Heart River, N. Dalf., 19, 26, 27, 73Hearths, 51, 75, 78, 77, 78, 85, 86, 93, 94,99basin-shaped, 36, 75rocli, 30, 31, 51Heclier, Thad G., 19Helena, Mont., 90neliosonia trirolvis, 235Hell Canyon, S. Dak., 23Hell Creek, Kans., 90Henrietta, Tex., 274Hewes, Gordon W., 19, 47, 48, 53, 96, 97Hickoria cordiformis, 156High Plains area, Nebr., 55Hill, A. T., 38, 107, 111, 130Hodges site, N. Mex., 271Area A, 272, 273 (fi?r.), 274, 275,276, 277, 278, 279, 280, 281, 291Area B, 272 (flg.), 275, 277, 278,279, 280, 281, 291conclusions, 282-283, 297-301excavation and stratigraphy, 272(flg.), 273 (fig.), 274, 291 (fig.)faunal material, 280-281geology of, 285-302physical setting, 289, 290 (map),291 (fig.)sequence at, 294-295, 296 (fig.), 297,299 (table), 300 (table)Hoe, bison scapula, 16, 41bone, 17shell, 37Holden, W. C., 274Holder, Mr. and Mrs. Preston, 38, 42Holdredge burial site 5 (25FR9), 112,129-130 (table)Hollywood, Ga., 315Homestead National Monument, Nebr.,12Hopewellian sites, 37Horn industries, 57Horrell, John, 107Horse, acquisition of, 53, 56fossil, 94native, 50, 235, 237Horsehead Creek, Angostura, S. Dak.,11, 21, 22, 31, 32, 74, 76, 77, 100Horsepen Creek, Tex., 151, 155, 174, 182Horse Thief Canyon, Kans., 35Horticultural peoples, 39, 54, 58, 100, 110Hot Springs, S. Dak., 20, 21, 32, 74House floors, 41pits, 28, 29sites, 40, 55, 76Houses, grass, 57jacal type, 282Housing, 162Houston, Tex., 151, 155, 156, 163Huddleston, T. E., 2Hudson Farm (14EW24), Kansas, 36 Hughes, Jack T., 8, 11, 20, 29, 32, 67, 68,70, 71, 74Hunters, 24, 58Hunting, 16, 57, 162, 234, 240, 277camps, 20, 21, 54, 55Hunting and gathering peoples, 39, 52,56, 57, 75, 85, 91, 99, 162, 245, 282,283Huron Indians, 118Hut rings, 27Implements, bone, 117chipped stone, 83, 200 (table)Indian Hill (14EW1), Kans., 35Indian Knoll, 220Indian Skeletal Remains from theDoering and Kobs Sites, AddicksReservoir, Texas, (Newman),25.3-266Indian-White contacts, 100Insectivores, fossil, 32, 43Irene, Ga., 310, 315Iron, surface, 41Isla Malhado, Tex., 160Isto, Neil, 68, 72, 74, 84Jackson Narrows, S. Dak., 74, 78Jar, elongated globular, 195, 289pottery, 178, 194, 196, 314, 316Jasper, 121, 122, 123Jemez Mountain area, 27SJennings, J. D., 2, 281, 282 .Johnson 'Siding Reservoir site, 20Johnson Siding, S. Dak., 10, 49Jones, Charles C, Jr., 307, 309-310, 311,314, 320Judson, Sheldon, 271, 280, 289(Geology of the Hodges Site, QuayCounty, New Mexico) , 285-302Judson's Modern Sand, 274No. 2/3 Sand, 272, 274Juniper, 19, 72, 85, 86Juniperus amcricana, 309Kanopolis Reservoir area, Kans., 34, 35,37, 49, 52, 54site 14EW6, 54Kansas River Basin, 54, 58Kansas, sites investigated, 34-37, 88-90Karankawa Indians, 160, 161, 162, 195,196, 246, 285Kay, J. Leroy, 32Keith Focus, Woodland culture desig-nation, 16, 52, 136Kelley, J. Charles, 149, 197, 202, 219Kelly, A. R., 315Keyhole Reservoir site, Wyo., 10, 29,30-31, 49, 56, 85Kidder, A. V., 281Kiowa Indians, 23Kivett, Marvin F.. 8, 11, 12, 13, 16, 34,39, 40, 47, 52, 07, 70, 91(The Woodruff Ossuary, a Pre-historic Burial Site in PhillipsCounty, Kansas), 103-130Kleinsasser, Glenn, 68 INDEX 329Knife fragments, 72, 82Knife River, S. Dak., 27Knives, 73, 76, 77, 83, 88, 94, 200 (table) , 221 (table),229, 243, 275beveled, 223, 230, 240bifacial, 277chalcedony plate, 23chipped, 22, 30convexoconvex, 277flake, 37, 277, 278, 283flint, 85, 277leaf-shaped, 222. 