SMITHSONIAN INSTITUTIONBUREAU OF AMERICAN ETHNOLOGYBULLETIN 92 V\ V<2? V, Y) SHABIK'ESHGHEE VILLAGEA LATE BASKET MAKER SITEIN THE CHACO CANYONNEW MEXICO BYFRANK H. H. ROBERTS, JR. UNITED STATESGOVERNMENT PRINTING OFFICEWASHINGTON: 1929 ADDITIONAL COPIESOF THIS PUBLICATION MAY BE PROCURED FROMTHE SUPERINTENDENT OF DOCUMENTSU. S. GOVERNMENT PRINTING OFFICEWASHINGTON, D. C.AT$1.00 PER COPY (CLOTH) LETTER OF TRANSMITTAL Smithsonian Institution,Bureau of American Ethnology,Washington, D. C, July 10, 1928.Sir: I have the honor to submit the accompanying manuscript,entitled " Shabik'eshchee Village, a Late Basket Maker Site in theChaco Canyon, New Mexico," by Frank H. H. Roberts, jr., and torecommend its publication, subject to your approval, as a bulletin ofthis bureau.Respectfully, H. W. Dorsey, Chief Clerk.Dr. C. G. Abbot,Secretary of the Smithsonian Institution. ni CONTENTS PageForeword 1Introduction 2Basket Maker I 3Basket Maker II 3Basket Maker III 4Pueblo I 4Pueblo II 5Pueblo III 5Pueblo IV 6Pueblo V 7Present status of the study 7The San Juan region 8The Chaco Canyon 9Shabik'eshchee VillageThe ruins 10House A 10House B 16House C 17House D ' 22House E 24House F 26House G 34House H 37House I 39House J 41House K 44House L 46House M 49House N 53House O 55House P 57House Q 58Structures excavated in 1926 61Protokiva House 62House X 68Arroyo House 70Court near storage bins 71Large circular structure, the kiva 73Influence of early dwellings on later kivas 81Storage structures 91Refuse mounds 105v VI CONTENTS PageMaterial culture 106Pottery 107Designs on the painted pottery 118Additional clay objects 124Bonework 126Awls 127Spatulate awls 128Punches 129Flaking tools 129Scraper or dressing tool 130Spoonlike spatula 130Needles and bodkins 130Whistles and beads 131Objects of stone 132Metates 132Mauls or hammers 134Chipped implements 136Other objects of stone 139Ornaments 142Shell ornaments 142Human burials 143Summary and conclusions 145Appendix?Catalogue number and provenience of objects illustrated 151Bibliography 155Index 159 ILLUSTRATIONS PLATES Page1. Map of Shabik'eshchee village 10.2. House remains, a, House A. b, House C 203. House C. a, Holes for roofing poles, b, Corner storage box. c, Wallrecess 204. House remains, a, Sipapu in floor of House F. b, House D 286. House remains, a, House F-l. b, House H 286. House remains, a, House L. b, House N 547. House remains, a, Protokiva house. 6, Doorway from passage side. c, Doorway from room side 548. House remains, a, Arroyo house, b, House X 729. a, Southeast portion of kiva. b, Bins Nos. 8 and 9 7210. Storage bins, a, Bin No. 11. b, Bin No. 5 11011. Pottery fragments 11012. Fragments from culinary vessels 11013. Bowls, a, Polished black interior bowl, b, Bowl with painted decora-tion 11014. Potsherds with painted designs, black-on-white series 12015. Designs from bowl interiors 12016. Designs from bowl interiors 12017. Bowls, a, Undecorated. b, With painted design 12018. Cloud blowers and additional clay object 12819. Bone implements, awls 12820. Bone implements, awls 12821. Bone implements, awls 12822. Bone implements, spatulate awls 12823. Bone implements, punches 12824. Bone implements, a, Spoon spatula, b, Scraper or dressing tool, c,Flaker 12825. Bone implements, needles, bodkins, whistles, and beads 12826. Stone objects, metates 13227. Stone implements, mauls 13228. Chipped stone implements 13629. Miscellaneous stone objects 13630. Ornaments, stone and shell 14231. Human burials 142TEXT FIGURES 1. House A 112. Postulated method of house construction 123. House B 164. House C 185. House D _ _ __ 22VII VIII ILLUSTRATIONS Page6. House E 257. The F House group 278. House F ! 289. House F-l 3010. House G 3511. House II 3712. House I 4013. House J 4214. House K 4515. House L 4716. House M 5017. House N 5318. House O 5619. House P 5720. House Q 592 1 . Protokiva house 6322. House X 6823. Court ?_ 7224. Kiva 7825. Constructional feature in kiva 7926. Construction of Pueblo-type kiva 8327. Storage bins 5, 7, 8, 9 9528. Storage bins 36, 37, 39 10229. Outlines of culinary vessel shapes 11130. Nonculinary vessel shapes 11431. Vessel with lateral spout 11732. Painted design from half of a bowl 122 SHABIK'ESHCHEE VILLAGEA LATE BASKET MAKER SITE IN THE CHACO CANYON, NEWMEXICO By Frank H. H. Roberts, Jr.FOREWORD During the summer of 1926, while engaged in excavating a smallpueblo ruin in the upper Chaco Canyon, N. Mex., for the NationalGeographic Society's Pueblo Bonito Expedition, the writer dis-covered the remains of a number of storage cists on top of the mesawhich forms the southern wall of the canyon at that place. Potteryfragments lying on the surface and the character of the cists them-selves suggested that at one time the site had been occupied by acruder, less highly developed people than those of the Pueblo cultureswhose ruined villages, great communal centers, and refuse mounds areto be found in such abundance throughout the length and breadthof the Chaco district. Two houses and several storage cists wereuncovered, but extensive examination of such a site was not pro-vided for in the program of excavations then being brought to com-pletion. It was felt, however, that greatly to be desired informationon an imperfectly known phase of southwestern archeology was to beobtained by careful investigation of this mesa top. Consequently,when the opportunity for field work was presented in the summer of1927 the Chaco Canyon site was chosen.The remains of the village had never been reported and it wasunnamed. At the foot of the trail leading up to the site is a largerock upon which the Navahos have pecked a sun symbol. Navahoworkmen employed in the excavations referred to the location asSun Picture place. This has been deemed as fitting a name as anythat could be applied to the one-time community. Hence the designa-tion Shabik'eshchee village.The writer wishes to thank Mr. Neil M. Judd, director of theNational Geographic Society's Pueblo Bonito Expedition, who con-tributed in no small measure to the success of the summer's investiga-tions by sharing his camp as well as transportation equipment withthe Bureau of American Ethnology Expedition. Because of thiskindness and whole-hearted cooperation much more was accomplishedthan would have been possible had the maintenance of a separatecamp been necessary. l INTRODUCTION In order that the village excavated during the summer of 1927 maytake its proper place in the sequence of prehistoric southwestern cul-tures and that the reader may have some appreciation of the prob-lems which confront the archeologists who are striving to reconstructthe stages in their development, a brief review of our present knowl-edge of the area may not be amiss.That portion of the United States in which are found the PuebloIndians and traces of their former homes, in the ruins of the pre-Columbian culture centers of their ancestors, comprises the territoryincluded in the States of Arizona, New Mexico, southwestern Colo-rado, almost all of Utah, a small section of southeastern Nevada, andthe great inland basin of the northern Chihuahua district of Mexico.Topographically the area is a high and arid plateau sloping away tothe south and west from the Rocky Mountains.The region has four large subareas in the drainage systems of thefour main rivers?the San Juan, the Rio Grande, the Little Colorado,and the Gila-Salt. These streams, and the tributaries of each, are allthat carry water the year around. It is in great part a desert countrywith little vegetation. There is some scanty grass, sagebrush, mes-quite, and, where conditions are more favorable, cedar and pinontrees. On the slopes of the higher mountains spruce and pine are tobe found.There was comparatively little game in any but the mountainousdistricts and it was never extremely abundant there. It consistedchiefly of deer, bear, rabbit, wild turkey, grouse, and quail, with anoccasional mountain lion and wildcat.The soil of the area is such, however, that where water can besecured fairly abundant crops can be produced. That the prehis-toric settlers of the region took full advantage of this productivityis shown by the countless ruins scattered throughout the area. Thereare the remains of great cliff dwellings, of communal structures builton the tops of mesas or on the canyon floors, small house sites,cavate lodges, and even in caves where there were no buildings thereare additional evidences of the occupation of the region.Each of the great river basins seems to have developed a culturewith more or less distinctive characteristics in the peculiarities ofhouse construction and grouping, in pottery types, baskets and tex-tiles, while still retaining its relationship to the general widespreadculture. The outstanding feature of the whole area may be summed2 koberts] SHABIK'ESHCHEE VILLAGE 6 up in the statement that it was the home of a sedentary, agriculturalpeople who built their houses either of stone or adobe, or both, madepottery, and wove textiles.This house-building, pottery-making culture did not spring fullblown from the soil but grew by degrees from a very simple begin-ning, covering a long period of time in its unfolding. The stagesin its growth and development are shown in a number of periodsduring which the people passed from a simple hunting type to arelatively highly cultured group. To facilitate study of this growththe several cultural levels have been classified under two main head-ings, called the Basket Maker and the Pueblo. As might be ex-pected, there are a number of phases, with distinct characteristicsin each. The Basket Maker group, which is the older, has three,while the Pueblo has five.1Basket Maker IThe Early Basket Makers were probably a more or less nomadicgroup sparsely scattered over the area. They depended to a largeextent on the natural caves of the region for shelter, although, ifnecessity demanded, they may have erected flimsy, perishable dwell-ings to shield themselves from the vagaries of the weather. Smallgame, wild vegetable products, and such fruits as the country affordedfurnished them with a meager supply of food. Later developmentswere foreshadowed, however, in their baskets and textiles. The ap-pearance of corn, presumably coming from the highlands of Mexicoor Central America to the south, led to the beginnings of agricultureand the second stage of their culture.Basket Maker IIThe Basket Makers proper were a semihunting. semiagriculturalpeople. They grew a single type of hard, flintlike corn and had avariety of squash, but apparently built no permanent houses andmade no pottery. They stored the grain from their crops in stone-lined pits located in the floors of caves. These cists were often putto a secondary use for burial purposes, and it is from these gravesthat the knowledge of their arts and industries has come down to usthrough the objects which they interred with their dead.The chief industry seems to have been that from which the grouptakes its name, basket making, and in this they were skilled crafts-men. Their baskets, twined and woven bags, ropes, sandals, andother objects are of a high degree of excellence. In addition, theydid some work in stone chipping. They made javelin points and 1 Kidder, Southwestern Archeological Conference, p. 490. 4 BUREAU OF AMERICAN ETHNOLOGY [bull. 92 various kinds of cutting edges. Several types of wooden implementswere fashioned, but the bow and arrow apparently were unknown.The main offensive weapon was a rather short spear used in con-junction with a spear thrower or atlatl.Present evidence indicates that it was at the very end of the periodthat the idea of making clay vessels developed. There was no truepottery, however. Thick, unfired containers of clay were crudelymolded in baskets. To prevent cracking when the clay dried, smallpieces of cedar bark were used as a binder. The idea of potterymaking possibly came from the centers of higher culture to thesouth, but the development of it was unquestionably local.2Basket Maker IIIAt about the time when the people were experimenting with mudcontainers and laying the foundation for the ceramic industry ofsucceeding periods another change is noted. They began to enlargethe excavations for their storage pits, to erect large stone slabs toserve as walk, and to construct pole, brush, and plaster coveringsover them, so that the dwelling of a more or less permanent natureevolved. This marked the beginning of the Late Basket Maker orBasket Maker III phase. The period witnessed a slight decline insome of the woven materials and an advance in other industries.Ceramics progressed to a point where fired pottery with painteddecoration made its appearance. Several varieties of corn weregrown. Beans were added to the supplies of staple foods. Featherrobes began to replace those of fur. The bow and arrow came intouse. It was toward the end of this period that the village excavatedin the Chaco Canyon was built and inhabited. Intervening betweenits abandonment and decay is the whole Pueblo cycle.The Basket Makers are known to have occupied parts of south-eastern and south central Utah, northeastern and southern Arizona,some sections of Nevada, portions of New Mexico and southern Colo-rado, and extended southward into Mexico. Indications are, how-ever, that they reached their highest development in the San Juanarea. Pueblo I Several features mark the beginning of the Pueblo era. The firstperiod, frequently referred to as the pre-Pueblo, saw the arrival of anew group of people, a round-headed stock, who practiced cranialdeformation. The preceding Basket Makers were long headed, doli- 2 Guernsey and Kidder, Basket-Maker Caves of Northeastern Arizona, p. 98.Nusbaum, A Basket-Maker Cave in Kane County, Utah, p. 144.Morris, Beginnings of Pottery Making in the San Juan Area, pp. 138-152, 198. kobebts] SHABIK'ESHCHEE VILLAGE 5 chocephalic, and their skulls show no occipital flattening. The crude,one-room domiciles gave way to structures with several contiguousrectangular rooms the slab walls of which were frequently augmentedby true masonry. Cotton was introduced, and fabrics made from itassume an important role in the material culture. The wild turkeywas domesticated. Necks of large pottery vessels were left un-smoothed, showing the coils of clay from which they were made.Pueblo IIThe second Pueblo phase was the small-house period, which wasmarked by widespread distribution of life in small villages. In theSan Juan area the general type of dwelling was that which has beencalled the unit or one-clan house. 3 The walls were completely con-structed of masonry and the dwelling was entirely above ground.The houses contained several rooms, were one story in height, andthere was generally a court on the south or southeast side. Beneaththe courtyard was a circular ceremonial room or kiva. Corrugatedpottery, so plentiful in southwestern ceramic collections, with elabo-rate indentations covering the entire surface of the vessel, was oneof the distinctly characteristic features. The painted wares werequite generalized in style and form. Pueblo IIIThe greatest period in the Southwest was that of Pueblo III whenthe large urban centers developed. It was during this phase thatthe finest examples of Pueblo architecture were completed. On theMesa Verde were Spruce Tree House 4 and Cliff Palace. 5 In theChaco Canyon Pueblos Bonito 6 and Chettro Kettle 7 were thrivingcenters. The inhabitants of the Mimbres Valley in southern NewMexico were making their pictorially decorated pottery,8 CasaGrande 9 in Arizona was the scene of great activity. The proto-Kayenta villages 10 in the northeastern section of the same State wereprospering. It was, in fact, the heyday of the Pueblo peoples.Pottery had become specialized to such an extent that each centerhad its own definite forms, forms so characteristic that each vessel 3 Prudden, Prehistoric Ruins of the San Juan Watershed, pp. 224?288.4 Fewkes, Antiquities of the Mesa Verde National Park : Spruce-tree House.6 Fewkes, Antiquities of the Mesa Verde National Park : Cliff Palace.9 Pepper, Pueblo Bonito. Judd, Everyday Life in Pueblo Bonito ; Archeological Investi-gations at Pueblo Bonito.7 Hewett, Chaco Canyon and Its Ancient Monuments ; The Excavation of Chettro Kettle,Chaco Canyon, 1920 ; The Chaco Canyon in 1921.8 Fewkes, Archeology of the Lower Mimbres Valley ; Animal Figures on PrehistoricPottery from Mimbres Valley ; Designs on Prehistoric Pottery from the Mimbres Valley.8 Fewkes, Casa Gran-de, Arizona.10 Kidder, Southwestern Archaeology, pp. 72-73. 6 BUREAU OF AMERICAN ETHNOLOGY [bull. 92 tells at once the place of its provenience. It is because of this factthat the trade relations between the various centers can be recorded ; that the boundaries of stylistic influence can be traced.The closing phase of the period was characterized by a marked de-cline in the great centers, the abandonment of many of them, andmany migrations which as yet have not been traced. It is fairlycertain that Kitsiel and Betatakin u in northeastern Arizona andmany of the black-on-white sites, so called from the pottery, nearPecos, N. Mex., represent this phase.Because of certain apparent relationships between the last stages ofthe great period and the more highly developed cultures to the southin Old Mexico, phase B, so to speak, of Pueblo III may be given theapproximate dates of 1100 to 1300 A. D. These dates must not beconsidered as absolutely accurate, but they are based on evidencewhich makes them a fairly close approximation. 12Pueblo IV There are two phases in the fourth period of the Pueblo cycle. Thefirst is the protohistoric which extended from about 1300 A. D. to thetime immediately preceding the arrival of the Spaniards in 1540. Itwas undoubtedly during1 this interval that Sikyatki 13 in the Hopicountry and the early glaze ware sites of the Pajarito district alongthe Rio Grande flourished. 14 The latter part of Pueblo IV, from thearrival of the Spaniards to the end of the reconquest in 1696, is theEarly Historic period which was characterized by a still further de-cline from the preceding cultural peak and the gradual disappearanceof the corrugated pottery. This period is best represented by oldAwatobi in the Hopi country which was destroyed in the autumn of1700,15 Hawikuh, near the present village of Zuiii, N. Mex., whichwas abandoned in 1670,16 and Pecos,17 not far from Santa Fe in thesame State. There is, of course, a certain overlapping of periods.Hawikuh and Awatobi were thriving villages before the Spaniardsappeared and Pecos continued to be occupied until 1838, but for gen-eral purposes they fall into the grouping suggested. It should beborne in mind that any classification must be to a certain extentarbitrary and often is not wholly satisfactory. 11 Fewkes, Preliminary Report on a Visit to the Navaho National Monument, Arizona,pp. 12-17. "Tozzer, Time and American Archaeology, pp. 210-221.13 Fewkes, Expedition to Arizona in 1895, pp. 631-742.11 Kidder, Southwestern Archaeology, p. 86. ,B Fewkes, op. cit., pp. 592-631.16 Hodge, F. W., Excavations at Hawikuh, N. Mex. ; Excavations at the ZuniPueblo of Hawikuh in 1917.17 Kidder, op. cit., pp. 4-35. bobebts] SHABIK'ESHCHEE VILLAGE 7Pueblo V Pueblo V is the period of the modern villages and covers theinterval from the final subjugation of the Pueblos by the Spaniardsto the present* day. It is the period which shows the gradual break-ing down of Pueblo life under the influence of the white man.Traces of the Pueblo I period are found throughout the entire SanJuan region, in parts of the Upper Gila and Rio Grande districtsand along the Little Colorado. There seems to have been a rathermarked increase in population. Pueblo II ruins are found over theentire area, while the later development of the many-roomed, com-pact structures and great centers of Pueblo III show a contractionof the population. There was an even greater contraction in PuebloIV with a rather marked decrease in population. Following thisstage the tendency to dispersion again appeared and in very recenttimes there has been a marked breaking up of villages and develop-ment of small communities. Present Status of the StudyThe story of the rise and decline of southwestern sedentary culturesis in the main as outlined in the foregoing paragraphs. As furtherinvestigations bring more evidence to light there no doubt will bechanges in some portions of the account, but the framework willremain essentially the same. The Basket Makers constitute a stillimperfectly known chapter. This is especially true of their firstperiod which, thus far, is largely postulated, although evidence of ithas been found in one, possibly another, locality. The second stage,that of the Basket Makers or Basket Maker II, was until recentlythe best known of the three horizons in the culture. Now, however,there are more data on, and there is a fuller knowledge of, the LateBasket Makers, or, as they sometimes are called, the Post-BasketMakers.It was for the purpose of obtaining much-needed information onthe house and village types of the final period in the Basket Makerera that the excavations in the Chaco Canyon were conducted. Mostof the previous work had been done at sites located in caves wherelater occupants had to some degree disturbed the older remains.Occasional Basket Maker III houses had been excavated but not insufficient numbers to warrant definite conclusions as to their types.The Chaco Canyon site was an especially fortunate one because itwas in the open and no later buildings had been erected upon it. Itwas an isolated example of a single cultural stage.The results obtained from the excavations far surpassed expecta-tions and from the information secured a much clearer picture can be 8 BUREAU OF AMERICAN ETHNOLOGY [bull. 92drawn of the houses and house life of the days immediately preced-ing the appearance of the Pueblo peoples.There is still a large problem to be solved in connection with theBasket Makers, however, and that is the question as to what becameof them. Were they absorbed by the incoming peoples or did theydrift away to settle in other sections of the country and there carryon their established industries? This is a question which, in thepresent state of our knowledge, can not be answered. Some of theinvestigators profess to see in the modern Mohave and Paiute peoplesthe remains of the old Basket Makers. Others are of the opinionthat they were completely absorbed by the newer arrivals in theSouthwest and that all traces of them were obliterated through theresulting mixture and the widespread practice of cranial deforma-tion. The problem is as yet untouched.On the cultural side there seems to have been a very even andgradual development from the one group into the other and in placesthe change is so slight that there is no perceptible break. It is truethat some new features were introduced, but on the whole thePueblo peoples seem to have taken over most of the things thattheir predecessors developed and to have built their own cultures onthe firm foundation already established.The San Juan RegionAs has been suggested in the foregoing discussion, the San Juanregion was one of the most important in the entire area. It seemsto have been the center from which many of the characteristic featuresof the sedentary cultures were diffused and for a long time was theleader in their development. It is the northernmost of the four greatterritorial divisions of the area and unquestionably is the best knownarcheologically.The