277long narrow, 223ovoid, 223planoconvex, 277prismatic, 89refined, 277-278, 283stemmed, 220, 221, 222, 230, 239triangular, 222Kobs site (42/66A6-3), 174, 176 (map),177 (map), 178, 179 (figs.), 180(figs.), 181, 182, 190, 191 (table),193, 194, 215 (graph), 218, 219,229, 233, 236, 237, 238, 240, 244,245, 265skeletal remains from, 257-266Krieger, Alex D., 149, 197, 219Lake McConaughy, Nebr., 43Lamar site, Ga., 315, 317, 319, 320Lampsilis ventricosa ocoidens, 136Lances, feathered, 78Lander, Wyo., 86Langham Creek, Tex., 151, 155, 163, 174,176, 182Lasmlgona complanata, 136Lav^rence, Barbara, 280Leavenworth (39C09), S. Dak., 28Lemmon, S. Dak., 83Leon, Alonso de, Jr., 159L'^pidosteus sp., 235Leuty, William O., 34Lime Creek, Nebr., 15, 38, 41, 42, 50, 93,99site 25FT41, 50, 93, 99site 25FT50, 50, 93Lincoln Municipal Airport, cooperationof, 8, 68Lincoln Office, River Basin Surveys, 34,66, 67, 70Lindenmeier Folsom horizon, 50Lissie Formation, Pleistocene age, 154Literature cited, 141. 248-252, 266, 283-284, 301-302, 320Lithie artifacts, 220-230stratigraphy and discussion, 228-230Lithie complex, 196-233discussion, 218-220Little Arkansas River, Kans., 34Little Knife River, N. Dak., 96site 32MN9, 96Little Missouri River, 55Little Popo Agie River, Wye, 86 Little Wind River, Wyo., 86Llano Estacado (Staked Plains), N.Mex., 271, 292, 294Lodges, earth, 17, 37, 39, 53, 57Lodge sites, 22Loess Plains area, Nebr., 43, 109Logan Museum of Beloit College, 27Long site (39FA65), Angostura Reser-voir, 49, 51, 77, 99Loretto, Boone County, Nebr., Reser-voir site. Investigations, 44, 45, 49Loseke Creek, Nebr., 53, 137Lost Cabin faunal zone, 32Loup River, 43, 44, 45Lower Cheyenne Village (39ST1), S.Dak., 28Lower Mississippi Valley culture, 195,220Lower Platte Basin, Nebr., 38, 43, 54archeological reconnaissance, 43-46Ludlow Cave, S. Dak., 83Lundelius, Ernest L., 8, 31Lusk, Wyo., 51Lynx rufus texanus, 157Maize, 56, 57, 110culture, 16, 282Maize-bean-squash horticulture, 53Maize-growing peoples, 53Malouf, Carling, 90Mammalian remains, 319Mammals, fossils, 32, 93JMammoth, Columbian, 50Mandan Indians, 20, 55, 58, 83Mandibles, badger, 128rabbit, 128Mankato substage of Wisconsin glacia-tion, 50:^.Ianos, stone, 30, 36, 51, 76, 77, 78, 79,80, 82, 86, 279subrectangular, 23Maple, ash-leaved, 110Marcasite, 33Marshall Ossuary, Nebr., 126Marshall site (25HN1), 112, 118, 127,129-130 (table), 135Matagorda Bay, Tex., 158filaterial culture, 184-248Maveye Indians, 161McAllister, J. G., 149McCann, Franklin T., 289McCook, Nebr., 93McKenzie, Gordon F., 8, 20Mealing slabs, 22Medicine Creek Dam, description of, 13,14, 52r.Iedicine Creek, Nebr., 3, 9, 11, 12, 14,15, 38, 42, 48, 51, 54, 68, 69, 93,94, 98, 132Medicine Creek Reservoir site, investi-gations, 12-18, 38, 41-43, 91, 93-94 99site '25FT18, 52site 25FT40, 94 330 BUREAU OF AMERICAN ETHNOLOGYMedicine Creek Reservoir site?Con.site 25FT41, 93, 94site 25FT42, 93, 94site 25FT49, 94site 25FT50, 49, 93, 94site 25FT70, 52Mera, H. P., 281, 282Metates, 51, 76, 77, 279-280Metcalf, George, 8, 12, 13, 68, 71Mexican Creek Valley, S. Dak., 81Mica, 310, 318Micrurus fulvius fulvius, 157Middens, 41, 54Middle Loup River, Nebr., 94Miller, Carl F., 307Milling pits, 291Milling stones, 201, 221, 228, 230, 240,242Minerals, miscellaneous, 228Mink, 235, 237Missions, founding of, 160San Francisco, Tex., 160San Ildefonso, Tex., 160Missouri Basin Indians, 9Missouri River, 9, 25, 26, 27, 32, 53, 55,56, 58, 59, 65, 68, 69, 72, 74, 80,81, 92, 96, 97, 101Missouri River Basin Survey, 2, 3, 4,11, 12, 34, 65, 66, 67, 70, 74, 90. 97cooperating agencies, 87-97field work and explorations, 70-87laboratory activities, 8-11, 68-70personnel, 1948, (list) 8, 67-68resume and conclusions, 48-59,97-101sites investigated, 5-7 (table)Mitchell Creek, Nebr., 94Mobridge, S. Dak., 27, 28Mollusca, 69Montana, sites investigated, 90-91Montana State College, 31Montana State University, 90, 91Moorcroft, Wyo., 30Moorhead, Montana-Wyoming border,56, 84Moorhead Reservoir area, Montana-Wyoming border, 70, 71, 84investigations, 84-85Moreau River, S. Dak., 74, 80Morhiss site, Texas Gulf Coast, 220Morrison, J. P. E., 107, 125, 149Morse, Stanley, 165Mortar, bedrock, 279, 280, 283Moscoso, 158Mule Creek, Wyo., 81Mullen, Nebr., 43, 94, 95, 98, 100Mullen Reservoir site, Nebr., 21, 67,68, 91investigations, 94-96Museum Comparative Zoology, HarvardUniversity, 280Muskogean-speaking peoples, 309, 319Mussels, fresh-water, 57, 274, 281 Mutch, C. L., 13Nacatoches settlement, Tex., 159Narvaez Expedition, explorations of,158National Park Service, 10, 12, 34, 65, 66,67, 69, 71, 87Chief Historian's Office, 66Nebraska, sites investigated, 12-18, 38-43, 91-96Nebraska State Archeological Survey,111Nebraska State Historical Society, 4, 9,10, 12, 38, 39, 41, 49, 67, 69, 91,93, 95, 107, 111, 112, 131, 133Neches River, Tex., 154, 161Needles, bone, 42Newcastle, Wyo., 29, 51Newman, Marshall T. (Indian Skele-tal Remains from the Doeringand Kobs Sites, Addicks Reser-voir, Texas), 253-266Niobrara chalk. Cretaceous depositsfrom, 93Niobrai'a River, 55, 58Nixon, A. E., 8, 68Nonhorticultural peoples, 53Nonpottery-making peoples, 45, 51, 79,80, 82, 85North Dakota Historical Society, 19,27, 47, 49North Dakota, sites investigated, 19-20, 46-48, 71-74, 96-97North Fork, Bad River, S. Dak., 80, 81Grand River, S. Dak., 82, 83Little Wind River, Wyo., 86Popo Agie River, Wyo., 86Solomon River, Kans., 109, 110North Loup River, Nebr., 91Norton, Kans., 10Nuestra Senora de la Candelario Mis-sion, Tex., 160Nuestra Senora de la Luz Mission, 160,246Nunez, Cabeza de Vaca, Alvar, 158, 161,162Oahe Reservoir site, S. Dak., 11, 20, 25-29, 49, 67, 68, 69, 97description of, 26, 28Oak, 26, 156, 174yellow (Quercus velutina), 156Ohovaria oUvaria, 136Obsidian, 18, 30, 31, 75. 