San Juan itself and its northern tributaries, the Piedra, Pine,Animas, and Mancos, have their sources on the slopes of the SanJuan Mountains in southern Colorado. They are the only streamswhich carry a never-failing supply of water. The three great tribu-taries from the south, the Largo, Chaco, and Chin Lee, are almostwholly dry for a great part of the year.In the northeastern part of the drainage where the mountains andhigher hills abound there is a comparatively thick forestation. Pine,spruce, cedar, and pinon trees are rather abundant. On the lowerlevels are the small-growing oaks, mesquite bushes, and sagebrush.There is quite an abundance of yucca, the Spanish Bayonet, in boththe broad and narrow leaf varieties, also many forms of cactus.Where the river bottoms open out cottonwoods and willows flourish.Once the mountains have been left behind, however, the region Roberts] RHABIK'ESHCHEE VILLAGE 9becomes one of great barren stretches, with only occasional cedarsand pinons on the mesa tops, clumps of sagebrush, and rather scantygrass. The summers are very hot and dry and the winters cold.The Chaco CanyonThe Chaco Canyon bears an important relation to the San Juanregion in that some of the finest examples of buildings of the greatPueblo period, Pueblo III, are to be found there. Not only is thistrue, but there is evidence to show that it was occupied for a longperiod of lime, unquestionably from Basket Maker III throughPueblo I, II, and III. Even then its history was not completed,because during the tempestuous days of the Pueblo revolt, 1680-1696, it served as a place of refuge for some of the warring factionswhich were forced to flee from the avenging Spaniards. In morerecent times it has become the home of small groups of NavahoIndians.From a purely scenic standpoint the Chaco Canyon is not im-pressive and can not be compared with the Mesa Verde to the northor with the Kayenta country to the west. The canyon itself isquite narrow, at no point being a mile wide. Its walls are of redsandstone and the mesa tops on either side are almost barren. Thereare a few stunted cedar and piiion trees, some sagebrush, and scantygrass. Near its upper end, however, there are pines and the smallertrees are more numerous. In all directions the region is marked byshifting sand, great dry washes, deep arroyos, and a lack of vege-tation. For one reason or another, however, this was a favoredspot in the eyes of the early peoples, as is shown by the 11 largepueblo ruins and the almost countless numbers of small house sitesscattered along its length.9215??Bull. 92?29 2 SHABIK'ESHCHEE VILLAGETHE EUINSShabik'eshchee village was located on top of the mesa which formsthe southern wall of the canyon 9 miles (14.49 k.) east of the greatpueblo ruins of Bonito and Chettro Kettle. It was very close to theedge of the escarpment and originally consisted of 18 houses, a smallcourt, a large circular ceremonial structure or kiva, and 48 storagebins.Before excavations were started the only surface evidence that theremains of a once thriving community lay buried beneath the sandwas in the slightly projecting tops of a few of the stone slabs usedin the construction of the storage cists and in the scattered fragmentsof pottery. There were no indications of houses or kivas. Soonafter work was commenced the trenches began to reveal the remainsof crude one-room domiciles, and at the end of the season there stoodexposed to view all that was left of a small village, one in which thedwellings followed the curve of the mesa top, extending northeastand southeast from an almost central kiva. (PL 1.)The houses had indeed been primitive. They had consisted of acircular, oval, or rectangular excavation roofed over with a pole,brush, and plaster superstructure. The earth walls of the excava-tion were covered with thick plaster made from adobe which occa-sionally had a slight admixture of ashes. Oftentimes large stoneslabs were used to line the excavation. When this was done theplaster was applied to the faces of the slabs. It was this use ofstone slabs which led, in earlier reports on this cultural stage, towidespread references to the slab-house people. Close to the housein most cases were small circular, oval, or rectangular pits in which * the corn, raised on the canyon floor below, was stored. These, too,had had pole, brush, and plaster coverings.House AA good example of the rectangular dwelling was found in house A.(PL 2, a; fig. 1.) All that remained was the excavation, but therewas sufficient evidence in the debris-filled interior to indicate what thestructure had been like. Large slabs had been used to line one andportions of two other sides of the excavation, the remainder havinghad the plaster applied directly to the native earth.10 CTT A DTTriPOTTPIIPP IZTI T \ n T7 11. eUREAU OF AMERICAN ETHNOLOGY o . o' BULLETIN 92 PLATE 1 Or ! 4- a '5 -~ . HOUSE C HOUSE A II. 7m KIVA OM6 17 c18 J P HOUSE D REFUSEMOUND * "N,,,,,,,,,,,,^* /y P. .2w n *~ ac3 CI +j? 3 to ? Q c3 CO 3CO r>O?"aOa a,go6 -ASi <" 5CO te03 ^Ah -a -M . 8!t* C3s PUa ?03 ???V o .a 'C & a> g-S o?.s aX si ~02 o . -|-J ?t*a ?a o 3 [Qa a fl :r fe and a thick coating ofplaster (fig. 2), indi-cated by the finding, inthe debris which filledthe room, of largepieces of plaster bear-ing the imprint ofleaves, sticks, and bark.This type of super-structure is known ?from other localities aswell. It varies some-what in specific fea-tures, but the generalcharacteristics are thesame. Doctor Kidderand Mr. Guernseyfound the remains of acircular room in HagoeCanyon, northeasternArizona, in which therewas a small offset orbench into which thebutts of 2 to 3 inchpoles were driven atsuch an angle that they-must have met over thecenter of the room at aheight of about 6 feet(1.8288 m.) above thefloor. There were nosupport posts in thisstructure, however, andthe roof seems to havebeen conical ratherthan flat-topped. Itwas of the type sug-gested by the storagebins of the Chaco site. 18 Mr. N. M. Juddfound a form of super- 18 Kidder and Guernsey, Archeological Explorations in Northeastern Arizona, pp. 43-44,fig. 18. Roberts] SHABIK'ESHCHEE VILLAGE 13 structure practically identical with those indicated by the ruins ofShabik'eshchee village in the remains of a dwelling not far fromPueblo Bonito which he excavated in 1922. 19 Judd also reports asimilar construction for the coverings of the earth lodges which heinvestigated near Willard, Utah. 20 The latter structures, however,seem to belong to a cultural phase more nearly approximating PuebloI than the Basket Maker III horizon. Doctor Cummings found anumber of excavated circular rooms which had been roofed over withpoles, brush, cedar bark, and plaster in the same region as the Kidderand Guernsey example just described. 21 Mr. Morris reports the sametiling from Mummy Cave in Canyon del Muerto, northeastern Ari-zona, where he found roughly circular pits which had had upperwalls "of posts set leaning inward and plastered over with mud, andthe roofs of logs covered with earth." 22 Doctor Hough found poleand brush coverings supported by posts in the pit ruins which heexcavated near Luna, N. Mex., but the roofs were quite different inthat there seems to have been a central pillar. These houses repre-sent a later period, however, and are not to be considered as BasketMaker III dwellings. 23 Dr. Kalph Linton found a conical type ofsuperstructure without interior support posts in Earth Lodge A onthe Mesa Verde. This structure was unquestionably of the BasketMaker III period.Near the center of the room of house A was a rectangular fire pitlined with small stone slabs with the tops flush with the plasteredfloor. It was well made but quite shallow. Close to the fire pit, inalmost the exact center of the room, was a small oval hole filled withclean sand. Just what its purpose was is not clear. There was asimilar feature in a majority of the houses. It is possible that itrepresented the same thing as the small hole found in about thesame position in the floors of later-day kivas. The latter, called thesipapu, symbolized the mythical place of emergence through whichthe Hopi, Zurii, and other southwestern Indians believe their ancestorspassed on their journey from the underworld to the surface of theearth. It may well be that in the days when Shabik'eshchee villagewas inhabited each house had its own shrine, whereas in later timesit was deemed essential only to the ceremonial house, the kiva. xVmore practical explanation would be that it was merely a place toset vessels where the heat from the fire would keep their contentswarm. The rounded bottoms of the containers would necessitatesome support to keep them in an upright position, and a hole filledwith sand would admirably serve such an end. Because of the small 19 Judd, Two Chaeo Canyon Pit Houses, p. 405.30 Judd, Archeological Observations North of the Rio Colorado, p. 8.? Cummings, Kivas of the San Juan Drainage,, p. 274.23 Morris, Exploring in the Canyon of Death, p. 272.31 Hough, Explorations of a Pit House Village at Luna, X. Mex. 14 BUREAU OF AMERICAN ETHNOLOGY [boll. 92 size of the holes in some of the houses, however, it seems more likelythat they had some such significance as that first suggested.There were other small holes in the floor which no doubt servedas storage places for small objects. At one side of the room wasa fairly large oval depression in which there was a metate or millingstone upon which the corn was ground into meal. In addition tothe depression for the metate there was a smaller one at one sidewhich no doubt served as a rest for the mano or hand stone used inthe milling operations.At the northern corner of the room there was a ridge of plasterextending out from the wall to the edge of the hole where the supportpost had been placed. It was quite low and narrow but marked off that portion of the floor space from the remainder of the room.What purpose it was intended to fulfill is not known. It was toolow to have been the retaining wall for a storage bin.One feature generally found in most of the houses was missingfrom this particular structure. There was no compartment or sec-tion partitioned off from the rest of the room. This was usuallythe portion nearest the door. As this feature is clearly shown insome of the other houses to be described, further' consideration ofit will be reserved for a later paragraph.A distinctive characteristic of many of the houses of the Chacogroup was the entryway on the south or southeast side.24 At house Ait was more elaborate than in most of the structures. The doorwayof the main room gave access into a short raised passage, which inturn opened into an oval room. The latter was roofed in much thesame fashion as the main dwelling, except that there were no interiorsupport posts. Because of its smaller size the poles sloping from thesides to the center could carry the weight of the superstructure with-out additional support. They probably met at the center, giving aconical shape to the roof. The exterior opening of this antechamberseems to have been at one side, the west, rather than opposite thedoorway into the house proper.One curious feature associated with the doorway to the ante-chamber was a large slab of stone set upright in front of it a short 24A good illustration of this feature in a slightly different form is to be found in threestructures on top of the bluff between McElmo and Yellow Jacket Canyons, just east oftheir juncture, in southwestern Colorado. They were first referred to by Mr. W. H.Jackson (Ruins of Southwestern Colorado and Adjacent Territory, pp. 413, 414), then byDoctors Kidder and Morley (Archaeology of McElmo Canyon, pt. 1, pp. 44, 45), and mororecently by Dr. J. W. Fewkes (Prehistoric Villages, Castles, and Towers of SouthwesternColorado, pp. 60-62). The writer inspected them on two occasions and is convinced thatthey represent a later stage in the period, perhaps are on the very border line betweenBasket Maker III and Pueblo I cultures.Dr. Ralph Linton (The Small Open Ruins of the Mesa Verde, unpublished manuscriptin files of Bureau of American Ethnology) found a quite similar feature in Earth Lodge Aon the Mesa Verde. The vestibule was rectangular in this instance, however, and some-what different from the ones in the Chaco, although comparable to them, boberts] SHABIK'ESHCHEE VILLAGE 15distance from the wall. (Fig. 1.) It may well have served as astep, although a little high for such a purpose. This was one of twohouses where such a slab was found in the antechamber. In manyof the others there was a similar stone in front of the main doorway.At the south side of the short passage leading from the ante-chamber to the main room there was an interesting suggestion of thebeginning of true masonry. Several large slabs had been placed flatabove the tops of the vertical stones in that portion of the wall.The horizontal slabs were embedded in thick masses of adobe mortar.They had been used to increase the wall height to a slight extent,and their upper surfaces probably formed a narrow shelf betweenthe edge of the wall and the slanting poles of the superstructure.There is evidence here that even at this early stage in house buildingthe idea of horizontally laid slabs and walls of true masonry wasbeginning to develop. It was some time, however, before it pro-gressed to the point where it was an important factor in the con-struction of domiciles. A similar feature is to be observed in twoother structures on this mesa top.The main room of house A was practically square. It measured15 feet 6 inches by 14 feet 6 inches (4.7244 m. by 4.4196 m.). Thefloor of the room was 2 feet 6 inches (76.2 cm.) beneath the originalground level. On the door side the slabs rose 3 feet (91.44 cm.)above the floor, while the plastered walls opposite were only 2 feet(60.96 cm.) high. The ceiling in the flat portion of the superstruc-ture must have been about 6 feet (1.8288 m.) above the floor. Theslant of the upper walls, as shown from the measurements of thepostholes in the adobe, was at an angle of 56?. At such an angle,and with the supporting posts at the distance that they were fromthe wall, the height of the ceiling can be fairly accurately postulatedas being about the figure given. The fire pit measured 1 foot 3inches (38.1 cm.) by 1 foot 10 inches (55.88 cm.), with an averagedepth of 6 inches (15.24 cm.). The oval hole near the fire pitmeasured 1 foot (30.48 cm.) on the long diameter, 8 inches (20.32cm.) on the short diameter, and had a depth of 6 inches (15.24 cm.).The doorway was 16 inches (40.64 cm.) wide and the passage leadingfrom the room to the antechamber was 2 feet 6 inches (76.2 cm.)long on the floor level. It no doubt had greater length along its topbecause of the slant of both the main room and antechamber super-structures. The floor of the passage was 2 feet (60.96 cm.) abovethe floor of the dwelling. The antechamber measured 11 feet by 8feet (3.3528 m. by 2.4384 m.). Its floor was 2 feet 6 inches (76.2cm.) below the original ground level. The height of the passageis not known and there was nothing in this particular ruin to indi- 16 BUREAU OF AMERICAN ETHNOLOGY [BULL. 92 cate what it might have been. From evidence secured in gome of theother houses, however, it is fairly safe to state that it probably wasabout 2 feet (60.96 cm.). House B This structure gave evidence of having been abandoned at somedate prior to that of the entire village. Most of the slabs which had o ? oV C c ?cd rTF'l \ i +-JL?ft Fig. 3.?Plan of House B. a, Oval pit in floor. b, Fire pit. c, Support posts, d,Deflector, e, Storage hole in floor, f, Sipapu. g, Bin erected after abandonmentof house, h, Remains of passagebeen used to line the excavation had been removed, probably for usein other structures. Some of them possibly found their way intothe walls of house A. From what remained of the dwelling, however,it was possible to determine that it had in a general way correspondedto the others. It had been irregular in shape (fig. 3), although itmight be referred to as roughly rectangular. koberts] SHABIK'ESHCHEE VILLAGE 17The interior features seemed to have been quite similar to thosealready mentioned in connection with house A. There was a centralfire pit, circular in this case, sipapu, post holes, depression for themealing bin, and indications of an entrance on the southeast side.There was sufficient evidence of the passage, but if it had at one timehad a chamber at its outer end all traces of the latter had vanished.In the interior of the room there was an upright stone slab set in thefloor in front of the doorway. This slab stood between the doorwayand the fire pit and recalls the deflector of the later kivas. It nodoubt served the same purpose?kept cold air from rushing throughthe doorway and directly onto the fire.During the excavation of the room it was found that the bin indi-cated by the slabs in the north corner (fig. 3) was of later date. Thebottom of this small cist was 1 foot (30.48 cm.) above the floor ofthe dwelling. At some time after the house had been abandoned astorage cist had been built on the same site. This feature was ap-parent in several of the dwellings and indicates that all of them werenot occupied contemporaneously.Another secondary feature was a burial which had been made inthe oval pit which had served as a mealing bin. The pit had not beenlarge enough for the body and instead of enlarging the hole the headhad been placed on the floor at one end and the feet projected in thesame fashion at the other.House B was slightly longer than it was wide. It measured 15feet 6 inches by 18 feet (4.7244 m. by 5.4864 m.) and averaged 2 feet(60.96 cm.) in depth. It must be borne in mind, however, that theactual measurements of the room were somewhat smaller at the timewhen it was occupied. The removal of the slabs made it impossibleto secure any but the dimensions of the hole. The circular fire pitwas 2 feet 6 inches (76.2 cm.) in diameter and had a depth of 1 foot5 inches (43.18 cm.). The lower 10 inches (25.4 cm.) had been filledwith clean sand. The sipapu was considerably north of the centerof the room. It was quite large, measuring 10 inches (25.4 cm.) indiameter. Its depth was 6 inches (15.24 cm.). The mealing bin was4 feet (1.2192 m.) long by 2 feet (60.96 cm.) in width and had adepth of 8 inches (20.32 cm.). The holes in which the support postsfor the superstructure had been set averaged 2 feet (60.96 cm.) indepth. The entrance passage appeared to have been about 1 foot 6inches (45.72 cm.) in width and 2 feet 6 inches (76.2 cm.) in length.House CFrom several different standpoints house C had been a very inter-esting structure. (PI. 2, b; fig. 4.) It differed from the two dwell-ings previously described in that no slabs were used to line the walls 18 BUREAU OF AMERICAN ETHNOLOGY [bull. 92 of the excavation. Plaster had been applied directly to the nativeearth. The excavation had been made in such fashion that therewas a small encircling bench at the top. In this bench were placedthe ends of the small poles which formed the framework for thesloping upper walls. (PI. 3, a.) The four posts which supportedthe superstructure were not placed in the floor some distance fromthe walls, as in the other houses, but were incorporated in the bench.5" 0.X \ ~ i i i ? > / *? Fig. 4.?Plan of House C. a, Ventilator shaft. 5, Ventilator passage, c, Supportposts, d, Corner bin. e, Fire pitThey were several incnes in front of the smaller poles, and on thewestern half of the room projected from the walls, so that there was adecided bulge in the plaster.The removal of the support posts from the interior of the roomitself would have been a decided advantage in giving greater floorspace, but it does not seem likely that in this particular case itwas the impelling motive behind such a construction. The obviousexplanation lies in the fact that the floor of the room was the caprock of the mesa top, and the digging of holes for setting posts wouldhave been an extremely difficult task. As a consequence the main ROBERTS] SHABIK'ESHCHEE VILLAGE 19 supports were placed in the wall. Just such an occurrence as thisno doubt led to the discovery that a better room was to be had whenthe roof was supported by posts which did not occupy valuable floorspace.Another circular pit dwelling showing the same feature, but prob-ably of a somewhat later date, was discovered in the floor of thecanyon about 1 mile (1.6 k.) east of Pueblo Bonito and excavated byMr. N. M. Judd in the spring of 1922. 25 In this case, however, themain support posts were completely incorporated in the bench ofthe structure and did not project from its face. The general featuresof this house seem to have been quite similar to the structure underconsideration, although there was no sipapu and, unfortunately, thewhole southern half of the dwelling had been cut away by an arroyoencroaching upon it at that side. Hence it is impossible to tellwhether or not it had the compartment and entryway features soapparent in the present group.The change from support posts set at some distance from the wallto those incorporated in it was not as abrupt in all localities. Thewriter excavated a number of ruins along the Piedra River in south-western Colorado where the support posts were placed against thewall but not included in it. In the general development of the housethe intervening step would appear to be the natural one taken. Thestage was finally reached where the support posts were actually apart of the upper wall. The latter feature was found in house J ofthe group in the Chaco Canyon. When walls of true masonry re-placed those of slabs and adobe the flat roof was carried by the wallsthemselves and additional stories became a possibility. It was un-questionably in this manner that the type of building so common inthe Pueblo periods was evolved.There was no sipapu in house C. The circular fire pit had beenhollowed out of the solid rock and had been plastered at the sametime as the floor.On the southeast side of the room there was a binlike compartmentwhich had been formed by the erection of an adobe wall containing afew large stones. (PL 2, b.) Most of the latter were at the topof the wall, where they formed a kind of coping. The large blockof stone at the bottom, on the right side of the doorway, which isshown in the photograph, was a part of the cap rock of the mesa.The northern and southern ends of the walls had upright slabs join-ing them to the main wall of the room. In the center of the lowpartition was an opening or doorway. This was directly in front ofthe opening into the passage.At the northern end of the small partition, on the room side ofthe wall, was a small inclosure formed by a ring of adobe plaster. 25 Judd, Two Chaco Canyon Pit Houses, p. 405. 20 BUREAU OF AMERICAN ETHNOLOGY [bull. 92 (PI. 3, b.) It no doubt had served the purpose of a storage placefor things connected with the daily life of the inhabitants. On theopposite side of the room, at the southern end of the low cross wall,again on the room side of the partition, was a small recess in themain wall. (PI. 3, c.) This probably served as a cupboard. It hadbeen constructed by digging a rectangular hole in the wall and plas-tering the earth over with adobe. It was the only example of a wallcupboard found in the entire village.The passageway and entrance of house C differed from the othersin the village in that it could not possibly have been used for anentrance. It was too small and could have served only as a ventilator.In fact it was distinctly comparable to the ventilator in the kivasof the Pueblo area. The floor of the passage was slightly higherthan that of the room and sloped upward toward the shaft at itsouter extremity. The lower part of this tunnel was cut in the caprock and bits of wood in the debris taken from it suggested that ithad had a timbered covering.The vertical shaft at the outer end of the passage had been slightlyoval in form and was lined with stone.The deflector associated with this passageway ventilator was notset in the floor as was the case in house B, but was a movable stonewhich closed the opening in the low cross Avail. At the time of ex-cavation this slab was in position in front of the opening in thepartition.As has been said before, the passage was too small to serve asan actual entrance, except possibly for small children, so it musthave served solely as a ventilator. With a fire burning in the pitin the center of the room the heat rising from it and passing out ofthe hole in the roof above would draw cold air down the shaftthrough the tunnel and into the room. Here, then, is what may beconsidered an example of the step which led from the true entranceto the ventilator. Because of the fact that the ventilator had atone time been the actual entrance to the dwellings there must havebeen some ceremonial significance attached to it and it was in allprobability retained at first more for this reason than because of its utility.Criticism frequently has been made of the term " ventilator " whenused in connection with similar features in the kivas of the Puebloperiods. The objection raised is that the Puebloans were not suf-ficiently versed in the laws of physics to construct ventilators. Asa matter of fact, they did not need to be conversant with such laws.They profited by their operation and no doubt came to realize in timethat the retention of the old passage, even though in a reduced andaltered form, had distinct advantages from a purely practical stand- BUREAU OF AMERICAN ETHNOLOGY BULLETIN 92 PLATE 2 sm -V a, Ilouse A 6, House CHOUSE REMAINS BUREAU OF AMERICAN ETHNOLOGY BULLETIN 92 PLATE 3 a, Bench at top of wall, showing holes for roofing poles 6, Corner storage box formed by an adobe ridgeHOUSE C c, Wall recess or cupboard Roberts] SHABIK'ESHCHEE VIT/LAGE 21 point. In fact, there is quite a little evidence to suggest that theventilator was retained long after it had ceased to have a markedsignificance ceremonially. There was so much about house C whichsuggested the later-day kivas that it might be called a protokiva.House C was somewhat smaller than houses A and B. Its diameteron the floor level was 9 by 9 feet (2.7432 m. by 2.7432 m.). Theaverage width of the small bench into which the roofing poles hadbeen set was 10 inches (25.4 cm.). There had been 24 of the smallpoles in the portion of the periphery extending from the northern endof the low cross wall, around the main portion of the room, to itssouthern end. Judging from the size of the holes in the top of thebench these small poles had an average diameter of 3 inches (7.62cm.). The larger support posts seem to have averaged about 6 inches(15.24 cm.) in diameter. They did not rest on the floor. Their buttends were, on an average, about 10 inches (25.4 cm.) above the floorlevel.The plastered wall of the room was 2 feet 6 inches (76.2 cm.) highon the west side and 3 feet (91.44 cm.) on the east side between thetwo support posts located in that section of the periphery. Thegreater height of the wall on the entrance or ventilator side of theroom seems to have had some significance, as it was found in practi-cally all of the ruins. It may, in fact, have had a direct bearing onone of the features in the later kivas. Discussion of this point, how-ever, will be reserved for a later section of this paper where thequestion of kivas will be considered.The ridge of plaster bounding the inclosure in the corner formedby the wall of the room and the partition was 4 inches (10.16 cm.)high, 6 inches (15.24 cm.) thick at the base, and an average of 3inches (7.62 cm.) thick at the top. The inclosure which it formedmeasured 1 foot (30.48 cm.) on the north side and 1 foot 3 inches(38.1 cm.) on the east. From the corner to the center of the curvedridge, on the floor, was 1 foot 3 inches (38.1 cm.).The cross wall averaged 2 feet (60.96 cm.) in height and 4 inches(10.16 cm.) in thickness. The opening in the center was 8 inches(20.32 cm.) wide.The doorway or opening into the passage or ventilator was 1foot 4 inches (40.64 cm.) high by 11 inches (27.94 cm.) wide. Thefloor of the passage was 3 inches (7.62 cm.) above the floor of theroom at the sill of the doorway. The passageway measured 3 feet6 inches (1.0668 m.) in length. The vertical shaft at its outwardend was 2 feet by 2 feet 3 inches (60.96 cm. by 68.58 cm.) indiameter. Its depth from the original ground level was about 2feet 6 inches (76.2 cm.).The wall cupboard was 10 inches (25.4 cm.) above the floor andmeasured 8 inches (20.32 cm.) high, 9 inches (22.86 cm.) wide, and 22 BUREAU OF AMERICAN ETHNOLOGY [BOLL. 92 10 inches (25.4 cm.) from front to back. It was 1 foot (30.48 cm.)from the end of the cross wall.The fire pit was 1 foot (30.48 cm.) by 1 foot 4 inches (40.64 cm.)in diameter and had a total depth of 8 inches (20.32 cm.). Thesides were not straight up and down, but sloping, so that the pitwas slightly bowl-shaped. c^ Of \ L?ft. Fig. 5.?Plan of House D. a, Antechamber remains, b, Passage, c, Support posts.d, Deflector, e, Fire pit. f, SipapuHouse DHouse D, like house A, had been fairly rectangular in form andshowed much in common with the first structure described. (PL 4, b;fig. 5.) The excavation had been lined with large stone slabs andthe superstructure had been supported on four posts set in the floornear the corners of the room. In this dwelling, however, the corner BOBHBTS] SHABIK'ESHCHEE VILLAGE 23 posts were not as far from the wall as in house A, a fact which is ofno great importance in this instance. The post at the southwestcorner was incorporated in the wall which formed the compartmenton the southern side of the room. This appears to have been a rathercommon practice, as many of the structures showed that the lowcross wall was so constructed as to include the support posts on thatside of the room. At the northeast corner a slab projecting from thewall suggested that there had been a bin inclosing the post therelocated. This feature is reminiscent of the inclosed corner in houseA, where a low adobe ridge was used instead of stone.There was a roughly circular fire pit in the center of the room.It had been lined with small stone slabs which were covered withplaster. North of this fire pit was the sipapu. It occupied a posi-tion very similar to those noted in the previously described houses.The binlike compartment on the southern side of the room had beenformed by the use of upright slabs covered with plaster. The centralopening in the cross wall was much larger in this dwelling. In theportion of the compartment at the eastern side of the room a largemetate was in place. This suggested that the inclosure may havebeen used as the family milling place.Originally there was a deflector slab in front of the doorway intothe passage, but it had fallen, and when the room was excavated itwas found lying flat on the floor in front of the opening. The im-pression in the plastered floor showed clearly where it had beenplaced.The floor of the passage in house D was somewhat higher thanthat of the room. A complete antechamber was not found at theouter end but there was evidence that there had been one. It haddiffered from the antechamber of house A, however, in that its floorlevel was the same as that of the passage. Apparently there hadbeen no excavation. Judging from what remains there were of theantechamber it had been considerably smaller than the one athouse A.Four posts placed at the corners where the side walls met the wallsof the room and the antechamber had supported the covering for thepassage. Portions of these posts still remained at the time of exca-vation. These posts had undoubtedly had forked tops and carriedthe long beams which supported the smaller cross poles of the roofand sides of the passage. The sides had been formed in part ofupright slabs. There were two of these slabs on each side of thepassage.House D measured 12 feet by 14 feet 6 inches (3.6576 m. by 4.4196m.). The wall on the north side was 2 feet (60.96 cm.) high andthat on the south 3 feet (91.44 cm.). The cross walls averaged 2 feet 24 BUREAU OF AMERICAN ETHNOLOGY [bull. 92(60.96 cm.) in height and the opening in the center was 4 feet 4inches (1.3208 m.) across. The cross walls at their interior endswere 3 feet 6 inches and 3 feet (1.0668 m. and 91.44 cm.) from thewall of the room.The doorway to the passage measured 1 foot (30.48 cm.) in widthand the slabs forming the side walls of the passage rose to a heightof 1 foot 6 inches (45.72 cm.) above its sill. The latter was 1 foot10 inches (55.88 cm.) above the floor of the room. The passage was4 feet 6 inches (1.3716 m.) in length. At its outer end the passagewidened to some extent and measured 1 foot 10 inches (55.88 cm.)across.The fire pit averaged 2 feet (60.96 cm.) in diameter and was 1foot (30.48 cm.) in depth. The sipapu was 1 foot (30.48 cm.) fromthe fire pit and measured 8 inches (20.32 cm.) in diameter. It wasslightly deeper than most of them, as its bottom was 10 inches(25.4 cm.) below the floor level.The deflector slab had stood about 10 inches (25.4 cm.) from thedoorway. Judging from the stone itself, the top of the deflectormust have been on about the same level as the sill of the doorwayinto the passage. The slab was 1 foot 6 inches (45.72 cm.) wideand 1 foot 10 inches (55.88 cm.) long.Judging from the holes, the support posts seemed to have averagedabout 6 inches (15.24 cm.) in diameter and to have been set at adepth of 2 feet 6 inches (76.2 cm.).House E Very little remained of house E beyond the excavation, fire pit,postholes, and deflector slab. (Fig. 6.) Almost all of the stoneslabs which had lined the interior of the excavated portion of thedwelling had been removed. As in the case of house B, the remainsof house E suggested that it had been abandoned and dismantledprior to the final desertion of the entire village.The circular form was more closely approached in house E thanin any of the structures described in preceding paragraphs. Thefour support posts were quite closely oriented to the cardinal direc-tions. The circular fire pit had a raised rim of plaster around it.This feature was observed in a number of the houses /ind is onewhich is found quite frequently in structures of the Late BasketMaker period.The raised rim around the fire pit was generally a continuationof the ridge of plaster in which the slabs forming the compartmenton the southeastern side of the room were set. In this particulardwelling these slabs had been removed, but the impressions madewhen they were placed in position in the ridge were quite evident. ROBERTS] SHABIK'ESHCHEE VILLAGE 25 These radiating ridges of adobe are not an uncommon feature inthe construction of the houses of the period. They seem to varyin the region to the west, however, in that poles were buried in thefloor, the adobe being applied to their upper surfaces and the ridgesformed in that way. Where that was done no upright slabs wereused to form bins or a true compartment at that side of the room.A good example of the use of logs in such a fashion is to be found in ?)? *\ ) (Be//? .^ 10'ft. Fio. 6.?Plan of House E. a, Oval depression in floor. 6, Adobe ridge, e, Holes forsupport posts, d, Deflector, e, Fire pit a structure located in cave 5, Cottonwood Canyon, Utah, which wasexcavated by Mr. N. M. Judd.26There were no indications of an entrance on the southeast sideof house E. This portion of the ruin was on a rather steep slopeof the mesa top, where it was subject to the erosive action of con-siderable drainage water and what might have remained at thetime the house was dismantled could easily have been washed away. * Judd, Archeological Observations, p. 113 ; also fig. 29.9215??Bull. 92?29 3 26 BUREAU OF AMERICAN ETHNOLOGY [bull. 92The deflector slab, which was braced by a second stone, was still inposition in the space between the fire pit and the side of the roomwhere such an entrance would have been.Between the deflector slab and the fire pit was a slight oval-shapeddepression in the floor. Similar depressions were noted in otherhouses. Just what it was for is not clear. The most logical explana-tion is that the foot of a ladder rested in it. As suggested in a pre-ceding paragraph, the smoke hole at the top of the superstructure waspossibly used on occasions as an entrance, and when such was thecase a form of ladder would have been essential. A depression ofthe type under consideration would have prevented the ladder fromslipping on the smooth surface of the plastered floor. One argumentagainst such a belief, however, is that in some of the kivas of thePueblo periods the ladders have been found crossing to the oppositeside of the fire pit. Where this was true the sipapu was at a greaterdistance from the fire pit and the ladder did not interfere with it.If in the case of the houses in Shabik'eshchee village the ladder hadrested on the floor at the north side of the fire pit, it would have beena hindrance to the performance of any ceremonies connected withthe sipapu, provided that is what the circular hole north of thefire pit was used for. It is ajl pure conjecture, but it seems morelogical to suppose that the depression between the deflector and thefire pit was for the base of the ladder than to try to explain it on thebasis of a ceremonial or other utilitarian function.The floor on the north side of the fire pit was so damaged that itwas impossible to determine the limits of the sipapu. There wereindications that one had been present, but its rim was missing.The excavation for house E was 14 feet (4.2672 m.) in diameter.When the slabs which had lined it were in place it was somewhatsmaller. The few which were in position along the northern arcaveraged 2 feet (60.96 cm.) in height. All of the stones had beenremoved on the south side, where the earth wall rose but 1 foot 6inches (45.72 cm.) above the floor level. There was every indicationthat a considerable portion of the wall had been washed away. Thecircular fire pit was 2 feet (60.96 cm.) in diameter and 5 inches (12.7cm.) deep. Its raised rim was 2 inches (5.08 cm.) above the floor andwas 6 inches (15.24 cm.) thick on an average. The support postsseem to have averaged 8 inches (20.32 cm.) in diameter and wereset at a depth of a little over 2 feet (60.96 cm.).House FOne of the most interesting problems encountered during theexcavations was in connection with house F. It was found that asmaller structure had been erected inside of a much larger one and ROBERTS] SHABIK'ESHCHEE VILLAGE 27 that the former had incorporated but a small portion of the latterin its construction. (Fig. 7.) Indications were that the older dwell-ing, the larger one, had been partially burned and that subsequentto the fire the smaller one had been built inside of it. This secondaryconstruction saved the laborious effort which a new excavation wouldhave entailed.The smaller, more recent dwelling was the first one to be dis-covered, and it was only after it had been completely cleared ofdebris that suggestions of another structure having occupied the site 70 ft Fig. 7.?Plan of the F House group, showing both structures in complete relationship were noted. To simplify discussion the houses will be describedin the order in which they were uncovered. Both were as fine ex-amples of the house type of the period as could be desired. Thesmaller of the two will be referred to as house F, while the largerwill carry the designation of house F-l.The subterranean portions of the walls of house F were con-structed entirely of stone slabs. (Fig. 8.) These had been coveredwith a coating of plaster. It was again found that the superstructurehad been supported on four posts. The two on the entrance side ofthe room had been incorporated in the partition which formed thecompartment about the doorway. The post near the southern corner 28 BUREAU OF AMERICA-NT ETHNOLOGY [bull. 92had been placed in the fire pit of the original house occupying thelocation, and for that reason had been more thoroughly braced withstones than was generally found to be the practice.It had not been necessary to line the fire pit with stone because ithad been pecked out of the mesa top cap rock. The pit was oval inshape and, like the one in house C, had been finished with a coating ofplaster over the rock. Close to the fire pit was the small circular Q a\ ~L?ft- Fio. 8.?Plan of House P. a, Antechamber. 6, Passage, c, Support-post holes, e,Fire pit. /, Sipapuhole which has been likened to the sipapu. The latter also was cutinto the living rock. (PI. 4, a.)At some distance from the so-called sipapu, between it and thewall on that side of the room, was another circular hole, smaller in size. An explanation for the second hole is not readily forthcom-ing. The same feature was found in one or two additional struc-tures. There was another circular hole, probably a place for thestoring of small articles, in the floor in front of the north supportpost for the roof. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 92 PLATE 4 >., / ^"rv a, Sipapu in floor of House F 6, House DHOUSE REMAINS] BUREAU OF AMERICAN ETHNOLOGY BULLETIN 92 PLATE 5 a, Ilouse F-l during excavation. Rough masonry may be noted at sides of doorway. Uouse F inleft foreground v 6, House HHOUSE REMAINS Roberts] SHABIK'ESHCHEE VILLAGE 29The compartment on the southeastern side of the room was some-what different from those previously described. In house F thepartition was continuous. It did not have the break or doorwayin the center opposite the entrance to the passage. The necessity forsuch an opening was removed by the manner in which the cross wallhad been constructed. At each end where it joined the walls of theroom there was the usual upright stone slab the same height as themain wall. Between these slabs the wall was considerably lower, somuch so, in fact, that a person could have stepped over it with ease.In addition to this the deflector had been incorporated into thepartition. It stood in its' proper position between the entrancedoorway and the fire pit.The deflector differed, also. Instead of being a stone slab, asin the other houses, it was of pole and plaster construction. Twosmall poles had been placed upright in the low cross wall and thespace between them was filled with adobe mortar. Indications inthe debris of the room were that the top of the deflector had been atabout the same height above the floor as the slabs on either side ofthe doorway into the passage.The rather short passage opened into an antechamber which wassomewhat rectangular in form. The latter shape was more or lessimposed upon the builders by certain features in the earlier house.The outer wall of the antechamber, the one opposite the passage,was formed by the wall of the older house. The side of the ante-chamber toward the northeast was formed by slabs which had beenset in the compartment of the older house, ostensibly for making abin in that portion of the room. Because of this the size and shapeof the antechamber for house F was decidedly limited. The floorof the passage was higher than the floor of the room, at the doorway,and sloped upward to the antechamber. This higher, upward-sloping floor was intentional, as it had been necessary to fill in thepassage and antechamber to secure such a result. The floor level ofthat portion of the older house had been the same as that of themain room in house F.On the floor level house F measured 12 feet 6 inches by 11 feet(3.81 by 3.3528 m.). On the side opposite the doorway the slab wallwas 1 foot 7 inches (48.26 cm.) high and at the entrance to thepassage it was 3 feet (91.44 cm.) from the floor level to the tops ofthe slabs.The cross wall of adobe plaster which formed the compartment onthe entrance side averaged 8 inches (20.32 cm.) high, between theupright slabs at its ends, and 3 inches (7.62 cm.) wide. As men-tioned previously, the height of the deflector was probably aboutthe same as that of the slabs on that side of the room. There is nodefinite information, however, on that point. 30 BUBEATJ OF AMERICAN ETHNOLOGY [BULL. 92The doorway into the passage was 1 foot 6 inches (45.72 cm.)wide. The floor of the passage was 4 inches (10.16 cm.) higher thanthat of the room, making a slight step at the doorway. The passagewas 1 foot (30.48 cm.) long and the antechamber measured 3 feet(91.44 cm.) by 2 feet (60.96 cm.) on an average. Because of theirregular shape of the latter the measurements are somewhat vari-able. The floor of the antechamber was 2 feet 8 inches (81.28 cm.)beneath the old ground level. Fig. 9.?Plan of House F-l. a, Antechamber, t, Passage, c, Holes for supportposts, d, Deflector, c, Fire pit. f, Oval hole in position normally occupiedby sipapu. g, Second fire pit. h, Fire pit placed iu antechamber at datesubsequent to abandonment of houseThe fire pit in the room measured 2 feet by 1 foot 6 inches (60.96by 45.72 cm.) on two diameters and had a depth of 6 inches (15.24cm.). The sipapu was 7 inches (17.78 cm.) in diameter and 7 inches(17.78 cm.) deep. It was 4 inches (10.16 cm.) from the fire pit.The smaller hole was 1 foot 5 inches (43.18 cm.) from the sipapu andwas 5 inches (12.7 cm.) in diameter. It was 3 inches (7.62 cm.) indepth. The holes for the support posts averaged 10 inches (25.4cm.) in diameter and had an average depth of 2 feet 6 inches(76.2 cm.).House F-l differed little in its general aspects from the smallerstructure which had been erected within its confines or from theother houses described. It was one of the largest excavated. The Roberts] SHABIK'ESHCHEE VILLAGE 31 walls of the pit were only partially lined with stone slabs. Thegreater part of the north wall and portions of the eastern one weresimply plaster. (Fig. 9.)The septilateral fire pit in the center of the room had been linedwith small stone slabs the tops of which were flush with the floor.The large oval-shaped pit in the floor near the fireplace occupiedthe position which the sipapu ordinarily did. Perhaps it served thesame function, in spite of its greater size. No excavation had beenmade for this pit. Instead, the housebuilders had taken advantageof a crack in the cap rock of the mesa. The crack itself was muchlarger than the finished pit but had been filled in partially with stonespalls and adobe mortar. The walls of the pit had been plastered care-fully and it would have made an ideal storage place. It had beenwithin the confines of house F also, but it had been filled with largestones and earth and the floor of the later house extended over it.The same was true of the rather shallow, oval pit near the bin atthe eastern corner. The latter pit gave every evidence, however, ofhaving been used as a fireplace.The compartment at the southeast side of the room had been moreelaborate than in many of the houses. The usual row of partitionslabs had been set in the floor, but in addition to these other stoneshad been used to make small bins in the compartment. Part of thispartition had been removed to make way for the later house, and asa result a portion of the east cross wall was missing. Its formerpresence could be traced in the grooves which it left in the floor.The latter indicated how far it had extended into the room. On thewest side the support post for the superstructure had been incor-porated into the cross wall but on the east side the post had only beenpartially inclosed. As in so many of the houses, the other two postswere inclosed in small stone boxes formed of slabs whose tops wereflush with the floor level.The wall of house F-l on the side where the doorway opened intothe passage was interesting because it had some masonry, albeit of acrude sort, on top of the slabs. This feature was the same as notedin the construction of house A. Its prime purpose seems to havebeen that of leveling the wall near the doorway. Some of the slabswhich were used in that portion of the wall were not as high as themajority and it was necessary to add to them. This was donethrough the use of the horizontally laid stones in a crude form ofmasonry. The fact that this was done at the doorway might beconsidered as indicating a desire for a little better finish to the wallnear that opening. The character of this masonry is clearly shownin the photograph taken at the time the ruin was being excavated.