275, 277, 278Occupational level, 1, 42Occupational sites, 35, 83, 94Occupational zone C, 93Odocoileus virginianus, 319sp. 235, 236virginianns tcwanus, 157Ogallala, Keith County, Nebr., 43Olivella nivea, 126, 106sp., 126, 136Olson, James C, 107 INDEX 331Onion Flat Reservoir site, Wyo., 68, 70,71, 84, 85investigations, 85-86Opossums, 110, 235 (table), 237Oregon Basin, Wyo., 10Orleans, Nebr., 38, 137site 25HN3, 129-130 (table)Ornaments, mussel-shell, 48, 117shell, 118Orobio Bazterra, Joaquin de, 159, 162Oso Mound, Nueces County, Tex., 264,265Oto Indians, 111Oyster Creek, Tex., 154Pactola Reservoir site, S. Dak., 10, 20,49Paint pigments, 201, 221, 228Paleo-Indians, 293Paleontological surveys, 73, 74, 80, 81,82, 83, 84, 85, 86, 87Pamcum virgatum, 156Paradise Creek, Kans., 90Paspaluni sp., 157Paste, manufacture of, 184-185, 196Pastures, 39Patiri Indians, 161Pawnee Indians, 16, 17, 39, 40, 58, 110Pearls, fresh-water, 112Pecos Pueblo, 278Pecos River, N. Mex., 275, 276, 277, 279,281Penasco Blanco River, 281, 282Pendants, biperforate shell, 37oval, 15shell, 52, 112, 117, 124, 125, 128, 283,281Perissodactyls, fossil, 32, 43Personnel, list of, 8Pestles, 279, 280Petroglyphs, .35, 37, 78, 87, 90recent, 78Pheasants, 110Philip Reservoir site, S. Dak., 70, 71,74,80investigations, 80-81Philip, S. Dak., 81Pickstown, S. Dak., 24Pickwick Basin, 220, 230Pierre, S. Dak., 9, 20, 25, 26, 27, 28, 97Pierre shale, 81Pike, Zebulon M., IllPine, 87, 156, 174lodgepolo. 86short-leaf (Pinus echinata), 156yellow, 30, 85Pineda, Alonzo de, 158Pine Ridge Indian Reservation, S. Dak.,81Pinus echinata, 156Pioneer, Colo., 10Pipe, clay, 314, 317Pipestone, 73Plains Apache Indians, 22, 55, 58, 92,95, 100, 101, 111 Plainview, Tex., 93, 99Plants, domestic, 56Platte River watershed, Nebr., 43, 45Plaza Larga Creek, N. Mex., 271, 272,289, 290, 294, 296, 297, 298Pleistocene sequence, 292-294, 298Plum Creek Reservoir site, GosperCounty, Nebr., investigations, 45-46,49Plums, wild, 109Points, Folsom, 96obliquely flaked, 93Plainview, 93, 99Pomme de Terre Reservoir site, 68Popo Agie River, Wyo., 86.Post molds, 15, 17, 40, 95Post-White era, 35Potatoes, wild, 162, 234Potsherds, 21, 36, 37, 41, 44, 45, 46, 53, 76,77, 78, 83, 89, 111, 118, 128, 174,176, 183, 184, 308, 313, 314, 315,316, 317Pottery 17, 21, 57, 72, 75, 91, 96, 118-119,128, 129-130 (table) , 162, 171, 178,196, 246, 281-282, 310, 314, 315-317, 318-319 (list)black-and-white ware, 282brown ware, 276, 281, 282, 283calcite-tempered, 15, 37, 52, 118, 119,132, 184, 239, 242Chupadera, 274, 275, 282, 283coiled, 194colors, 184, 185, 239, 316cord-roughened, 15, 19, 21, 52, 53,54, 76, 77, 95, 100, 119, 128, 317decorated, 44, 45, 53, 189, 195, 239,316dentate-stamped, 37discussion, 192-196distribution, 189El Paso polychrome. 282Etowah Stamped, 317fabric marked, 25Geneseo Plain, 36Geneseo Simple Stamped, 36Goose Creek Incised, 189, 191, 195,239Goose Creek Plain, 184-189, 239gray to bufE, 80, 119, 316grit-tempered, 25, 37, 44, 46, 80, 95Harlan cord-roughened, 131, 132,133, 134, 135Hopewell, 37incised, 21, 189, 316Irene Incised, 316Lamar Bold Incised, 315, 316Lamar Complicated Stamped, 308,309, 315, 316, 317Lamar Plain, 315, 316, 320"Mandan-Hidatsa type," 83Mandeville Plain, 194manufacture of, 184-185mica-tempered, 41Mississippi complex, 394 332 BUREAU OF AMERICAN ETHNOLOGYPottery?ContinuedNoi-thei-n Woodland, 195paddle stamped, 195, 316Plain Red, 274prehistoric, 96protohistoric, 96puebloan, 