(PI. 5, a.) 32 BUREAU OF AMERICAN ETHNOLOGY [bull. 92The entrance to the passageway was larger than in the smaller,later house. There was a slight step and the floor of the passage-had a gentle upward slope toward the antechamber. The deflectorbetween the doorway and the fire pit was quite large, and because ofits size and the fact that the floor of the house was on the cap rock ofthe mesa it had been necessary to brace it with other stones. Thesewere placed on the side toward the fire pit and with the adobe plasterserved to keep the deflector in an upright position.The antechamber of house F-l had been quite large, although notas large as the one of house A. It was not intact at the time ofexcavation, however, as some of the slabs which had lined the pithad been removed. Their imprints were still to be seen in the earthwalls, and there is no question that the entire excavation had been solined. The large slab projecting from the east side of the ante-chamber seemed to have been of a later date than the main portionof the structure. It stood considerably higher than the slabs whichlined the antechamber pit and its base was well above the floor of thelatter. It probably had been associated with a fire pit which hadbeen placed in the antechamber at a date subsequent to its abandon-ment. This fire pit, like the cist described in connection with houseB, had its floor at a much higher level than that of the antechamber.The space between was filled with debris such as is wont to accumu-late around an Indian village. It was in large part house sweepings,ashes, sand, containing bits of broken bones, and a few potsherds.It is quite possible that this fire pit was used by the occupants of thelater house, F.On the east side of the fire pit, at about the same level as its bot-tom, was a burial. The body had been placed close to the pit andwas within the confines of the old antechamber. Advantage was nodoubt taken of the ease with which a grave could be dug in thedebris which had accumulated in such a place. Excavations in the soil which topped the mesa must have been quite difficult with thecrude implements with which the people had to work. Conse-quently, wherever possible, they seem to have deposited the bodiesof their dead in accumulations of refuse.At the west side of the antechamber there was a feature whichwas quite suggestive of the antechamber in house A. An old, worn-out metate or milling stone had been set on end in the floor in aboutthe same position as the large slab noted in the house A entryway.It was much smaller, however, than the one in the latter. Such astone would be of considerable assistance to a person stepping downinto the antechamber or stepping up to the ground level when leav-ing it. The position of the stone would suggest that here, as in thefirst house discussed, the door to the antechamber had not been on a Roberts] SHABIK'ESHCHEE VILLAGE 33 direct line with the passage but was at one side, a feature whichwould have had distinct advantages from several points of view.From the present evidence, however, it can only be suggested thatin the case of the two antechambers, those of houses A and F-l, theentrance to the chamber may have been at one side rather thandirectly in front of the passage into the main room of the structure.The break in the southern wall of the antechamber is quite sug-gestive of a doorway, but there was nothing to indicate that such anopening had been present. The slab at the east side was in a posi-tion which might be considered indicative of a jamb, but otherfactors favor the belief that this break in the wall of the pit was dueto the removal of slabs from that point rather than to its havingbeen an entrance to the antechamber.House F-l had been the largest domicile in the entire village. Itmeasured, on the floor level, 21 by 18 feet (6.4008 by 5.4864 m.). Thewall on the northwest side averaged 1 foot 6 inches (45.72 cm.) inheight and on the southeastern side of the room, from corner boxto corner box, it was 3 feet 6 inches (1.0668 m.) in height. Nearthe center the upright slabs were 3 feet (91.44 cm.) high and thecrude masonry extended 6 inches (15.24 cm.) above their tops.The fire box measured 2 feet 10 inches (86.36 cm.) by 2 feet(60.96 cm.) and was 1 foot (30.48 cm.) in depth. The oval pit in thefloor, north of the fire pit, was 2 feet 6 inches (76.2 cm.) long and 1foot (30.48 cm.) wide. Its rounded bottom was 2 feet (60.96 cm.)below the floor level.The oval fire box near the eastern corner of the room measured3 feet (91.44 cm.) the long way and 1 foot (30.48 cm.) in width.It was 10 inches (25.4 cm.) deep. On an average it was 1 foot (30.48cm.) from the wall of the room and 1 foot 2y2 inches (36.83 cm.)from the corner box.At the southeastern corner the bin was 4 feet 4 inches (1.3208m.) in length and 2 feet 6 inches (76.2 cm.) wide. It had beendivided into two compartments by the use of an additional slab.One of these compartments was approximately twice as large as theother. The traces of the partition wall, in the floor, showed that ithad extended into the room 3 feet (91.44 cm.) from the corner ofthe storage bin. This made a total length of 7 feet 4 inches (2.2352m.) for the compartment wall on that side of the room. The openspace between the end of this wall and that on the opposite side was7 feet 6 inches (2.286 m.).The partition wall on the west side of the room was 6 feet 6 inches(1.9812 m.) long. A small bin had been formed in the corner bysetting two slabs of stone, one an old metate, across the inclosure,3 feet (91.44 cm.) from the west wall of the room. 34 BUREAU OF AMERICAN" ETHNOLOGY [bull. 92The deflector slabs stood 10 inches (25.4 cm.) from the doorwayinto the passage. The main slab was 2 feet (60.96 cm.) high and 1foot 6 inches (45.72 cm.) wide. It averaged 2 inches (5.08 cm.) inthickness.The doorway into the passage was 2 feet (60.96 cm.) wide. Thepassage was 3 feet (91.44 cm.) long on the east side and 3 feet 6inches (1.0668 m.) along the west wall. The passage widened whereit opened into the antechamber, measuring 2 feet 6 inches (76.2 cm.)from wall to wall. The floor of the passage was 3 inches (7.62 cm.)above the floor of the room at the doorway and sloped slightlyupward to the floor of the antechamber. The latter was 8 feet 6inches (2.5908 m.) in diameter on its long axis and 6 feet (1.8288m.) on the short. The floor of the antechamber was about 4 inches(10.16 cm.) higher than that of the house. The slabs which were still in position rose to a height of 3 feet (91.44 cm.) above thefloor level.The fire pit, which had been placed in the antechamber at somedate subsequent to its abandonment, was roughly circular in form.It had an average diameter of 4 feet (1.2192 m.). The bottom ofthis pit was 1 foot 6 inches (45.72 cm.) above the floor of the ante-chamber.With the exception of a few features, the two houses had verylittle in common. Both had had the same floor. Three slabs in theeast wall of the old house had been incorporated into the newerstructure. The same hole had been used for the northeast supportpost in each house. A portion of the bin in the southeast corner ofthe older building had formed one side of the antechamber of thesmaller domicile. Another side of the antechamber consisted of twoslabs in the wall of the larger dwelling. Beyond these features,however, the buildings had been separate structures.House GThis structure was very similar in its general aspects to the othersin the village. (Fig. 10.) It was somewhat smaller than some ofthose already described but in all respects had been a characteristicdomicile. The excavation had been lined with the customary stoneslabs, although at the time of excavation a number of them weremissing. Some of them had fallen in and were lying on the floor ofthe room while others had been completely removed. Here again,perhaps, material had been removed from a dwelling no longeroccupied for use in other constructions.An interesting feature associated with the apparent abandonmentof house G was the finding of quite definite indications that the ante-chamber had been worked over to a certain extent to serve as a storage BOBERTS] SHABIK'ESHCHEE VILLAGE 35 bin. This was the only example of such a procedure found in theentire community.In the main room the various features indicated that the usualsupport posts had been used to carry the superstructure. The fourholes near the corners still retained portions of those posts in theform of fragments of decayed wood. The sipapu was observed in itscustomary position near the fire pit. The fire pit was somewhatmore irregular in shape than those which have been described for theother houses. It had been lined with small stone slabs, the tops of *jpAFig. 10.?Plan of House G. a, Antechamber, b, Possible short passage, c, Holesfor support posts, d, Deflector, e, Fire pit. /, Sipapu which were flush with the plastered floor, and had been paved withthe same material. The pit was much closer to the opening in thecross wall forming the compartment on the southeast side of the roomthan those in the other structures. In fact the rim on the passage-way side of the pit was almost on a line with the two ends of theslab partition. The deflector consisted of two slabs, one having beenplaced in such a position as to support the other.The passage between the main room and the antechamber had beenvery short. It also seems to have been on the ground level. The 36 BUEEAU OF AMERICAN ETHNOLOGY [bull. 92 floor level of the antechamber was slightly higher than that of themain room and it would have been possible, without a great deal ofinconvenience, to step over the higher section in passing from oneroom to the other.The antechamber had been quite large, almost as large as the onefor house A, and was closer to the main portion of the dwelling thanin most cases. There was clear evidence that the doorway to thepassage had been blocked up or filled in. This suggested the beliefthat the antechamber may have been used as a storage bin after thehouse was no longer occupied. The debris filling the two excavationswas totally different in character.The house portion of the structure had been filled in with refuseof the type commonly found in dump heaps, while the antechamberpit contained clean, wind-blown sand. The latter would suggest thatthe antechamber had been in use down to the end of the occupationof the village. If the antechamber had been abandoned prior to thedesertion of the village, it too would have contained debris of accumu-lation such as always collects around an Indian settlement. Themodern Pueblo Indians are quick to throw their refuse into rooms nolonger used, and it is quite likely that the early communal dwellerswere followers of a similar practice. There was considerable car-bonized corn on the floor of the antechamber, which was good evi-dence of its having been used as a storage place.House G measured, on the floor level, 11 by 12 feet (3.3528 m. by3.6576 m.). The wall on the northwestern side was 2 feet 6 inches(76.2 cm.) high and on the southeast it was 3 feet (91.44 cm.) high.The sipapu was 1 foot (30.48 cm.) in diameter and 6 inches (15.24cm.) deep. It was 4 inches (10.16 cm.) from the edge of the fire pit.The latter measured 2 feet 4 inches (71.12 cm.) by 2 feet (60.96 cm.)and had a depth of 1 foot (30.48 cm.).The opening between the ends of the compartment walls was 4 feet(1.2192 m.) wide. The inner end of the south wall was 2 feet (60.96cm.) from the main wall of the room and that of the north partitionwas 1 foot 10 inches (55.88 cm.) from the same wall. The averageheight of these walls was 2 feet 6 inches (76.2 cm.).The main deflector slab was 10 inches (25.4 cm.) from the wall ofthe room. The base of the bracing slab was 5 inches (12.7 cm.)from the wall. The deflector was 1 foot 6 inches (45.72 era.) wideand 1 foot 6 inches (45.72 cm.) high.The passage seemed to have been about 1 foot (30.48 cm.) long.Its width could not be determined correctly. It must have been ap-proximately 2 feet (60.96 cm.), however.The antechamber measured 8 feet 6 inches (2.5908 m.) by 7 feet(2.1336 m.). Its average depth was 2 feet 6 inches (76.2 cm.). ROBERTS] SHABIK'ESHCHEE VILLAGE 37Three of the support posts averaged 9 inches (22.86 cm.) in diam-eter, and the fourth one was 6 inches (15.24 cm.) across. The aver-age depth to which they had been sunk in the floor was 2 feet(60.96 cm.). House HThere were certain individual characteristics in house H, but theywere of such a nature that the general aspects of the dwellings coin-cided quite closelywith those of theother structures. C^^^(PI. 5, I; fig. 11.)Stone slabs had beenused to face the en-tire periphery ofthe excavation. Thesuperstructure hadbeen supported onfour upright posts,there was a sipapu,the usual fire pitnear the center ofthe room, and therehad been a compart-ment on the south-east side.House H differedfrom those previous-ly described in hav-ing had no deflector.Furthermore, therewas no indication inthe plaster on thefloor that there everhad been one. Inaddition to this, notraces could befound of an entry-way or passage onthe southeast side. Fig. 11.?Plan of House H. a, StorageOval depression in floor for end of ladder,support posts, e, Fire pit. f, Sipapu hole in floorc, 6,Holes forThis is not definite proof, however, that onenever existed. In view of the fact that so many of the houses had hadsuch a feature in their construction it seems more in keeping with thecharacteristic house type to believe that house H had an entryway.Had it been constructed wholly of perishable materials and on theground level all traces of it could easily have been obliterated during 38 BUREAU OF AMERICAN ETHNOLOGY [bull. 02 the centuries which must have passed since the village was deserted.On the other hand, it would be more interesting to believe that ithad not had the usual entrance and that access to the interior hadbeen possible only by means of a ladder extending through the smokehole at the top of the superstructure. If the latter were true, houseH would be an example of the step which was made when the oldside entrance was abandoned and the only entrance to the housewas through the roof, an outstanding feature in the rectangularabove-ground dwellings of the earlier Pueblo periods which followed.Certain evidence suggested the belief that house H was one of thelate structures in the village and it is quite possible that there is tobe observed in it one of the stages in the transition from the typicalLate Basket Maker structure to those of the early Pueblo period.The sipapu and fire pit were cut into the cap rock of the mesa andhad been plastered when the adobe had been applied to the surfaceof the floor. The fire pit had a raised rim of plaster. The latter wasa continuation of the ridge into which the slabs of stone forming thecompartment on the southeast side had been set.In the floor at the southeast side of the fire pit was an oval depres-sion similar to that in house E, which may have been the place wherethe lower end of the entrance ladder was placed. Some such pro-vision would have been necessary here, as well as in the precedinghouse, to keep the ladder from slipping on the smooth surface of thefloor. Inasmuch as the cap rock served as the floor it would havebeen rather difficult to embed the foot of the ladder in the floor, aswas frequently done in houses of a later period. An oval depressionwould serve the purpose and would be much easier to make than aslanting hole in the stone. The plaster which covered the rock wassomewhat broken and roughened on the side of the depression wherea ladder could have rested.The slabs forming the eastern end of the compartment wall hadbeen broken off. The same was true of one of the slabs near thewestern end. The southern support post for the roof had been incor-porated in the partition wall but the eastern post had not. Thelatter stood against the low cross wall, inside of the compartment.There was a small circular hole, well plastered, in the floor at thesouthwestern end of the compartment. It had probably been usedfor storage purposes.The base of the northern supporting post in this structure showedan interesting treatment. It had been surrounded by a small raisedring of plaster. * This same feature was observed in house K, whereit was also a north post which had been so treated. None of theother posts gave any evidence of having had this ring of plaster atthe base. Its usefulness as a brace and support for the post is readily robert3] SHABIK ,ESHCHEE VILLAGE 39 apparent, but why it should have been applied only to the post atthe north side of the house is not. There may have been some specialsignificance attached to it or it may have been just an individualpeculiarity so expressed in the construction of the two houses inwhich it was found.House H measured 12 feet 6 inches (3.81 m.) by 11 feet 6 inches(3.5052 m.) on the floor level. There was no marked variation in thewall height. On the north side the wall was 2 feet 6 inches (76.2 cm.)high. The south side, which in a majority of the houses was higherthan the north, measured the same, 2 feet 6 inches (76.2 cm.).The sipapu was 8V2 inches (21.59 cm.) in diameter and 6 inches(15.24 cm.) in depth. It was 6 inches (15.24 cm.) from the fire pit.The latter was more heart-shaped than circular, but its greatestdiameter on two axes was 2 feet (60.96 cm.). Its depth was 6 inches(15.24 cm.). The raised rim around the pit was quite damaged, butit was possible to ascertain that it had been 4 inches (10.16 cm.) inwidth and approximately the same height. The oval depression onthe southeast side of the fire pit was 1 foot 4 inches (40.46 cm.) longand 6 inches (15.24 cm.) wide. It was quite shallow, as its greatestdepth was but 3 inches (7.62 cm.).The opening in the cross wall was 4 feet (1.2192 m.) wide. Theeastern part of the partition forming the compartment had been 3 feet6 inches (1.0668 m.) long and the western part measured 3 feet(91.44 cm.). The single slab standing in the partition was 1 foot 5inches (43.18 cm.) high. The partition wall had been slightly higher,however, as the plaster which had covered the slabs rose severalinches above the tops of the stones. The inside end of the eastern sec-tion of the partition was 2 feet 6 inches (76.2 cm.) from the south-eastern wall of the room. The inside end of the partition on the westmeasured 2 feet 10 inches (86.36 cm.) from the wall of the room. Thehole in the floor at the western end of the compartment was 8y2inches (21.59 cm.) in diameter and 7 inches (17.78 cm.) in depth.The holes for the supporting posts averaged 6 inches (15.24 cm.)in diameter and were about 2 feet (60.96 cm.) deep.House I This structure showed greater variation from the general housetype than any other in the village. Most of the features which mightbe considered as common were missing, and the general plan wasquite simple. (Fig. 12.) The earth walls of the excavation, whichwere roughly circular in form, had been treated with a coat of adobeplaster. No slabs had been used in the construction of the lowerportion of the house. The superstructure had been supported onfour posts, as in the other domiciles. The fire pit was circular in 40 BUREAU OF AMERICAN ETHNOLOGY [BULL. 9S f form and lined with plaster. No stones were used to line its interior.There was no indication in the floor of a sipapu. The compartment-like inclosure found on the southeastern side of so many of the roomswas absent in this dwelling, and there were no indications that therehad ever been one. There was no deflector, and no trace whatsoevercould be found of an entryway or passage.An interesting feature of this house was the storage bin or recessat the western side of the room. Its form is readily determined fromthe profile view of the ruin shown in Figure 12. At the ground levelthe horizontal outline ofthe structure was thatof a circle, although arather crudely shapedone. The recess hadbeen cut out of the earthwall in such a way thatit was roughly dome-shaped. Plaster hadbeen applied to the sidesand top of this excava-tion, and a very goodstorage place thus ob-tained. At the center ofthe opening into this re-cess was a circular de-pression in the floorwhich also could haveserved for storage pur-poses. This was the onlyhouse in the entire vil-lage which showed sucha recess. Wall niches onthe floor level are notunknown, however. Mr. Morris found them in structures of thesemisubterranean type 70 miles east of La Plata River in southernColorado,27 but in the latter the bottoms of the bins were slightlybelow the floor level. Doctor Kidder and Mr. Guernsey found awall niche on the floor level in a structure which they excavated inpart in White Dog Cave in northeastern Arizona. 28House I varied slightly in diameter. On an east-and-west lineit measured 9 feet 8y2 inches (2.9591 m.) and on the north-southdiameter 9 feet 6 inches (2.8956 m.). 27 Morris, Preliminary Account of Antiquities Between the Mancos and La Plata Rivers,p. 186.28 Guernsey and Kidder, Basket-Maker Caves, p. 25, fig. 7 b. Fig. 12.?Plan of House I. a, Wall recess. T), Storagepit in floor, c, Support-post boles, e, Fire pit koberts] SHABIK'ESHCHEE VILLAGE 41There was some variation in wall height. On the east side ofthe room the plaster rose to a height of 1 foot IV2 inches (34.29 cm.)above the floor, and on the west, where the recess was located, itmeasured 1 foot 6 inches (45.72 cm.) in height. This variation wasno doubt due to the fact that there was quite a slope to the hill atthat point, and in order to secure a floor which was fairly level theexcavation would have had to be somewhat deeper on the westernside. From the top of the recess to the top of the wall, at the centerof the recess, was 6 inches (15.24 cm.).The recess, on the floor level, measured 5 feet (1.524 m.) on itslong axis. The back wall of the recess was 3 feet (91.44 cm.) froma line connecting the two sides of the opening into the room. Thelatter measured 4 feet 9 inches (1.4478 m.) at the point of its greatestwidth.The circular hole in the floor in front of the opening into therecess was 1 foot 2y2 inches (36.83 cm.) in diameter and had a depthof 6 inches (15.24 cm.). The eastern edge of this pit was 2 feet 6inches (76.2 cm.) from the western edge of the fire pit. The latterwas 2 feet (60.96 cm.) in diameter and had an average depth of 4inches (10.16 cm.).Because of the small size of the structure large support posts forthe superstructure were not necessary, and judging from the sizeof the holes in which they were set they must have measured, on theaverage, about 6 inches (15.24 cm.) in diameter. The average depthof these postholes was slightly over 2 feet (60.96 cm.).House J House J when excavated revealed certain features which were ofdecided importance from the standpoint of their bearing on the typeof house construction which quite frequently has been observed inthe first of the Pueblo periods. (Fig. 13.) In erecting this dwell-ing the builders made the usual excavation, roughly oval in contourin this instance, but instead of applying the plaster directly to thenative earth walls or lining the pit with stone slabs they made useof a series of upright posts, set in the floor around the periphery.These posts formed a framework for the walls, which consisted ofthick adobe plaster placed between and around the posts. This is theso-called jacal type of construction which has been found in quite afew ruins of the Pueblo I phase of southwestern sedentary cultures.The introduction of this type of wall construction apparently didaway with the sloping upper portion of the walls noted in the houseswhich have been described in preceding paragraphs. The posts setaround the sides of the excavation were probably tall enough to pro-9215??Bull. 92?29 1 42 BUREAU OF AMERICAN ETHNOLOGY [bull. 92 vide the necessary above-ground portions of the walls. With wallsof this type, and a house no larger than the one under consideration,interior support posts for the roof would not be necessary. Thebuilders of house J seem to have realized this fact, because there wasno indication that such posts had been used in the structure. Theroof timbers apparently were carried by the walls. There is then,in house J, what may be considered as an illustration of one of thesteps which led from the old type of domicile with sloping walls and 10'/* Fig. 13.?Plan of House J. a, Antechamber, b, Doorway to room,posts used in wall construction, d, Possible deflector, e, Fire pit. c, Holes forf, Sipapu flat roof, which is so marked a characteristic of the Basket Maker IIIperiod, to one of the forms of dwellings with straight side walls anda flat roof which is found in the first Pueblo period.Numerous examples of structures with walls of the type justdescribed have been observed in the San Juan area. Mr. Morrisfound rather definite suggestions of them in the La Plata district,east of Spring Canyon, in southwestern Colorado. 29 The writer un-covered similar features in the remains of structures excavated along ^Morris, Antiquities of Southwestern Colorado, p. 187. boberts] SHABIK'ESHCHEE VILLAGE 43 the Piedra River in the same State,30 while Mr. Judd found a slightvariation of this type of wall construction in southern Utah.31 Inthe latter dwellings, however, the four central support posts wereretained.Remains of most of the posts which had formed the framework ofthe walls of house J were found but, due to the fact that the floorat the northern end of the room had been damaged greatly, the exactnumber and location of several, possibly four or five, could not bedetermined. Hence in the ground plan of the house they are omitted.The sipapu was present at the north side of the fire pit, but at asomewhat greater distance from it than in most of the structures ofthe village. The fire pit was circular in form and was located inabout the center of the room. There was no deflector or fire screenbut there had been an entryway with an antechamber at the south-east corner.It is possible that there had been in this structure, as in house C,a movable deflector, so to speak. A large stone slab was found inposition across the opening to the antechamber. Perhaps it merelyhad been placed in position when needed and not set in the floor.The passage from the antechamber to the main room of the dwell-ing was so short that it could hardly be called anything but a door-way. The floor of the antechamber was higher than that of theroom and there was a slight step at the doorway. There was asecond step from the sill of the doorway to the floor of the; ante-chamber. The stone slab which has been considered as a possibledeflector of the movable type was resting on the sill of the doorwayand against the riser of the second low step.The excavation for house J measured 13 feet (3.9624 m.) by 10 feet6 inches (3.2004 m.), but the inside of the house had been somewhatsmaller because of the thickness of the post and plaster walls. Theinside measurements, on the floor level, were found to be 11 feet(3.3528 m.) by 8 feet 6 inches (2.5908 m.).It was impossible to tell how high the walls had been, but fromthe amount of debris which filled the interior of the excavation itwould seem that they must have been about 6 feet (1.8288 m.). Theexcavation was 2 feet (60.96 cm.) and it is fairly probable that theupper walls extended 4 feet (1.2192 m.) above the surface of theground. The posts set around the periphery averaged about 6inches (15.24 cm.) in diameter and the average depth to which theyhad been set was 1 foot 6 inches (45.72 cm.). 30 Robert3, Report on Archaeological Reconnaissance in Southwestern Colorado In 1923,P. 11, fig. 1.31 Judd, Archeologlcal Observations, p. 71. 44 BUREAU OF AMERICAN ETHNOLOGY [boll. 92The sipapu was 10 inches (25.4 cm.) in diameter and 8 inches(20.32 cm.) deep. From the edge of the sipapu to the edge of thefire pit was 2 feet (60.96 cm.). The fire pit was 1 foot 6 inches(45.72 cm.) in diameter and 6 inches (15.24 cm.) deep.The doorway to the antechamber was 2 feet 4 inches (71.12 cm.)wide. The passage, if such it may be called, was just 1 foot (30.48cm.) in length. The step at the doorway was 2 1/? inches (6.35 cm.)high. The second step, that from the sill of the doorway to the floorof the antechamber, was 5 inches (12.7 cm.) high. Both steps hadbeen cut from the earth and treated with a coating of plaster.The antechamber measured 5 feet by 5 feet 6 inches (1.524 by 1.6764m.). The floor of the antechamber had an almost imperceptibledownward slope from this point and at the opposite wall was slightlylower than at the step. The wall was 1 foot 4 inches (40.64 cm.)high. The slab which closed the doorway measured 2 feet 6 inches(76.2 cm.) in length and was 1 foot 7y2 inches (49.53 cm.) high and2 inches (5.08 cm.) thick. House K Little remained of house K at the time of excavation. (Fig. 14.)This fact can be explained on the same grounds as was the similarcondition of houses B and E, namely, house K gave evidence ofhaving been dismantled after it had been abandoned. Perhaps, aswas suggested in the discussion of this feature in the two otherhouses, the material was needed in the construction of a later dwell-ing, and for that purpose many of the slabs which had lined theexcavation had been removed. There were clear indications aroundthe walls of the excavation of the slabs which had been used to faceit. Most of them were missing, however, when the remains of thedwelling were uncovered. Nevertheless, there was enough of thestructure left to show that it had been, in a general way, quite likethe others of the village.The excavation for the main rooms of the dwelling had been morenearly oval in form than in most of the structures of the village.Its interior features were much the same as those described in pre-ceding pages. The superstructure had been supported on four posts.There was one interesting feature associated with these posts. Oneof those on the north side of the room had been buttressed with aring of adobe plaster at its base. It will be recalled that a similartreatment of a north support post was observed in house H. Therewas a sipapu and a fire pit and there had been a compartment on thesouthern side of the room. There were additional holes in the floor,which must have served in the capacity of storage boxes or receptaclesfor small objects. IteBERTS] SHABIK'ESHCHEE VILLAGE 45 Although most of the slabs which had formed the compartmenthad been removed the deflector slab was still in position. The pas-sage leading to the antechamber was missing, however. The debrisremoved from this portion of the structure indicated that there hadbeen such a passage and that, as in the case of the others described, ithad been lined with stone. The latter was probably removed at thetime when the other slabs were taken from the main portion of thebuilding. o ? \ 6 XL- 10 ft. Fig. 14.?Plan of House K. a, Antechamber, b, Passage, c, Holes for support posts.d, Deflector, e, Fire pit. f, SipapuThe antechamber was of the type described for houses A and F-l.It was not as large as the former but corresponded quite closely tothat of the latter. There were indications at the west side of theantechamber of a storage cist. The latter appeared to be of morerecent date, as a portion of one of its slabs extended over those in thewalls of the antechamber, but as it was incomplete it was impossibleto tell just what its relation to the other structure may have been.Several worn-out metates or milling stones were found in the wallsof house K and its antechamber. This feature was observed in sev- 46 BUREAU OF AMERICAN ETHNOLOGY [bull. 92 eral houses, and it is one which is quite common in structures of thesame period found elsewhere in the Southwest. Metate stones werejust the size and shape of the slabs used in lining the excavatedportion of the structures, and even though their usefulness for pre-paring meal was vitiated by small holes worn through the groove ortrough portion, they were still serviceable for use in house construc-tion. Small holes or breaks in the stones would be covered by theplaster which was applied to their surfaces, and the worn and dis-carded metates would be as useful as new slabs.The excavation for house K measured 13 by 16 feet (3.9624 by4.8768 m.). When lined with slabs covered with plaster the insidemeasurements would have been several inches smaller than thosegiven for the excavation. The few slabs which remained in positionon the northern side of the room showed that the wall height at thatpoint had been 2 feet (60.96 cm.), while on the south side, near thepassage and antechamber, the height was 2 feet 6 inches (76.2 cm.).The sipapu was 1 foot (30.48 cm.) in diameter and 6 inches (15.24cm.) in depth. From the edge of the sipapu to the fire pit was 10inches (254 cm.). The fire pit was circular in form, although not aperfect circle, with an average diameter of 2 feet (60.96 cm.) and adepth of 5 inches ( 12.7 cm. ) . The distance from the edge of the fire pit to the deflector slab was2 feet 6 inches (76.2 cm.). The latter stood about 8 inches (20.32cm.) from the doorway to the passage. This last measurement is notabsolutely correct, however, as it was necessary to estimate from theindications of the wall as to just where it had been. The deflectorwas 2 feet (60.96 cm.) in height, 10 inches (25.4 cm.) wide and V/2inches (3.81 cm.) thick.The remains of the antechamber indicated that it had measured 7feet by 7 feet 6 inches (2.1336 by 2.286 m.). The passage seemed tohave been approximately 2 feet (60.96 cm.) in length. The floor ofthe antechamber was only slightly higher than that of the main room.The slabs which formed its walls were 2 feet 6 inches (76.2 cm.) inheight. House L House L was one of the best preserved of the structures in the vil-lage and might be considered as typical of the roughly circular formof dwelling. The periphery of the excavation had been lined com-pletely with stone slabs. (PI. 6, a; fig. 15.) The superstructure hadbeen supported, as in the other houses, on four posts and there wasthe usual sipapu and fire pit. The two support posts on the south-ern side of the room showed the features noted in several other ROBERTS] SHABIK'ESHCHEE VILLAGE 47 houses, namely, the one on the west was incorporated in the compart-ment wall while that on the east was just against the slabs.The fire pit was circular in form and had a raised rim of plasteraround it. This rim, as in houses E and H, was a continuation ofthe ridge of adobe into which were set the slabs forming the partitionsfor the compartment on that side of the room. Just south of the firepit was an oval depression in the floor which corresponds to those de- ,c^.X?c \ IKV ,/ :0.c { 10ft. Fiq. 15.?Plan of House L. a, Short passage, b, Step between passage androom, c, Holes for support posts, d, Deflector, e, Fire pit. f, Sipapu. g,Storage hole in floor, h, Oval depression, possibly for ladder end. i, Secondfire pit scribed for other houses and which may have been for the purpose ofpreventing the base of the ladder from slipping on the smoothly fin-ished surface of the floor.There was a rectangular box in the floor at the east side of theroom which gave indications of having been used as a fire pit. Ithad not functioned during the entire period of occupancy of thehouse, however. It had been filled with large stones, all of which,curiously enough, gave evidence of having been burned. The heat 48 BUREAU OF AMERICAN ETHNOLOGY [bull. 92had been sufficient to turn them red, and their exteriors were black-ened with soot. The small slabs which lined the pit were also burnedbrick red in color. After the loose stones had been placed in thispit ashes and other refuse were thrown in and the top was coveredwith plaster, so that there was an unbroken surface to the floor ofthe room at that point. It will be recalled that a similarly placedlire pit was found in house F-l. The same feature was also notedin one of the houses which was excavated the season before, duringthe progress of the investigations for the National GeographicSociety. At the present time the writer knows no satisfactory ex-planation for this secondary pit in the structures. There is a pos-sible explanation which will be considered when the house justmentioned is described.The deflector slab placed directly in front of the doorway openinginto the passage was somewhat smaller than those described in pre-ceding pages and could have been of little value as an actual firescreen. Its presence is interesting, however, and is a good exampleof the relationship existing between the whole complex of ante-chamber, passage, doorway, deflector, fire pit, and sipapu.The step at the doorway into the passage was well finished. Ithad a stone sill, but the riser was of adobe plaster. The latter curvedinto the room to a considerable extent, as will be noted from theground plan of the structure. The passage was not lined with slabsas in the other houses, but there was clear indication of the smallerwooden poles which had furnished the framework for its walls andalso the posts which had supported the roof. The existence of anantechamber is rather dubious. There were indications of a step atthe end of the passage, similar to that in house J, and some sugges-tions of a circular excavation, so that it may well be that there wasoriginally an antechamber, but definite evidence of such a structurewas lacking.House L measured 11 feet 6 inches (3.5052 m.) by 13 feet (3.9624m.). The slabs on the north side of the room rose to a height of 2feet (60.96 cm.), while at the south side they were 2 feet 10 inches(86.36 cm.) in height.The sipapu was 10% inches (27.4 cm.) in diameter and 5 inches(12.7 cm.) in depth. From the edge of the sipapu to the edge of thefire pit rim measured 8% inches (22.22 cm.). The rim of the fire pitwas 3% inches (9.52 cm.) thick and 2 inches (5.08 cm.) high on thenorth side. There was some variation in the width and height of thering around the periphery of the pit, but the measurements taken onthe north side may be considered as about the average. The interiorof the fire pit had a diameter of 1 foot 8% inches (52.7 cm.) on itsnorth and south axis and 2 feet (60.96 cm.) on the east and west. Roberts] SHABIK'ESHCHEE VILLAGE 49The depth of the fire pit was 6 inches (15.24 cm.). The oval de-pression in the floor just south of the fire pit was 6 inches (15.24 cm.)wide and 1 foot 2,y2 inches (36.83 cm.) Jong. It was 2% inches(6.35 cm.) deep.The secondary fire pit at the east side of the room measured 1 foot8% inches (52.7 cm.) on a north and south line and 1 foot ly2 inches(49.53 cm.) from east to west. It was 6 inches (15.24 cm.) in depth.This pit was 1 foot (30.48 cm.) from the east wall of the room.The small circular hole in the floor which was located about mid-way between the sipapu and the north wall of the room was 6inches (15.24 cm.) in diameter and 3 inches (7.62 cm.) deep.From the southern edge of the rim of the fire pit to the deflectorslab was 2 feet 8% inches (83.18 cm.). The deflector was not as largeas those found in other houses. It was just 11% inches (28.57 cm.)high. At its base it measured 1 foot 2y2 inches (36.83 cm.) across,but tapered down to 7y2 inches (19.05 cm.) at the top.The opening between the slabs forming the east and west walls ofthe compartment was 4 feet (1.2192 m.). The western end of theeast wall was 2 feet 6 inches (76.2 cm.) from the south wall of theroom. The east end of the west compartment wall was 2 feet 10inches (86.36 cm.) from the main wall. The east cross wall was 3feet (91.44 cm.) long and the western one was 3 feet 6 inches (1.0668m.) in length.The sill of the doorway to the passage was 5 inches (12.7 cm.)above the floor of the room. The doorway was 1 foot 8% inches(52.7 cm.) wide. The stone slab on the sill was the same length and1 foot (30.48 cm.) in width. Because of the curve of the plasterfacing of this step there were just 2y2 inches (6.35 cm.) between itand the deflector. The south wall of the room was 10 inches (25.4cm.) from the deflector.The passage was 3 feet (91.44 cm.) long and 2 feet (60.96 cm.)wide. The floor of the passage was on a level with the floor ofthe main portion of the structure, so that there was a step down of6 inches (15.24 cm.) from the sill of the doorway. The step atthe outer end of the passage was 1 foot (30.48 cm.) high.House MOne of the oddest structures in the entire village was house M.Its interior features were of such a nature that it might almost beconsidered as a double house. There had been two fire pits, two de-flectors, and two compartments. (Fig. 16.) The south side of theroom had been finished in much the same fashion as in other dwellings.It had the customary compartment, although of a somewhat differentnature. The fire pit had a raised rim which was a continuation 50 BUREAU OF AMERICAN ETHNOLOGY [bull. 92 of the low ridge of adobe which marked off that section of the roomand the deflector was in its proper position near the southern wall.At the north end of the room was a second fire pit and there wereclear indications of a partition which had separated this pit fromf-7/?' M? ?OC3c=)c=3esa e=ae=? \ o . . . toft Fig. 16.?Plan of House M. a, Passage, b, Storage pits in floor, c, Holes for sup-port posts, d, Deflector, e, Fire pit. f, Sipapu. gr Wall for second compartment.h, Fire pitthe remainder of the interior. The pit was inclosed, so to speak, ina compartment at that end of the room. The sipapu (there ap-peared to have been but one) was near the center of the roombetween the two fire pits.In its general construction features house M did not differ greatlyfrom the structures previously described. The excavation tendedtoward the rectangular shape, with rounded corners. It was not as bobbrts] SHABIK'ESHCHEE VILLAGE 51 deep, however, as in the majority of cases and the slabs which hadbeen used to line the periphery were much smaller. The superstruc-ture had been supported on four interior posts.Both fire pits were of the circular type. The one at the north hadbeen lined with plaster while that at the south had small stone slabscovered with plaster for its interior facing. The latter pit was muchmore irregular in form than the one at the north. This may be attrib-uted to the use of stones, which would render regularity of outlinemuch more difficult of attainment than when plaster alone was used.At the time of excavation the slabs which shut the north pit offfrom the rest of the room, and which formed the compartment atthat place, were not in position. They had fallen and were lying onthe floor. Their imprints were clearly distinguishable in the plaster,however, and the fact of their having stood in such a way as to forma low wall across that end of the room can not be questioned.The southern compartment was somewhat different from many ofthose already described. It had never been a complete inclosure inthe sense that those in the other dwellings were. No slabs had beenplaced in the plaster ridge which extended out from the walls andaround the fire pit. The southern end of the room had been marked off from the remainder but not separated from it by low walls ofupright stone slabs. Practically the same condition was found toexist in three other structures still to be described. Whether this isan example of a breakdown in essential house features, one whichwas to lead ultimately to a room without interior division, or ismerely the record of an expression of individuality can not be known.Further investigations may throw more light on the subject andfurnish some clue which will lead to the correct answer.There are certain factors which point toward the indicated ratherthan the actual form of compartment as preceding that in whichthe slabs were used. The fragments of pottery found on the floorof this structure were of the crudest type secured in the village andit is quite within the bounds of reason to suppose that the southernend of the room was first marked off by the use of plaster and thatlater the idea of actually separating it from the main portion byplacing slabs in the ridge came into vogue. There is just enoughevidence pro and con to make either theory fairly plausible as anexplanation but not enough to warrant the statement that eitherfurnishes the correct solution of the problem. The writer is ratherinclined to think that it represents a breakdown rather than thepreceding form, despite the potsherd evidence.At all events there can be no question of the fact that there wassome important significance attached to the compartmentlike fea-tures on the south side of the room, because even when an actual 52 BUREAU OF AMERICAN ETHNOLOGY [bull. 92 partition was missing it was, in many cases, suggested by a ridgeof plaster on the floor.No satisfactory explanation for the compartment on the northside is forthcoming, and in the light of our present knowledge it canbe attributed only to individual peculiarity. The fact that it wasthe only one in the village lends credence to such a supposition.Nothing definite could be determined with regard to the questionof an antechamber and passage leading into the structure from thesoutheast. There was sufficient evidence to show that there had beena passage but nothing to indicate an antechamber. It was also im-possible to determine the extent of the passage. It seemed to havebeen one of the pole, brush, and plaster type, with no stone used inits construction. The fact that it had been on the ground level alsocontributed to the difficulty with which its extent could be traced.There were simply a few scattered holes showing where the ends ofthe poles had been embedded.House M had been one of the largest structures in the village.It measured 19 feet (5.7912 m.) on its long axis and 16 feet (4.8768m.) across the short way. The wall heights on the north and southsides were 1 foot 6 inches (45.72 cm.) and 2 feet (60.96 cm.),respectively.The north fire pit was located 1 foot 6 inches (45.72 cm.) fromthe north wall and was about midway between the east and westwalls. This pit was 2 feet 2y2 inches (67.31 cm.) in diameter and 6inches (15.24 cm.) in depth. The compartment wall was just 6inches (15.24 cm.) from the southern edge of the fire pit.The sipapu was 1 foot 6 inches (45.72 cm.) from the north compart-ment. It was 8% inches (22.22 cm.) in diameter and 5 inches (12.7cm.) deep. From the edge of the sipapu to the south fire pit was4 feet 3% inches (1.3145 m.).The rim around the south fire pit averaged 6 inches (15.24 cm.)in breadth and 3 inches (7.62 cm.) in height. The diameter of thepit was 2 feet (60.96 cm.) and its depth was 10 inches (25.4 cm.).From the southern edge of the second fire pit to the deflector slabwas 4 feet 2% inches (1.2827 m.). The latter slab was 1 foot 6inches (45.72 cm.) wide and 1 foot (30.48 cm.) in height. Itsaverage thickness, like the great majority of the stone slabs usedthroughout the village, was 2 inches (5.08 cm.). The deflector stood1 foot (30.48 cm.) from the south wall of the room.The ridge of adobe which marked off the southern end of the roominto the semblance of a compartment averaged 6 inches (15.24 cm.)in width and 3 inches (7.62 cm.) in height.The average diameter of the holes for the support posts was 10inches (25.4 cm.). They had an average depth of 2 feet 10 inches(86.36 cm.). ROBEBTS] SHABIK'ESHCHEE VILLAGEHouse N 53 House N was more characteristic of the general type of structurefound in the village than was the dwelling just described. The exca-vation was of the rectangular shape with a tendency to slightlyrounded corners. (Fig. 17.) As in the case of so many of thedomiciles, the periphery of the excavation had been lined com- ^-JSft Fig. 17.?Plan of House N. a, Passage, o, Corner storage bin. c, Holes for sup-port posts, d, Deflector, e, Fire pit. /, Sipapu pletely with slabs before the coating of adobe plaster was appliedto the walls. At the time when the accumulated rubbish and debriswas removed from the interior of the pit a number of these slabswere found to be missing. Several of them had become loosenedand had fallen to the floor. Others could not be located at all buttheir imprints in the plaster were quite plain. 