18, 41, 276quartz-tempered, 316red ware, 281rocker-marked, 37Rockport Focus, 190, 195St. John's, 194sand-tempered, 41, 80, 184, 194, 239,281Savannah Check Stamped, 317shale-tempered, 37shapes, 185, 180 (fig.), 187 (fig.),188 (fig.), 189, 192 (table), 194,239shell-tempered, 46Socorro, 275Spanish wheel-made, 315stamped, 19, 21, 41, 46, 316, 317surface finish, 185, 189Tchefuncte Stamped, 178, 190, 193,194, 195, 239, 244tempering, 184, 239, 242texture, 184tooled, 77undecorated, 80unidentified, 190Upper Republican type, 36, 37, 44,45, 113, 131 "Valley cord-roughened, 131, 132,133, 134, 137Weeden Island, 194Woodland type, 36, 37, 44, 45, 131,183, 195Woodstock Diamond Stamped, 317zone-decorated, 37Pottery-bearing sites, 93Pottery complex, 184-196Pottery-making peoples, 24, 45, 52, 57,79, 85, 192, 282, 283Poverty Point, 220Powder River, Rlont. and Wyo., 84, 85Prairie Dog Creek, Nebr., 38, 40, 92,109, 110, 111, 113, 124site 25HN39, Nebr., 89Prairie Dogs, 110, 280Preceramic peoples, 57, 76, 82, 85Pre-Mankato climax dating, 99Prepottery cultures, 93Pre-White culture, 76, 98Primates, fossil, 32Procyon lotor, 157sp., 235Projectile points, 22, 23, 50, 51, 72, 76,85, 86, 91, 93, 96, 99, 121, 174,197, 201-220, 234, 239, 274-276,283Alba Barbed, 178, 198, 202, 214, 216,218, 220, 242, 244 Projectile points?Continuedantler, 232, 240, 241Baird beveled, 199, 214, 242chipped, 25, 30, 31, 198-200 (table)Clovis Fluted, 199, 205, 211, 216,217, 219, 220, 240, 241concave-based, 77, 122"contracting stem," 198Copena Point, 199, 213corner-notched, 15, 22, 31, 37, 52,82 276distribution, 216-217 (table)Eddv Stemmed, 176, 178, 198, 202,216, 217, 242espanded-base, 37, 239 "fish-tailed," 31Gary Stemmed, 171, 178, 183, 199,204-205, 214, 215 (graph), 216,217, 218, 219, 220, 239, 242Jumano Phase, 274Kobs triangular, 197, 198, 202, 203,216, 217, 242lanceolate, 23, 31, 76, 77leaf-shaped, 42long, 198metal, 83Middle and Lower Level types, 199,204-211Nebo Hill, 219, 220, 240, 241Pedernales Indented base, 210, 241Perdiz Pointed Stem, 171, 176, 178,183, 198, 201, 202, 214, 216, 217,218 219 220 242, 244Plainview Point, "199, 213, 219, 241provisional types, 198-199 (table),203, 204, 205, 206, 207, 208, 209,210, 211, 212, 213, 216, 217, 220,240, 241, 242, 243Scalhorn Stemmed, 178, 198, 201-202, 216, 242Scottsbluflf, 220, 240, 241serrate, 52side-notched, 73, 77, 83, 276stemmed, 17, 37, 48, 52, 88, 165stone, 318stratigraphy of, 214-218triangular, 36, 41, 77, 274, 275, 278Typological Concept of, 197unplaced types, 199, 211-214Upper Level types, 201-203 (table)Wells Contracting stem, 199, 205Pueblo of Pecos, N. Mex., 275, 278Pueblo region, N. Mex., 54, 275, 278Punches, bone, 48Quadrula quadrula, 136speciosa, 235Quartz, 37, 123, 316, 318Qnartzite, 23, 121, 274, 275, 276, 277, 278,279Quercus velutina, 156Quimby, George I., 190, 193 INDEX 333Kabbit, 42, 110, 235-236 (table), 237cottontail, 87, 280jack, 280Rabbit teeth, 121Raccoons, 110, 235 (table), 237{Procyon lotor) , 157, 319Radiocarbon dating, 193-194Raft Lake Reservoir site, Wye, 70, 71,84investigations, 86Rapid Creek, S. Dak., 20Rattlesnake, eastern {Crotalus horri-dus), 157Reagan, Ina May, 8, 68Rebecca Creek, Nebr., 112Recreation Survey OflSce, Region 2, 66Red River Drainage, S. Dak., 26Tex., 274Red Rock Canyon, Kans., 35Redondo formation, 290Red Willovp, Nebr., 10Refuse deposits, 27, 28Rehder, Harald A., 107Rembert, Captain, 314Rembert Mounds, The, Elbert County,Georgia (Caldwell), 303-320, 308(map)conclusions, 319-320the large mound, 309-311 (map),312 (figs.), 313 (figs.)the small mound, 314the village site, 314-315trait list, 318-319Renaud, B. B., 281Reptiles, 42fossil, 32, 93Republic, Kans., 39Republican City, Nebr., 38Republican River Basin, 10, 14, 38, 95,126Republican River, Nebr., 38, 39, 40, 46,91, 92, 93, 109, 110, 111, 112, 113,118Republican River Terrace, 2, 42, 50, 94Rhinoceros, fossil, 94Rio, Domingo del, 160Rio Grande River, 161, 274, 276, 278,2S9 283Rio Puerco, N. Mex., 275, 282River Basin Surveys, 2, 4, 12, 13, 20, 24,31, 32, 33, 34, 39, 40, 41, 42, 43, 47,48, 49, 54, 70, 71, 80, 82, 83, 84, 86,87, 88, 89, 90, 91, 93, 94, 96, 98,99, 107, 114, 151, 246former reports, 321-322procedure of, 3, 73Rivera, General, 159Roasting pits, 22, 30, 40, 51, 76Robb site (25WT4), 129-130 (table)Roberts, F. H. H., Jr., 1, 51, 65, 149, 278,279, 280, 282, 289Robinson, H. E., 13Rock Creek, Nebr., 10Rock shelters, 20 Rockville, Nebr., 43Rockyford Reservoir site, S. Dak., 70,71, 74, 81-82investigations, 81-82site 39SH1, 82site 39SH2, 82site 39SH3, 82Rodents, fossil, 32, 43, 94Rods, clay, 190Rogan, John P., 307, 310, 311Rosedale (formerly Clearwater), Ante-lope County, Nebr., 44, 49Rosedale Reservoir site, investigation,46Russell, Kans., 90Rygh (39CA4), S. Dak., 28Sabine River, Tex., 1.54Sagebrush, 30, 86, 87Salina, Kans., 34Saline River, Kans., 88, 90San Agustin de Ahumada presidio, Tex.,160San Bernard Creek, Tex., 154San Francisco Mission, Tex., 160San Gabriel River, Tex., 160San Ildefonso Mission, Tex., 160San Jacinto River, Tex., 151, 154, 1.55,161San Jon area, 274, 276, 277, 278, 279,280, 282, 283, 289, 290 (map), 299(table), 300 (table)Sandhill area, Nebr., 43, 55, 94, 95, 96Sandstone, 34, 35, 37, 229, 279Savannah River, Ga., 307, 309, 320Saws, sandstone, 201, 227, 229, 242Sayles, E. B., 150Schultz, C. B., 38, 91Sciurus carolinensis, 157niger rufiventer, 157Scottv Phillips Ranch site (39T14), S.Dak., 28Scraper-graver tools, 41Scrapers, 48, 76, 77, 83, 85, 86, 94, 136,174,221 (table), 240, 277chipped stone, 30, 36, 200 (table)concave, 276end, 15, 23, 35, .36, 37, 41, 82, 88, 89,127, 136, 200, 224, 230, 242flake, 200 224gougelike, 42nondiagnostic, 94planoconvex, 22, 37, 276side, 23. 37, 88, 200, 224, 242, 243,276, 283snub-nosed, 276, 283trapezoidal, 42Semihorticultural Indians, 52, 53, 54,57, 79Seventh Plains Conference for Arche-ology, 70Shadehill Reservoir, S. Dak., 70, 71, 74,82investigations, 82-84 334 BUREAU OF AMERICAN ETHNOLOGYShaft wreuches, bison-rib, 41Shale, 276Sheep, 87Shellfish, 162, 234, 236, 237, 240, 246Shell fragments, 77, 90, 94, 114, 115Shell industries, 57, 123-126, 129(table), 234, 242Shells, 280-281clam, 165conch, 125, 136, 319, 320fresh-water, 112, 116, 118, 126, 128,281marine, 112, 117, 118, 126mussel, 76, 93table of, 136Sheps Canyon, ?. Dak., 