54 BUREAU OF AMERICAN ETHNOLOGY r boll. 92The interior finishings of the structure -were much the same asthose described for houses D, F, G, and L. The superstructure hadbeen supported on four posts get in the floor some distance from thecorners of the room. There was a sipapu, a fire pit with a raisedrim, and the southern end had been set apart from the rest of theinclosure by the low ridge of adobe extending out from the eastand west walls. There was not a complete compartment but in thesoutheast corner there was a small bin formed of upright slabs.The latter were held in position by the ridge of adobe. The deflectorstone was present in its customary position in front of the doorwaywhich had opened into the house. (PI. 6, b.)The passage in house N appeared to be of a rather simple form.It had been lined with slabs and, as in some of the other structures,its floor level had been considerably higher than that of the mainroom. In fact it was on a level with the top of the deflector slab.The outer third of the floor of the passage had a rather sharp upwardslant. This is clearly shown in the profile drawing of the ruin.(Fig. 17.) Although the floor of the passage was higher than thatof the room it still had been .somewhat below the ground level,except at its outer end where its upward slope brought it almostto the old ground level. This passage floor was an exception to thegeneral type in that it had been paved with large stone slabs. Mostof the other structures had only a plaster floor in the passageway.Little could be found of an antechamber, although there was onelarge stone slab in position at the east side of the exterior doorwayinto the passage. The latter may have been a portion of the wallsof the antechamber, or, on the other hand, it may have been placedthere purely as reinforcement to that end of the passage. Becauseof the antechamber feature in so many of the structures it is quitepossible that there originally had been some form of it present inhouse N. That such was actually the case, however, can never beknown.House N measured 12 feet 6 inches (3.81 m.) by 14 feet 6 inches(4.4196 m.). There seemed to have been some difference between thewall heights on the north and south sides of the room. The north*slabs were 1 foot 6 inches (45.72 cm.) high on the average, while thoseon the south side had been 2 feet (60.96 cm.). The tops of manyof the slabs on the south side had been broken off.The sipapu was 8% inches (22.22 cm.) in diameter and 8 inches(20.32 cm.) in depth. From the edge of the sipapu to the edge ofthe rim around the fire pit was 8% inches (22.22 cm.). The raisedrim averaged 5 inches (12.7 cm.) in width and 2y2 inches (6.35 cm.)in height. The fire pit was roughly circular in form and measured2 feet (60.96 cm.) on a north and south diameter and 2 feet 3% BUREAU OF AMERICAN ETHNOLOGY BULLETIN 92 PLATE 6 a, House L BUREAU OF AMERICAN ETHN '"'n nr;v BULLETIN 92 PLATE 7 -miBZF? ** a, Protokiva house. (Photograph by Neil M. Judd. Courtesy of National Geographic Society) 6, Doorway from passage side. Note tim- c, Doorway from room side. This picturebers for passage covering, also plaster re- shows the manner in which the originalducing size of doorway opening was later reduced iu size by useof plasterHOUSE REMAINS BOBBBTS] SHABIK'ESHCHEE VILLAGE 55inches (70.49 cm.) on its east and west axis. The floor of the pitwas 7y2 inches (19.05 cm.) below the floor level of the room. Withits raised rim the pit had a total depth of 10 inches (25.4 cm.).The deflector slab was 2 feet 6 inches (76.2 cm.) from the edge ofthe fire pit and about 6 inches (15.24 cm.) from the wall of theroom. It was not parallel to the wall, but had been set in an obliqueposition. The slab itself measured 1 foot 6 inches (45.72 cm.) inwidth and 1 foot 5 inches (43.18 cm.) in height.The small bin in the southeast corner measured, on the interior,1 foot 6 inches (45.72 cm.) by 1 foot Z% inches (40.01 cm.).The holes in which the support posts had been placed averaged 10inches (25.4 cm.) in diameter and 2 feet 6 inches (76.2 cm.) in depth.The supports at the west side of the room had been slightly differentfrom the majority of those in the various houses in that there wasa circular depression in front of each. These depressions, or prefer-ably storage holes, were also 10 inches (25.4 cm.) in diameter. Theyhad a depth of 6 inches (15.24 cm.).The doorway to the passage was 2 feet 6 inches (76.2 cm.) wide.The sill of the doorway was 1 foot 6 inches (45.72 cm.) above thefloor of the room. The passage had a total length of 5 feet (1.524m.). For a distance of 3 feet (91.44 cm.) the floor had been 1 foot(30.48 cm.) below the original ground level. The last 2 feet (60.96cm.) sloped upward to a point 3 inches (7.62 cm.) below the oldground level. House OHouse O approached the circular form more closely than anyother structure in the village. (Fig. 18.) It was also interestingfor the fact that the plaster on its walls was in an excellent stateof preservation. This condition might perhaps be attributed to itshaving been covered so completely by the wind-blown sand whichcovered the mesa top where it was located. The interior of thestructure was filled entirely with sand which seemed to have driftedinto the excavated portion of the dwelling before it had been sub-jected to much attack by the elements. The plaster was very good.It had broken through at one place only, and for that reason it wasdifficult to determine how extensively slabs had been used in liningthe periphery of the excavation. At the spot where the plaster hadbeen damaged, however, two slabs were in evidence, and it is reason-able to suppose that the entire surface of the sides of the hole hadbeen lined with them. Because of the good condition of the plasterit was not thought that the question involved was important enoughto justify the removal of portions of it to see if the slabs extended all the way around the Avail. With or without slabs the house fittedinto the series. 56 BUREAU OF AMERICAN" ETHNOLOGY [bull. 92House O was rather small in size and quite simple in its interiorfeatures. The superstructure had been supported on four posts;there was a sipapu and a fire pit. The latter was circular in formand had had a raised rim of adobe plaster. There were no indi-cations whatsoever of a compartment on the south or southeasternside of the room and there were no traces of a deflector. Evidenceof a passage or antechamber was entirely lacking. It would seemthat the inhabitants of this dwelling cared little about the custom-ary fixtures for theinterior of a houseand had been con-tented merely witha fire pit and si-papu. It is quitepossible that thestructure may havebeen an unusuallylarge storage placewhich was adaptedfor habitation pur-poses after its useas a granary.The average di-ameter of house Owas 12 feet (3.6576m.). There was nodistinction in wallheight between thenorth and southsides of the room,and the averagewas 2 feet 6 inches(76.2 cm.).The sipapu in thisdomicile was ovalrather than circularin form and measured 8% inches (22.22 cm.) by 5 inches (12.7 cm.)in diameter and had a depth of 6 inches (15.24 cm.). From the rimof the fire pit to the edge of the sipapu was exactly 1 foot (30.48 cm.).The rim around the fire pit was 6 inches (15.24 cm.) broad at itsbase and 3 inches (7.62 cm.) high. The fire pit had a diameter of2 feet 6 inches (76.2 cm.) by 2 feet 10 inches (86.36 cm.) and atotal depth of 10 inches (25.4 cm.). The floor of the pit was 7inches (17.78 cm.) below that of the room. The greater depth wasmade possible by the ridge around the rim of the pit. Pig. 18.?Plan of House O. c, Holes for support posts.Fire pit e> EOBERTS ] SHABIK'ESHCHEE VILLAGE 57The holes for the .supporting posts of the superstructure averaged10 inches (25.4 cm.) in diameter and had a depth of approximately2 feet 6 inches (76.2 cm.). House P / This structure was quite similar to the one just described in thefeatures of its smallness, the perfection of the plaster on its walls,and the general simplicity of its interior furnishings. It differedfrom the former, how-ever, in that it ap-proached more closelyto the rectangular form,although there was con-siderable curve to itswalls, and it had no in-terior support posts.The latter were placedat the corners and therewas very much the samesituation in roof con-struction as that dis-cussed in connectionwith house C. In thatstructure, it will be re-called, the large sup-port posts were incor-porated in the walls ofthe dwelling. The sameheld true for this domi- cile. (Fig. 19.)The plaster on thewalls prevented the lo-cating of more than twoslabs. The latter werein evidence near thesoutheast corner of theroom, where some of the adobe had fallen away from their surfaces.Here again it was not thought that the question as to whether ornot slabs had been used to line completely the excavation was im-portant enough to justify the damage entailed. Hence no effort wasmade to locate additional stones.House P had no sipapu, but the southeastern portion of its in-terior was set off from the remainder of the room by a ridge of adobeextending out from the walls and forming a raised rim around the9215??Bull. 92?29 5 Fig. 19, -Plan of House P. c, Holes for support e, Fire pit posts. 58 BUREAU OF AMERICAN ETHNOLOGY [bdll.92 fire pit. No slabs had been used to make a compartment partitionwall, but such a place had been suggested by the ridge. There wasno deflector slab in position, nor were there any indications thatthere ever had been one in the room. No traces could be found of apassage, and there was no suggestion of an antechamber.House P had been the smallest dwelling in the entire community.It measured 9 feet (2.7432 m.) by 9 feet 4 inches (2.8448 m.). Therewas no variation in its wall height, excepting slight irregularities inconstruction, the average of which was 2 feet 8 inches (81.28 cm.).The average thickness of the plaster on the walls was a fraction over2 inches (5.08 cm.).The rim around the fire pit averaged 5 inches (12.7 cm.) in widthat its base and tapered to 2y2 inches (6.35 cm.) at the top. Its aver-age height was 3 inches (7.62 cm.). The fire pit was 1 foot 6 inches(45.72 cm.) in diameter and had a depth of 5 inches (12.7 cm.) belowthe floor level. With the rim its depth was 8 inches (20.32 cm.).House Q This structure had a number of unique features, although it fittedinto the general type of dwelling found throughout the village.The shape of excavation was rectangular with rounded corners.(Fig. 20.) With a single exception, no slabs had been used to linethe lower walls. The plaster had been applied directly to the nativeearth walls. The single exception was at the east side of the roomwhere the entrance may have been located. Here one slab had beenset in the wall at a point directly in line with the fire pit and sipapu.Along the top of the plaster at this side of the room was a singlecourse of horizontally laid slabs, a feature somewhat similar to theuse of masonry in houses A and F-l. Another curious feature wasthat there had been three holes which might be considered as sipapusand that all of them were on the west side of the fire pit instead ofbeing to the north as previously observed.Associated with the above combination was the location of thecompartment feature at the east side of the room. It was the onlystructure in the entire group which had this orientation. In allother cases this compartment, whether actually present or merelysuggested, was at the south or southeast side of the interior of thedwelling. There is no explanation readily forthcoming as to whythere was this difference in house Q.There was no true compartment in house Q, but merely the indi-cation of one. As in the two houses just described, there was a ridgeof adobe plaster extending out from the walls to the fire pit, markingoff part of the floor space. This ridge did not completely encircle BOBEBTS] SHABIK'ESHCHEE VILLAGE 59 the fire pit as in other structures, but formed a raised rim on theeastern quarter of the circumference only. No slabs had been setin this ridge to form a partition wall for a compartment. An addi-tional feature was that the floor at the east side, within the boundariesof the adobe ridge, was slightly higher than that of the rest of theroom. As a matter of fact it may be considered that the east sideof the room had had a low platform rather than a compartment. Fi<;. 20.?Plan of House Q. a, Small holes in floor, b, Storage pits in floor, c,Holes for support posts, e, Fire pit. f, Sipapu. y, Location of small pole, h,Hole in floorFour interior posts, as in most of the dwellings, had been usedto support the superstructure. The eastern posts had, at their bases,been partially included in the adobe ridge. In this respect theconstruction might be said to run true to form.In addition to the three holes which were in such a position as tobe considered in the light of possible sipapus, there were three holesat various places in the main part of the room, which probably 60 BUREAU OF AMERICAN ETHNOLOGY [bull. 92 served as storage places. There were two holes in the floor in thecompartment or platform portion of the room. One was locatedin a rather odd place and its purpose was not apparent. It wasimmediately in front of the single slab at the east side of the room.The very small hole between this one and the fire pit had decayedfragments of wood in it, and suggested that a small pole of somesort had been placed in it. This was in about the position whichnormally would be occupied by a deflector. If, as it may wellhave been, the deflector was of pole and plaster construction, thetype noted in the discussion of house F, it naturally would besupposed that there should have been two upright posts, while therewas an indication of but one. It is possible that one was all thatthe builders deemed necessary to anchor such a deflector in position.There was no indication of a plaster wall having risen above the floorat this point. All that can be stated definitely is that there hadbeen a single pole standing in the floor at about the position wherea deflector generally would be found.Trenches failed to reveal any indications of a passage extendingto the east and there was no evidence of an antechamber. Hereagain, as mentioned in preceding pages, the external portions of thehouse may have been constructed wholly of perishable materials, alltraces of which were destroyed during the length of time which musthave elapsed since the dwelling fell into ruin and was covered overby drifting sand. It is, of course, quite possible that the structurenever had an entrance on the ground level. It may be supposedthat under such conditions the slab which was set in the east wallat a point where the doorway would have been may have had certainsignificance through being ceremonially associated with such anopening. In line with this thought is the possibility that the slabwas really a deflector which, because of the absence of a doorway,from a practical point of view was entirely useless, but which from aceremonial standpoint was felt to be essential to the well-being of theinhabitants of the dwelling. All of this is, of course, pure conjecture,and it must be borne in mind that from the position of a strictlymatter-of-fact consideration the only answer to the question as towhy certain features were as found in this structure is that we donot know.House Q measured 14 feet 4 inches (4.3688 m.) by 14 feet (4.2672m.). The wall height on the west side was 2 feet 2^/2 inches (67.31cm.) and on the east 2 feet 5 inches (73.66 cm.). The plaster portionof the east wall was the same height as that on the west side of theroom, but the row of slabs lying along its top increased the totalheight.The westernmost of the three sipapu holes was 8% inches (22.22cm.) in diameter and had a depth of 2y2 inches (6.35 cm.). The robbrts] SHABIK*ESHCHEE VILLAGE 61 second, or middle one, was slightly smaller, having a diameter of 6inches (15.24 cm.), with the same depth. The hole which occupiedthe position of the true sipapu, and no doubt served in that capacity,was 1 foot 3% inches (40 cm.) in diameter and had a depth of 6inches (15.24 cm.). The first hole was 1 foot 6 inches (45.72 cm.)from the west wall. From the edge of the first to the edge of themiddle hole was the same distance, 1 foot 6 inches (45.72 cm.), andfrom the edge of the latter to the edge of the sipapu was 1 foot 2%inches (36.83 cm.). From the edge of the sipapu proper to the edgeof the fire pit measured 8% inches (22.22 cm.).The fire pit averaged 2 feet (60.96 cm.) in diameter and had adepth of 10 inches (25.4 cm.). The raised rim on the eastern arc ofthe pit was 3% inches (9.52 cm.) in height and 5 inches (12.7 cm.)wide. The latter measurements are a good average for the adoberidge extending out from the north and south walls to the fire pit.The two additional circular holes in the main part of the roomaveraged 8% inches (22.22 cm.) and 11 inches (27.94 cm.) in diam-eter and had a depth of 6 inches (15.24 cm.). The third storagepit in the floor of this part of the dwelling was ovoid in shape. Itslongest diameter was 1 foot (30.48 cm.), while the short one meas-ured 7!/2 inches (19.05 cm.). Its depth was 5 inches (12.7 cm.).The floor of the section at the eastern end of the room, that markedoff by the ridge of adobe, was 2 inches (5.08 cm.) higher than thatof the rest of the room.The small hole in which the pole occupying the normal positionof a deflector had been set was 3 inches (7.62 cm.) in diameter andhad a depth of 6 inches (15.24 cm.). The hole at the base of the slabset in the east wall was 7Yo inches (19.05 cm.) in diameter and hada depth of 5 inches (12.7 cm.).The placing of the single slab in the east wall resulted in an ir-regular projection or pierlike feature at that point. The lattermeasured 1 foot 5 inches (43.18 cm.) across the face. At its northside it projected 6 inches (15.24 cm.) from the wall, while at thesouth it stood out but 2y2 inches (6.35 cm.). The slab was 1 foot(30.48 cm.) wide and 1 foot 6 inches (45.72 cm.) high.Structures Excavated in 1926 The ruins which were investigated during the summer of 1926 whenthe writer was conducting excavations for the National GeographicSociety's Pueblo Bonito Expedition did not occupy the same portionof the mesa top as did those of Shabik'eshchee village. The formerwere on a low knoll 215 feet (65.532 m.) northwest of the large cere-monial room. There was a distinct depression between the twomounds and extensive trenching failed to reveal anything in the na- 62 BUREAU OF AMERICAN ETHNOLOGY [bull. 92 ture of buildings in the area between the two sites. As a matter offact there. were only a few inches of earth covering the cap rock andit would have been impossible to make the necessary excavations fora dwelling, presuming conditions were essentially the same duringthe occupation of the site.The structures on the knoll to the northwest belonged unquestion-ably to the same cultural horizon and no doubt were occupied atabout the same time as those in the village. There were certaininteresting though minor variations from the general type of dwell-ing described in preceding pages, and for that reason these struc-tures should again be considered. Certain points in their construc-tion which were not easily explained when they were first excavatedbecome quite clear in the light of what has been learned from thehouses uncovered during the progress of the work- at Shabik'eshcheevillage. PROTOKIVA HOUSEWhen this structure was first described the term " protokiva " wasapplied to it because it contained so many features which have beenconsidered as characteristic of the specially constructed circular cere-monial rooms of the later-day Pueblos to which the name of kiva hasbeen applied. 