21Sheyenne River, N. Dak., 47Shippee, J. M., S. 12, 20, 25, 29, 34, 39, 40,68, 71, 107Shoshonean Indians, 24, 56, 84Shoshoni, Fremont County, Wyo., 31Signal Butte 1, 50Sims, John M., 149Sioux, 79Sixth Plains Archeological Conference,11Skeletal remains from Doering andKobs Sites, Addicks Reservoir,Tex. 257?266Skeletons, 47, 48, 69, 116, 127, 314analysis of, 264-266extended, 17female, 266pathology, 266Woodruff, 137-140?Skull fragments, 258 (flg.),259 (table),260, 261, 263Skulls, 116, 258 (fig.), 260 (fig.), 262(figs.), 263 (fig.), 314artificial deformation, 266long-headed, 264, 265measurements of, 139-140 (table),258pathology, 266round-headed, 264, 265Skunks, 110Smith, Carlyle S., 34, 35, 88Smoky Hill Basin, Kans., 10, 33, 34, 35,55Smoky Hill River, Kans., 34, 36, 38, 53,54, 88Smoothing tools, bison-rib, 48Snails, 235Snake, coral (Micrurus fulvius fulvius),157Society for Vertebrate Paleontology, 31Solomon River, Kans., 88, 89Solomon Valley, Kans., 89tSoral Creek Reservoir site, Wyo., 70, 71,84,86investigations, 86-87Soto, Hernando de, 158South Dakota Archeological Commis-sion, 27 South Dakota, sites investigated, 20-29,74-84South Fork, Grand River, S. Dak., 82,83South Mayde Creek, Tex., 151, 155, 163,164,167,168 (map), 174Southwestern Archeological Conference,197Spalls, stone, 31Spanish Diggings, Wyo., 23Spatula, ulna, 231, 233, 240, 241Spheroids, stone, 42Spring Creek, Nebr., 44Spring Creek, Tex., 161Spruce, 86Squash, 16, 57, 110Squash seeds, 15Squirrel, gray (Sciurus caroUnensis) , 157western fox ( Sciurus niger rufiven-ter) 157Stalling's Island, 220, 230Steatite, 18Stereograph, Schwarz, 258Sterns Creek, Nebr., 56, 137Stinking Water Creek, Nebr., 55Stone, chipped, 91fire-cracked, 88ground, 229industries, 57, 121-123, 129 (table)worked, 274-280Storage pits, underground, 16, 27, 57Strong, WilUam Duncan, 111, 112, 127quotations from, 28, 118, 123-124Subsistence, 234-236Sundance Creek Reservoir site, Wyo.,22Sunflower, cultivated, 16, 57seeds, 15Swanton, John R., quotations from, 161,162, 246Tatum, Webb, 307Taxidea sp., 235, 236Tchefuncte cultures, 178, 218, 220, 244,245Teeth, carnivore, 48complete dental arch of, 48perforated canine, 25, 48Telegraph Flat, S. Dak., 28Tents, bearskin, 162Terraces, tetragon, 314, 319Terrapene sp., 235, 236Terry, Mont., 85Teton Dakota Indians, 23Texas Council of Archeology, 152, 197Thomas, Cyrus, 308, 309, 310, 311, 312,313, 314Thompson Creek, Kans., 35, 36, 54Three Cave Farm (14EW7), Kans., 35Thunderbird, 78Tiber, Mont., 10, 67Tipi rings, 22, 23, 72, 83, 85, 91Toohey, Loren M., 94 INDEX 335Tools, stone, 31Topography, 152-153 (map), 154Tortoise, 234, 235, 236, 237Towns, fortified, 58Townsend, Broadwater County, Mont.,32Trading, 57, 58, 246Trait list, 318-319Trantina, John, 24Trees, hardwood, 81Triassic Age, 290sandstone, 291, 292 (fig.), 296 (fig.)Tri-County Project, P. W. A., 45Trinity River, Tex., 154, 160, 161Tubes, bird-bone, 48, 317Tucumcari Creek, N. Mex., 271Tiicumcari, N. Mex., 271, 282, 290Tularosa Basin, 283Tumlin Mound, Ga., 310, 317Turkey Creek, Nebr., 46Turkey Creek, Tex., 151, 155, 163Turtelot, Harry A., 31Turtles, 41, 157, 165Two Rock Shelters