32 Despite the fact that the season of 1927 showedthat practically all of the dwellings of that period had the same fea-tures, there are certain factors which make it seem that the term orname first applied to the structure is still quite appropriate.The excavation for this building was roughly circular in form andlike house C in the main village in that it had been made in such afashion that there was an encircling bench at vthe top. No slabs hadbeen used to line the periphery of the excavation, which had a thickfacing of adobe plaster. The superstructure had been supported onfour interior posts and the small poles which formed the slopingupper walls had been embedded in the earth at the back of thebench. (PI. 7, a; fig. 21.)There was a sipapu, a circular fire pit in the center of the room, adeflector slab, and a compartment. At the south there was a well-preserved entry and passageway. At the east side of the room therewas a second fire pit, rectangular in outline, similar to those describedfor houses F-l and L.One of the most interesting features in this structure was that ofthe compartment and deflector. In this case the partition wallformed from slabs extended all the way across the southern side ofthe room, except for a small opening in the center between the fire pitand doorway into the passage. It was in all respects a true compart- 33 Judd, Archeological Investigations in Chaeo Canyon [in 1926], pp. 165-166. ROBERTS] SHABIK'ESHCHEE VILLAGE 63 ment. The deflector slab was not immediately in front of the door-way into the passage, as noted in so many of the houses previouslydescribed, but stood between the opening in the low cross wall andthe fire pit. It was in a position where it could quite truthfully becalled a deflector. The row of slabs had been completely covered witha thick coat of plaster, and at the time the house was occupied it isprobable that none of the stone was visible. The south support posts jp/t. Fig. 21.?Plan of protokiva house, a, Antechamber, b, Passage, c, Supportposts, d, Deflector, e, Fire pit. f, Sipapu. g, Storage hole in floor, h,Second fire pitfor the upper walls and roof were incorporated in the partition walland at the time when the ruin was opened portions of these postswere still in position.In the eastern end of the compartment, leaning against the parti-tion wall, was a metate or milling stone, and beside it, on the floor,were the fragments of a decorated bowl. Another metate was foundon the floor east of the circular fire pit in the main portion of theroom.There was no difficulty in finding the passage or " ventilator " inthis structure. The doorway or opening into the passage was directly 64 BUREAU OF AMERICAN ETHNOLOGY [bull. 92 opposite the break in the screen of slabs which formed the compart-ment. The floor of the passage had a slightly upward slant andwas surfaced with a thick layer of plaster. The passage terminatedin an oval-shaped pit, an antechamber. Both the passage and theantechamber seemed to have been excavated at the same time as themain pit for the dwelling.The passage had been roofed over with a structure in which poles,brush, and plaster played a large part. Four notched posts had beenused. One was placed at each side of the doorway and one at eachside of the passage where it opened into the entry or antechamber.Crosspieces were then placed in the crotches to serve as supports.Longer poles running lengthwise of the passage were placed on topof these supports and formed the ceiling of the passage. The latterwere covered with bark and earth to the ground level. At the timeof excavation portions of the four supporting poles were still inposition. The cross pole at the aperture into the room was still inplace (pi. 7, b), and the remains of the six poles which had run thelong way of the passage could be traced without the slightestdifficulty.The oval pit or antechamber into which the passage opened gaveno indication of ever having been covered over, but there seems littledoubt but what it had had a superstructure similar to that whichwas erected over storage cists. Certain reconstructions which tookplace during the period of occupancy of the building may have oblit-erated traces of this covering to the antechamber. The earth wallsof the latter had been very carefully plastered, the plaster continuingaround and along the walls of the passage. The flooring wasunbroken from the antechamber to the doorway into the room.One of the instructive features of this dwelling was the manner inwhich the passage and its doorway had been reduced in size untilwhat had originally been an entrance capable of actual use becamea mere " ventilator." The changes which were made when the door-way into the room was made smaller were quite clearly shownbecause of the fact that when the reconstruction took place a reddish-brown plaster was used, while the original was gray in color, due tothe use of ashes in it. (PL 7, c.) At the time of the excavation ofthe ruin, in the summer of 1926, the reduced doorway was closed by aslab of stone which had been placed against it.The changes in the doorway were not the only ones which weremade. Presumably at the time when the opening into the passagewas made smaller, a section of wall was built on the southern arc ofthe periphery of the room. It was constructed of large slabs ofstone, laid horizontally, in the form of true masonry, held in posi-tion by thick red mortar. The line of overlap between the red Roberts] SHABIK'ESHCHEE VILLAGE 65 pj aster of the addition to the wall and the gray of the original con-struction was very plainly marked. Why such a change or addi-tion should have been desired is not at all clear. It certainly musthave caused some difficult}^ in the matter of roof rearrangement.Instead of having an even slope on all sides of the structure thesouthern portion would have a higher and less abrupt slant to it.This feature would have been even more marked than in the struc-tures in the main village, where it was observed that the wall at thecompartment side of the room was higher than at any other place.This variation, especially when it was as marked as in the protokivahouse, would necessitate some provision for bridging the gap be-tween the two sections of the superstructure. How this was donewas not indicated in the debris which filled the room. It could havebeen accomplished by a series of short poles, set upright, extendingfrom the lower to the upper portions of the roof, and diminishing inlength as the two sections of the roof came closer together near thecentral framework. The central portion of the superstructure, thatwithin the rectangle formed by the framework on top of the foursupporting posts, was probably flat.Entrance to the room in its revamped state must have beenthrough an opening in the roof. Prior to the reconstruction, en-trance to this structure could easily have been made through theantechamber and passage. When the size of the doorway was re-duced, however, it could not have been used for such a purpose. Thiswas not the only factor against the use of the passage. When thevarious additions and changes were made the builders went stillfurther and materially reduced the size of the passage and ante-chamber pit. This was accomplished through the use of large blocksof stone and great quantities of mortar forming a shaft which wasonly about 1 foot (30.48 cm.) square. The latter corresponded moreclosely to what in later structures has been termed the ventilator shaftof kivas than to an actual entrance. This rebuilt shaft was muchsmaller even than the one in house C, which was not considered largeenough to serve as an actual entrance. The work in both the shaftand passage was quite crude.The protokiva house definitely illustrated a shift from an olderto a newer style and what was once an actual entrance became merelya representative one,-a passage for the bringing of fresh air into thechamber and one which quite possibly had certain ceremonial signifi-cance. There seems to be little question but what this structurerepresents a domicile which had been remodeled with a view to morespecialized ceremonial functions.This structure exhibited another feature which was missing in allthe others excavated on top of this mesa. The excavation had notbeen carried down to the cap rock and there were 2 feet (60.96 cm.) 66 BUREAU OF AMERICAN" ETHNOLOGY [bull. 92 of clean sand between the floor and the top of the mesa. There wasno charcoal or ash in the sand, indicating that no other structurehad occupied the site and that the floor had been laid down soonafter the completion of the excavation.The two fire pits were interesting. The circular one in the centerof the room was quite deep and had been lined with stone slabswhose tops were flush with the floor. The rectangular one near theeast wall was rather shallow and had a facing of plaster only. Thelatter gave much more evidence of fire than did the former. Thiswould suggest that it had had greater use. Just why this was donecan not be explained. One interpretation might be that the struc-ture had not been given over entirely to ceremonial use and that theordinary fires of the day-to-day life of the people were kindled inthe rectangular pit while the circular one was saved for more specialoccasions. At all events there were the two pits, one showinggreater use than the other.The only other feature of interest connected with this structurewas not an architectural one. On the floor of the main portion ofthe room, lying at the west side of the circular fire pit, was theskeleton of a dog. There were no indications that it had been aburial and it is possible that the animal either fell in through theopening at the top and, being unable to escape, died there, or thatwhen the people left the site it was forgotten and left in the roomto die of starvation. The latter does not seem likely, however. Inmany instances the remains of dogs have been found indicating care-ful burial, a final mark of affection on the part of a sorrowing master.Although frequently harsh in the treatment of their animals thepresent Indians often show a decided fondness for them, and thesame may well have been true in the past. Because of this it isperhaps more fitting to think that the dog may have returned to itsformer home and in prowling about fell into the room, whence it wasunable to escape.Because of the bench there was some variation in the diameter ofthe protokiva house. For a depth of 1 foot 6 inches (45.72 cm.)from the original ground level the diameter averaged 18 feet (5.4864m.). The size of the excavation was then reduced to an averagediameter of 16 feet (4.8768 m.). The smaller circle was still furtherreduced by the application of a 6-inch (15.24 cm.) coat of plaster.This gave a final average diameter of 15 feet (4.572 m.) for thelower portion of the room.The top of the bench was 3 feet (91.44 cm.) above the floor andhad an average width of 1 foot 6 inches (45.72 cm.). The masonrywall on the south side of the room rose 2 feet (60.96 cm.) above thetop of the bench, making a total height of 5 feet (1.524 m.) forthe wall on that side. Roberts] SHABIK'ESHCHEE VILLAGE 67The sipapu was 8 inches (20.32 cm.) in diameter and had a depthof 6 inches (15.24 cm.). From the edge of the sipapu to the edgeof the circular fire pit measured 10 inches (25.4 cm.). This pit was2 feet Zy2 inches (69.85 cm.) in diameter and 1 foot 6 inches (45.72cm.) deep. The lower 1 foot (30.48 cm.) had been filled with cleansand and only the upper 6 inches (15.24 cm.) used for fire.The rectangular fire pit at the east side of the room measured 2feet 11 inches (88.9 cm.) on a north-and-south line and 1 foot 8y2inches (52.07 cm.) on an east-and-west line. It was 5 inches (12.7cm.) deep. The distance from the side of this pit to the wall of theroom averaged 1 foot (30.48 cm.).The deflector slab was 1 foot (30.48 cm.) wide and 1 foot 10 inches(55.88 cm.) in height. It stood 9V4 inches (23.49 cm.) from the edgeof the circular fire pit and was 10 inches (25.4 cm.) from the openingin the partition wall forming the compartment. The latter openingwas 1 foot 4 inches (40.64 cm.) wide.The average height of the row of slabs forming the compartmentwall was 1 foot 11 inches (58.42 cm.), but at the time "of the occu-pancy of the dwelling this had been increased 6 inches (15.24 cm.)by the plaster which covered them. The opening into the compart-ment was approximately 3 feet 6 inches (1.0668 m.) from the door-way into the passage.The original doorway into the passage had been 1 foot 11*4 inches(59.05 cm.) in height and 1 foot 4 inches (40.64 cm.) wide. Whenreduced it measured 1 foot iy? inches (33.65 cm.) in height and 11inches (27.94 cm.) wide. The slab which was in position in frontof this smaller opening at the time when the excavations were mademeasured 1 foot 2 inches (35.56 cm.) by 1 foot 5 inches (43.18 cm.).The sill of the doorway was a stone slab whose upper surface wasSy2 inches (8.89 cm.) higher than the floor of the room. The sec-ondary floor in the passage was 7 inches (17.78 cm.) above the floorlevel of the room.The passage was 5 feet 10 inches (1.778 m.) long, averaged 2 feet(60.96 cm.) in width, and had an average height of 2 feet (60.96cm.) for its original dimensions. The height and width were ma-terially diminished when this portion of the structure was remodeled.The poles used in the construction of the framework for the passagecovering averaged 4 inches (10.16 cm.) in diameter.The oval pit of the antechamber at the outer end of the passagemeasured 3 feet 8y2 inches (1.1303 m.) on its long axis and 3 feet5 inches (1.0414 m.) across the short way.The large support posts for the main superstructure were 8y2inches (21.59 cm.) in diameter. They were set at a depth of 2feet (60.96 cm.). The small poles forming the sloping upper walls 68 BUREAU OF AMERICAN ETHNOLOGY [BOLL. 92 averaged 2 inches (5.08 cm.) in diameter and had been set at anaverage of 7 inches (17.78 cm.) apart.The small storage hole at the east corner of the compartment was8V2 inches (21.59 cm.) in diameter and 8 inches (20.32 cm.) deep. , loft. Fig. 22.?Plan of House X. b, Corner bins, c, Holes for support posts, d, Deflector. e, Fire pit. /, Sipapu. a, Fire box of later date, built on site after abandonmentof house HOUSE XThis structure belonged to the same group as the protokiva housejust described but had greater similarities to the majority of thestructures in the main village. The excavation for the lower portionof the house was fairly circular in form, except for the southeastern Roberts] SHABIK'ESHCHEE VILLAGE 69quarter of the circumference where it was decidedly flattened. (PL8, b; fig. 22.) The curved portions of the wall had been faced withslabs before the application of the plaster but the straight section atthe southeast side was composed entirely of plaster. The slopingpoles for the upper walls were set back about 1 foot from the tops ofthe slabs, making more of an offset or bench than was found in mostof the structures on the mesa top.The interior furnishings had run quite true to type. There hadbeen four posts to support the superstructure. These posts wereroughly placed in the directions of the four cardinal points of thecompass. The sipapu was on the northwest side of the fire pit, whichwas crudely circular in shape. There was a deflector slab in positionnear the south wall. There had been the suggestion of a compart-ment. The latter was quite like the compartment in house F-l inthat there had been two bins, one at the east and one at the west end.The east bin was in a good state of preservation but that at the westhad completely fallen apart. Its general size and shape could bedetermined, however, from the imprints of the slabs in the adoberidge and from the stones themselves. The latter were lying on thefloor where they had fallen.The east bin had been constructed of five slabs of stone. Thelatter formed two of its walls while the other two were suppliedby the walls of the room itself. The spaces between the slabs hadbeen filled with adobe plaster, which also had covered their sur-faces. On the floor of this bin was a large milling stone or metate.The hand stone or mano was lying in the groove or trough of themetate, presumably where the woman who last had used it left it.Two more of these milling stones were found on the floor in sucha position as to indicate that they had been within the confines of(he bin at the opposite side of the room. The latter also had theirhand stones in place in the grooves.There was no indication of an entrance or passage at the southside of the room, and it is quite possible that this structure, as inthe case of several previously described, had had one of the highlyperishable type. There is also the possibility that there was noentrance on the ground level and that access to the interior waslimited to the smoke hole at the top. The wall at this side of thehouse was higher than that beyond the confines of the compart-ment. There was about the same variation as noted in some of thedwellings of the main village on the knoll to the southeast.The hexagonal cist indicated by dotted lines in the northwestportion of the ground plan of the room was of later date. (Fig. 22.)After the house had been abandoned and had fallen into decay alater group occupying the same site constructed a large fire pit. In 70 BUREAU OF AMERICAN ETHNOLOGY [bull. 92 making this pit they placed it, perhaps unknowingly, within theboundaries of the former domicile. The bottom of this fire pitwas several inches above the floor of the room. The space betweenwas filled with debris which had accumulated during the intervalelapsing between the time the house fell into ruin and the fire pitwas constructed.House X had been fairly large. It measured 17 feet 6 inches(5.334 m.) by 18 feet 3% inches (5.5816 m.). The wall at the northside was 1 foot 6 inches (45.72 cm.) high, while that at the south orcompartment side measured 2 feet (60.96 cm.).The sipapu was 7y2 inches (19.05 cm.) in diameter and 5 inches(12.7 cm.) in depth. From the edge of the sipapu to the rim of thefire pit measured 7 inches (17.78 cm.).The average diameter of the fire pit was 2 feet 6 inches (76.2 cm.)and its depth was 8 inches (20.32 cm.). The rim around the pit wasso greatly damaged that it was impossible to tell what its measure-ments had been, except that the traces on the floor suggested a widthof 5 inches (12.7 cm.). From the edge of the fire pit to the deflectorslab was 4 feet 3% inches (1.3144 m.).The deflector was 1 foot 8% inches (52.7 cm.) wide and stood 1foot 6 inches (45.72 cm.) high. It was not parallel with the wall ofthe room. At its east end it was 1 foot (30.48 cm.) from the walland at the west end 1 foot 2y2 inches (36.83 cm.).The bin at the east end of the compartment measured, along itsslab walls, 4 feet (1.2192 m.) by 5 feet 6 inches (1.6764 m.). Its eastwall, a portion of the room wall, was 1 foot 2y2 inches (36.83 cm.)long and the fourth wall, also a part of the room, measured 5 feet 10inches (1.778 m.). The average height of the slabs was the same asthat of the southeast wall of the room, 2 feet (60.96 cm.).The traces of the west bin showed that the slabs which formed thewalls inside the room made a partition 6 feet (1.8288 m.) long by5 feet (1.524 m.) in width.The interior support posts averaged about 8 inches (20.32 cm.) indiameter and were set at a depth of 2 feet (60.96 cm.) in the cap rock.The secondary fire pit measured 3 feet (91.44 cm.) by 2y2 feet(76.2 cm.), with a depth of 2 feet (60.96 cm.). Its floor was 7 inches(17.78 cm.) above the floor of the room. ARROYO HOUSEAt the foot of the eastern end of the escarpment upon whichShabik'eshchee village is located there i^ a small rincon extendingback into the mesa for some distance. In the floor of this narrowside canyon surface water has cut an arroyo or gully of considerable ROBERTS] SHABIK'ESHCHEE VILLAGE 71 depth. The washing of this gully was a fortunate circumstancefrom the viewpoint of the archeologist, because it exposed the ruinsof a small pueblo structure and, at a much lower level, the remainsof a dwelling of the type found in the main village on top of themesa. The floor of the Basket Maker III structure is 13 feet 6inches (4.1148 m.) beneath the present surface.33 (PI. 8, a.)While a great part of the remains of the structure had been washedaway there was enough of it left to enable the observer to identifyit as belonging to the period represented by the one-room domiciles ontop of the mesa. The fragments of pottery found on its floor wereof the same type as the potsherds and vessels from the mesa topand there can be no question as to its belonging to the Late BasketMaker group.The chief interest in this structure lies in the fact that its flooris 6 feet 6 inches (1.9812 m.) beneath the foundations of the smallpueblo. This shows beyond doubt that the type of structure andculture which it represents is much older than the pueblo forms.The superposition of the Pueblo cultures over Basket Maker groupshas been found in so many places that the greater antiquity of thelatter can no longer be questioned. The strata in the arroyo bankshowed quite clearly that the Basket Maker structure had gone intoruin and had been completely covered over and all traces of itobliterated long before the small pueblo structure had been erectedat the same site. 34Because of its location and the danger from overhanging banksit was possible to do little more than trace the floor level and walllocations of this structure. The remains of one of the support postswere found, however, as well as a central circular fire pit withdeflector in place at the south side. The doorway, passage, and ante-chamber, if there had been such features, had been completelywashed away. Court Near Storage BinsFrom what is known of the customs and practices of more recenthouse-building Indians of the Southwest, it may be assumed witha fair degree of certainty that the inhabitants of the village on themesa top lived the greater part of their lives in the open air. Thehouses were no doubt used as places of refuge in time of storm andM Judd, Archeological Investigations in Chaco Canyon [in 1926], pp. 166-167.31 Doctor Cummings (Kivas of the San Juan, p. 274), in referring to similar structuresin northeastern Arizona, says : " In some instances these older habitations were filled upand well-constructed rooms of stone and clay of the so-called Cliff Dwellers built abovethem.'* Mr. Morris (Exploring in the Canyon of Death, p. 272) reports a similar condi-tion existing in Mummy Cave, and the writer has seen evidences of such a superpositionin Tseahatso Cave, in the same canyon, where Mr. Morris also worked. Another undis-putable example was found by Mr. Guernsey and Doctor Kidder (Basket-Maker Caves,p. 3) in their Sunflower Cave. 72 BUREAU OF AMERICAN ETHNOLOGY [bull. 92probably were occupied quite continuously during the winter monthswhen the bleak mesa tops were swept by chill winds and occasionalsnow. When the weather was more propitious the structures prob-ably served only as places for sleeping and the storage of personal \ 12 vJ \ JV '?=?*f7\ ? U(| \+\ %\ /M\ 1 **?-* r ?^Fig. 23.?Plan of court. 12, 13, 14, and 15, Storage bins, e,Fire pitspossessions. Judging from the fire pits scattered throughout thevillage, much of the cooking was done in the open. An excellentillustration of the provisions which were sometimes made for thisphase of outdoor life was found in the remains of a court locatedat the east side of a group of storage bins. (Fig. 23.) BUREAU OF AMERICAN ETHNOLOGY BULLETIN 92 PLATE 8 a, Arrovo house. Floor of house is in right bank just below shoulderlevel of Indian boy. (Photograph by Neil M. Judd. Courtesy ofNational Geographic Society) - iV_ 6, House XHOUSE REMAINS BUREAU OF AMERICAN ETHNOLOGY BULLETIN 92 PLATE 9 a, Southeast portion of kiva 4 *tfl! >-a.