near Tucumcari, N.Mex. (Dick), 267-284Unio sp., 281Vniomerus tetralasmtis, 235, 281U. S. Army Engineers, 290United States Geological Survey, 31University of Arizona, 149, 150University of Kansas Museum of Na-tural History, 34, 35, 37, 49, 88, 89,90University of Michigan, 149, 281University of Nebraska Labratory ofAnthropology, 38, 39, 40, 43, 44,45, 46, 49, 50, 54, 66, 68, 70, 91, 92,100University of Nebraska, State Museum,2, 4, 8, 12, 23, 38, 41, 42, 43, 49,66, 91, 93, 94, 98University of North Dakota, 19, 27, 46,47, 49University of South Dakota Museum,27University of Texas, 31Department of Anthropology, 149,152, 197University of Wyoming, 31Upper Republican culture, 15, 16, 17,21, 36, 37, 42, 46, 50, 52, 54, 55,57, 76, 90, 93, 96, 111, 113, 119,120, 121, 122, 130, 131, 185, 136,137Upper Republican village sites, 14, 15,16, 17, 18, 39, 44, 54, 92, 100, 126Ursus americanus, 157Valley City. N. Dak.. 47Vertebrates, fossil, 43Vessels, elongate pottery, 25 Villages, ditched and palisaded, 17, 27earth-lodge, 27fortified, 27Village sites, 58Vivipartis intertextus, 235Wakeeney, Ti-ego County, Kans., 33, 88Wall plaster, clay, 317Ward, William V., 149Waterfowl, 110Wattling clay, 44Wedel, Waldo R., 70, 107, 112, 127(Prehistory and the Missouri Val-ley Development Program ; Sum-mary report on the Missouri Riv-er Basin Archeological Surveyin 1948), xv-59(Prehistory and the Missouri Val-ley Development Program ; Sum-mary report on the MissouriRiver Basin Archeological Sur-vey in 1949), 61-101Wheat, Joe Ben (An Archeological Sur-vey of the Addicks Dam Basin,Southeast Texas), 143-252Wheeler, Richard P., 67, 68, 71, 72, 74, 84Wheeler Bottom, S. Dak., 25White, George, 307, 314White, T. E., 8, 12, 13, 31, 33, 68, 149White contacts, proof of, 92, 100Whiteford, G. L., 34White Oak Bayou, Tex., 155Wliite Oak Creek, Tex., 152White Oak Dam site, Tex., 152White River, Nebr., 23White River, S. Dak., 74, 81, 82Wichita Indians, 36, 54, 58, 274Willey, Gordon R., 315Willow, 38, 84, 86, 87, 109Willow, long-leaved, 110Wilson Reservoir site, Kans., 10, 88, 90investigations, 90site 14RU2, 90Wind River Indian Reservation, Wyo.,84Wind River (town), Wyo., 86Witte, A. H., 274Wittry, Warren, 8, 19, 20Wolbach, Greeley County, Nebr., 44Wood Island, S. Dak., 25Woodland culture, 16, 37, 39, 42, 44, 52,53, 76, 100, 134, 136Woodland groups, 15, 18, 53, 93Woodland village sites, 17, 56, 57, 95,136site 25H07, 95site 25H09, 95site 25H021, 95site 25H024, 95W d 1 a n d-Upper Republican-LowerLoup-Pawnee sequence, 55 336 BUREAU OF AMERICAN ETHNOLOGYWoodruff, Kans., 113Woodruff Ossuary, a Prehistoric BurialSite in Phillips County, Kansas(Kivett), 103-140cultural relationships, 130-137environmental background, 109-111previous archeological work, 111-113site 14PH4, 69, 113-118, 124, 126,128, 129-130site 14PH5, 113skeletal remains, 137-140 Works Project Administration, 10, 27Wounded Knee Creek, S. Dak., 81Wrenches, shaft, 136 |Wyoming and Montana, sites investi-gated, 84-87Wyoming, sites investigated, 29-31Yellowstone River, 55, 56Yellowstone Valley, 85Yellowtail Reservoir area, 67Younkin mound, Kans., 133, 134,137Yuma complex, 56 135,O I