- / Fig. 27.?Storage bins Nos. 5, 7, 8, and 9 All three bins just described?7, 8, and 9?seem to have been builtafter the mesa top had been occupied for some time. This was shownby the fact that the excavations went down through debris of accumu-lation. It is quite probable that these three bins served house A,which was one of the last in the village to be occupied.Bin No. 10 was one of the few in the village which approachedthe rectangular shape. (PL 1.) It was not as regular as No. 3,due to the fact that one end was slightly curved. Measurements forthis structure were 3 feet 10 inches (1.1684 m.) by 6 feet (1.8288 m.),with an average height for the slabs which lined the pit of 1 foot 6inches (45.72 cm.).Bin No. 11 also approximated the rectangular form. It was not ina very good state of preservation, however, as all of the slabs whichhad lined the pit were missing from two of its sides. (PL 1.) That 96 BUREAU OF AMERICAN ETHNOLOGY [bull. 92 they had been present was indicated by their imprints in the wallsof the excavation. At a much higher level, and actually touching thesouth side of the pit, was a fire box. It is rather interesting tonote the close association in a number of instances of these fire pitswith the storage bins. It would have been impracticable, certainly,to build a fire in close proximity to a structure composed of poles,brush, leaves, and bark had the latter not been covered with a thicklayer of earth and plaster. Because of the latter, however, firecould be used with a fair degree of safety almost against thestructure. Bin No. 11 measured 6 feet 10 inches (2.0828 m.) by 7feet 10 inches (2.3876 m.), with an average depth of 2 feet 6 inches(76.2 cm.).There was a factor of some significance to be observed in the loca-tion of the fire pits near the corners of houses or close to storagebins. The general practice seems to have been to place them a littlesouth or east of the former structures. This suggests, as previouslymentioned in the discussion of the court, that the prevailing windswere probably from a westerly direction.Judging from the slab-lined pit which remained, bin No. 12 hadbeen a typical structure for the storage of corn or other products.(Fig. 23.) It had been roughly oval in form and was not very large.In addition to its function as a bin it had served as a partial wind-break for the court, which has been described in preceding pages.Bin No. 12 was completely lined with slabs originally, but when exca-vated during the process of uncovering the remains of Shabik'esh-chee village it was found that some of the stones were missing.From the size of the pit, however, it was possible to determine itsmeasurements. The latter showed that it had been 4 feet 10 inches(1.4732 m.) by 3 feet 10 inches (1.1684 m.). The average height ofthe slabs remaining in position was 1 foot 6 inches (45.72 cm.).Bin No. 13 also formed a part of the windbreak at the west sideof the court. (Fig. 23.) It was slightly different in shape fromthe majority of such structures in that the pit which had formedthe subterranean portion was a long, narrow oval in contour. Allof the slabs which had been used to line the excavation were inposition. Measurements showed a long axis of 7 feet 6 inches(2.286 m.) and a short one of 3 feet (91.44 cm.). The averageheight of the slabs was the same as for the preceding bin, 1 foot 6inches (45.72 cm.).The third bin at the western side of the court, bin No. 14, wassomewhat similar to bin No. 13, in that it was a rather long ovalin shape. (Fig. 23.) The latter feature was not so marked in thisinstance, however. In length bin No. 14 measured 6 feet (1.8288 m.) , while its breadth was 2 feet 10 inches (86.36 cm.). Its floor was Roberts] SHABIK^ESHCHEE VILLAGE 97 plastered and the side walls were slightly higher than those of thepreceding two bins. The slabs averaged 1 foot 8 inches (50.8 cm.)in height.Joining bin No. 14 at its south side was a much larger one whichwas rather irregular in shape and had had a second one built insideof it. Bin No. 15 appeared to have been discarded and allowed tofall into ruin, after which another and smaller one was constructedin its interior. The original bin had the most irregular shape of anyin the village. This was rather hard to account for; in fact, noreally plausible explanation for its contour presents itself. Someof the slabs which had lined the pit were missing and others hadfallen from the wall to the floor. The outline of the pit was markeddistinctly, however, and there can be no doubt but that its originalshape was as indicated. (Fig. 23.)Because of the large size of bin No. 15 it was thought for a timethat it might have been an inclosure around the smaller structure inthe center. When traces of roofing poles were found in the earthback of the slabs this supposition had to be discarded. Althoughnearly as large as some of the smaller dwellings there was nothingto indicate that it had been used as such. Hence it seemed that itshould be included in the group of storage bins. It certainly wasmuch larger than the average of these structures. Its measurementswere 11 feet (3.3528 m.) by 8 feet (2.4348 m.), with an average depthof 1 foot (30.48 cm.). The latter was not as great as the depth formost of the bin pits. The smaller structure in the center tendedtoward an oval shape with diameters of 4 feet Zy2 inches (1.3081 m.)and 2 feet 10 inches (86.36 cm.). Its slabs averaged somewhat morein height than those of the outer circle. Their tops were 1 foot 3*/2inches (39.37 cm.) above the floor.The interior of the pit for bin No. 16 had been finished in plasteronly. (PI. 1.) No slabs had been used in this structure. Not onlywas this true, but its depth was considerably greater than the average.This apparently was a characteristic of the plastered pits, as all ofthose excavated showed a greater depth than those lined with slabs.Bin No. 16 was 3 feet (91.44 cm.) deep and 6 feet by 6 feet 6 inches(1.8288 by 1.9812 m.) on two diameters. The difference in diameterswas due to the fact that the northwestern arc of the periphery wasconsiderably flattened.Bin No. 17 was several feet west of No. 16 and in contrast to thelatter, inasmuch as it was unusually shallow. (PI. 1.) Slabs hadbeen used in its construction and it had a slab-and-plaster floor. Atthe time of excavation a few of the slabs which had been placedaround the wall of the pit were missing, but most of them were inposition. The lower portion of the bin had a somewhat oval shape 98 BUREAU OF AMERICAN ETHNOLOGY [bull. 92 and its diameters were 5 feet 6 inches (1.6764 m.) by 7 feet 3 inches(2.2098 m.). Its average depth was 10 inches (25.4 cm.). This wasone of the shallowest pits in the village.Between bins No. 16 and No. 17, and slightly closer to No. 17, was asmall slab-lined fire pit. The stones gave evidence of having beensubjected to considerable heat and the pit was filled with wood ashes.There were a few broken bones in the ashes. They were from a fairlylarge animal, probably an antelope or mule deer. They were nodoubt the discard from some meal which had been prepared over andenjoyed near this particular fire pit. The latter was about the sizenoted for similar boxes. It measured 2 feet (60.96 cm.) by 2 feet 2inches (66.04 cm.), with a depth of 1 foot (30.48 cm.).Bin No. 18 was quite in accordance with the usual form. It wasslightly oval in shape and had been lined completely with large stoneslabs. (PI. 1.) Not all of the latter were in position, however, atthe time when the debris was cleared from its interior. The pitmeasured 4 feet 2y2 inches (1.2827 m.) by 4 feet 10 inches (1.4732 m.)and had an average slab height of 1 foot 10 inches (55.88 cm.).Midway between the kiva and the F houses was a group or clusterof three bins, Nos. 19, 20, and 21. Only two of them were cleanedout. (PI. 1.) The third, No. 21, could be traced on the surface bythe tops of the slabs which had been placed around the periphery ofits pit.Bin No. 19 measured 6 feet (1.8288 m.) by 6 feet 6 inches (1.9812m.), with an average wall height of 2 feet (60.96 cm.).Bin No. 20 was 3 feet (91.44 cm.) west and slightly north of No. 19.It measured 5 feet 10 inches (1.778 m.) by 7 feet 3y2 inches (2.2225m.). Its average depth was greater than that of No. 19. It was2 feet 6 inches (76.2 cm.) from the tops of the slabs to the floor.From measurements taken on the slabs at the surface of the groundthe diameters of bin No. 21 were determined to be 6 feet 6 inches(1.9812 m.) by 7 feet (2.1336 m.). Due to the fact that it was notcleared of the accumulated debris which filled its interior, it wasimpossible to determine what its depth had been, but it is quite likelythat it was approximately the same as for 19 or 20.Bin No. 22 was located a few feet northwest of house G (pi. 1)and no doubt belonged to that structure. There was nothing dif-ferent or outstanding about this bin. It was of the same generaltype as the majority and of about the same size. It, too, had beenlined completely with slabs, and although all of them were not inposition, the missing ones were lying on the floor in front of thespace from which they had fallen. Bin No. 22 measured 6 feet5 inches (1.9558 m.) by 7 feet 3^ inches (2.2225 m.) on two diam- kobekts] SHABIK'ESHCHEE VILLAGE 99 eters. The average height of the slabs above the floor was 1 foot10 inches (55.88 cm.).Another one of the bins in which there had been a small bench orshelf was found a short distance east of the antechamber for houseF-l. (PI. 1.) Bin No. 23 had a fairly circular pit partially facedwith slabs. There was no question but what the entire wall surfaceoriginally had been so treated. This bin differed from those withbenches previously described in that the bench was much lower andno slabs were used in its facing. The plaster floor ran up over thetop of the bench and then continued up the wall and covered theslabs of the upper portion of the pit. The bench in the presentinstance could be more properly considered as such than those in theother bins in that it was only 10 inches (25.4 cm.) high. The slabsrose to a height of 2 feet (60.96 cm.) above the top of the bench.The bench itself had an average width of about 11 inches (27.94 cm.),although there was considerable variation in it from side to side.The diameter of the pit below the level of the bench was 5 feet 2y2inches (1.5875 m.) on its short axis and 6 feet 1 inch (1.8542 m.) onits longest axis. Above the bench the diameters ranged from 6 feet8 inches (2.032 m.) to 8 feet (2.4384 m.). The floor of the pit was2 feet 10 inches (86.36 cm.) below the original ground level.Bin No. 24 was a rather small one located about midway betweenhouses G and H. (PI. 1.) The interior of the pit had a facing ofslabs for the walls and a thick layer of plaster on the floor. Thebin had been somewhat smaller than many in the village. Thegreatest diameter of the pit was 4 feet 2y2 inches (1.2827 m.) and theshortest was 3 feet 3% inches (1.0033 m.). The average wall heightwas 2 feet (60.96 cm.).In close proximity to house G was another pit, the remains of binNo. 25. (PL 1.) Like No. 24, it had been quite small but its interiorhad been finished completely with stone slabs, both floor and walls,over which there had been a heavy coat of adobe plaster. Measure-ments showed its greatest diameter to be 4 feet 6 inches (1.3716 m.)and its smallest to be but 4 feet (1.2192 m.). Its walls were notquite as high as those for bin No. 24, as they rose but 1 foot 10 inches(55.88 cm.) above the floor.Bin No. 26 was located well down on the slope of the mesa topsome distance east and a little south of house J. (PI. 1.) It hadtended to the rectangular form, although its southeastern wallshowed a distinct curve. A curious feature connected with the re-mains of this structure was the single slab which extended outwardfrom the wall at its northern side. There did not seem to be anyreason for the stone being in this position and there was no function,apparently, which it could have served. All that can be said is that 100 BUREAU OF AMERICAN ETHNOLOGY [bull. 92 there was such a stone at that point. There was little difference inthe two main measurements of the pit. The average length of thetwo straight sides, where the slabs were still in position, was 6 feet6 inches (1.9812 m.). The tops of the slabs were 2 feet 5 inches(73.66 cm.) above the floor level. The floor in this structure wascomposed entirely of plaster.Bin No. 27 was one of the few in the village which had not hadthe stone slabs as a facing for the walls of the pit. (PI. 1.) A thicklayer of plaster had been applied directly to the native earth andthis seemed to have sufficed. The pit was slightly oval in outline,although it closely approached the circular. Its diameters on thelong and short axes were 4 feet 6 inches (1.3716 m.) and 4 feet 3%inches (1.3081 m.). Its average depth was slightly over 2 feet(60.96 cm.).A combination of the two types of pits, the completely plasteredand the one which was lined with slabs, was found in bin No. 28.(PI. 1.) Here a few slabs had been used and the remainder of thepit had been treated with plaster only. The lack of slabs in partof the interior was not due to their having been removed but tothe fact that they never had been used. No stone appeared in thefloor, which consisted solely of a layer of plaster over the earth atthe bottom of the excavation. Bin No. 28 tended more to the ovalshape than did No. 27. Its long diameter measured 6 feet 8 inches(2.032 m.), while the shorter was 6 feet 3i/2 inches (1.9177 m.).The walls of the pit rose to a height of 2 feet (60.96 cm.) abovethe floor.Bin No. 29 originally had been completely finished with stone slabs,but a number of them had fallen from the wallsi and were lying onthe floor, while others were missing entirely. (PI. 1.) The subter-ranean portion of this structure had been slightly more irregular inshape than its neighbors. In general, however, it approximated theoval-circular scheme followed in the construction of the majority ofthe bins. Its floor was composed entirely of plaster. It measured 6feet 10 inches (2.0828 m.) by 6 feet 3y2 inches (1.9177 m.) and hadan average depth of 1 foot (30.48 cm.), considerably less than theaverage.Situated about midway between the remains of houses H and Kwas the pit for bin No. 30. (PI. 1.) This was another one in whichthe walls had been part stone and part plaster, bearing in mind thefact that even where stones were used they had been covered with acoating of plaster. In this case, as in that of bin No. 28, there wasno indication that slabs ever had been placed in that section of thewall which was composed entirely of plaster. At the time whenthe structure was built there may not have been a sufficient number Roberts] SHABIK'ESHCHEE VILLAGE 101 of slabs available for use. The builders possibly did not feel in-clined to work additional ones out of the cap rock of the mesa whereit was uncovered near the edge of the cliff, so they were content tocomplete the surfacing of the walls with plaster. The floor alsowas composed entirely of the latter material. Bin No. 30 measured5 feet 6 inches (1.6764 m.) and 5 feet 3y2 inches (1.6129 m.) on twodiameters. Its walls rose 1 foot 10 inches (55.88 cm.) above thefloor.Bin No. 31 had a pit in which the walls and floor were entirely ofplaster, no slabs having been used in this portion of the structure.It was slightly larger than the preceding one but inclined to theoval in the shape of its excavation. (PI. 1.) On its long diameterit measured 6 feet (1.8288 m.) and on the short it was 5 feet 6 inches(1.6764 m.). Its average depth was 2 feet 6 inches (76.2 cm.).Bin No. 32 was 4 feet 6 inches (1.3716 m.) west of bin No. 31. Itwas more characteristic in its form in that the periphery of the exca-vation had been lined completely with stone slabs. (PI. 1.) Therewas no paving on the floor, however. The latter consisted of plasterapplied to the earth at the bottom of the pit. This bin was some-what smaller than its neighbor. Its longest diameter was 5 feet 10inches (1.778 m.) and its shortest 4 feet 6 inches (1.3716 m.). Theaverage height of the slabs was 2 feet 6 inches (76.2 cm.), which gavethe same depth to the pit as that of bin No. 31.Bin No. 33, because of the system used in numbering these struc-tures, was located at the opposite side of the village from the pre-ceding one, No. 32. It was 26 feet (7.9248 m.) east of house N andpossibly had been the storage place for that dwelling. (PL 1.) Therewas nothing exceptional or outstanding about the remains of this bin.It had been lined with slabs, several of which were missing whenthe debris was removed from its interior. Its size was within theaverage range for the village. Its longest diameter was 6 feet 8inches (2.032 m.) and its shortest was 5 feet (1.524 m.). The averageheight of the wall slabs was 2 feet 6 inches (76.2 cm.).Between houses K and L there was a cluster of bins, seven in all,which gave every indication of having been granaries for the latterdwelling. (PL 1.) They were not all contemporaneous, however, asone had been constructed in part over the remains of another.The northernmost of this group was bin No. 34, which had theshallowest pit of any found in the village. It had a slab-paved floorand its walls had been of the same material, although several of thelatter stones were absent when the remains were uncovered. (PL 1.)The floor of the pit had been but 5 inches (12.7 cm.) beneath thesurface of the ground and the slabs in the walls had risen but 7 inches(17.78 cm.) above the floor, 2 inches (5.08 cm.) above the ground 102 BUREAU OF AMERICAN" ETHNOLOGY [BULL. 92 level. The outline of the pit had approached the triangular in shape,although the sides were curved. The measurements for two axes were5 feet 2y2 inches (1.5875 m.) and 5 feet (1.524 m.).The second bin in this group, No. 36, was almost due south ofNo. 34. It was one which had had all of the slabs which formerlylined its walls removed. Their imprints could, still be seen in theearth, however, and there was no question but that they had beenused. The pit was oval in form with a long diameter of 6 feet(1.8288 m.) and a short one of 5 feet (1.524 m.). Its floor was 4 feet(1.2192 m.) below the present surface of the ground, but its wallsonly rose 2 feet 6 inches (76.2 cm.) above the floor level. (Fig. 28.)Bin No. 37 was of later construction, as was shown by the factthat a portion of it extended out and over the pit for bin No. 36.(Fig. 28.) It is possible that the slabswhich had been used to line the latterwere removed at the time No. 37 wasbuilt and that they were employed inits construction. No. 37 was quitesimilar in other respects to the rest ofthe bin pits on the mesa top. It hadmeasured 5 feet (1.524 m.) by 5 feet8 inches (1.7272 m.) and the averagewall height had been 1 foot 7 inches(48.26 cm.). One reason for the be-lief that the slabs from the lower pit,that for bin No. 36, had been used inNo. 37 was that the slabs in the wallsof the latter, where it extended overthe lower pit, extended about 10 inches (25.4 cm.) below the floor.Their total measurement was approximately what would have beenrequired for slabs placed in the wall of the lower pit.Bin No. 38 had a typical slab-lined pit, although the latter was sosmall that it was at first thought to be an outdoor fire box. When ithad been cleared of accumulated debris, however, it was found thatthere were no traces of fire in its interior and that it gave the generalimpression, one which was heightened by the indication of pole holesin the ground around its exterior, of having been the subterraneanportion of a storage bin. The floor was of plaster and most of theplaster which had covered the surfaces of the slabs was still in place.The pit measured 3 feet 6 inches (1.0668 m.) by 3 feet (91.44 cm.) andhad an average slab height of 2 feet (60.96 cm.).The pit for bin No. 39 was one of the deepest in the village. (Fig.28.) When the hole was dug it had been carried down to the caprock of the mesa and the plaster floor was applied directly to it. Fig. 28.?Storage bins Nosand 39 Roberts] SHABIK'ESHCHEE VILLAGE 103The slabs in the walls were 3 feet (91.44 cm.) in height. North-and-south diameter measured 5 feet 6 inches (1.6764 m.) and theeast-and-west 6 feet (1.8288 m.).The bin nearest to house L, in the entire group of seven, was No.40. (PI. 1.) It appeared to have been one of the older structuresand all of the slabs which had lined the walls of the excavation, witha single exception, were missing. The stone paving on the floor was still intact, however. Measurements of the pit showed diameters of4 feet 6 inches (1.3716 m.) and 4 feet 10 inches (1.4732 m.) and adepth of 2 feet 6 inches (76.2 cm.).The largest and southernmost of the bins in this cluster was No. 41.It had been lined completely with stones, both the walls and thefloor, and was in a good state of preservation when it was uncovered.(PI. 1.) The floor was 2 feet 6 inches (76.2 cm.) beneath the presentsurface, but the tops of the slabs in the wall and the original surfacewere but 2 feet (60.96 cm.) above the bottom. On an east-and-westdiameter the bin pit measured 7 feet (2.1336 m.), while the north-and-south axis was 5 feet 10 inches (1.778 m.) long.Mention has been made of the apparent lack of contemporaneityof the seven bin pits forming the cluster lying between houses Kand L. The evidence obtained during the excavations of these pitssuggested that 36, 39, 40, and 41 had been built and used at aboutthe same time, while 34, 37, and 38 were of later origin. The mostconclusive argument for such a belief was noted in the relationexisting between the ground levels at the tops of the pits. There wasa distinct accumulation of debris of occupation between the surfaceof occupation at the tops of the first-named group of pits and thesurface of occupation at the tops of the second group listed. Asecond point of interest was observed in the probability that binNo. 41 functioned in both groups. Consequently it may be assumedthat while there were seven storage bins located at that part of themesa top not more than four were used at any one time. This factoralso might be considered as having a bearing on the matter of thetwo periods of occupancy in the village as discussed in connectionwith the character of refuse found in the kiva. It is quite possiblethat bins 36, 39, and 40 were the storage places for house K, whichwas one of the dismantled structures in the village. No. 41 mighthave been the last of the first group to be constructed and was insuch a good state of preservation that it was retained in the laterseries.Because it did not belong in the group of bin pits forming thecluster discussed in preceding paragraphs, bin No. 35 was not con-sidered in its proper numerical order. As a matter of fact there isbut little to be said of the remains of this bin. It was located at the 104 BUREAU OF AMERICAN ETHNOLOGY [bull. 92 west side of the antechamber to house K and had been built againstthat structure. (Fig. 14.) There was so little left of it, however,that no description or measurements of it can be given.Bin No. 42 was close to house O and no doubt had been the granaryfor that structure. It was in a rather dismantled state, as only twoof the slabs which had been used in its construction were in placearound the wall. (PI. 1.) The floor, which was of plaster, was ina good state of preservation, however. This pit had an averagedepth of 2 feet (60.96 cm.) and on two diameters measured 4 feet2y2 inches (1.2827 m.) by 5 feet (1.524 m.).Bin No. 43 gave indications of having had a bench similar to thefew examples previously discussed. (PL 1.) It differed from theothers in that the upper portion seemed to have been somewhat higherthan in bins Nos. 1 and 8. The lower section suggested a completebin in itself and it may be that the upper row of slabs had beenadded after the other had been in use for some time. Only three ofthe slabs which had risen above the top of the bench were in positionwhen the debris which had collected in the interior of the pit wasremoved. The others had fallen into the pit and were found inthe debris. The three slabs in position indicated that the benchat the top of the lower section of the bin had had an average widthof about 6 inches (15.24 cm.). The slabs in the lower section had anaverage height of 2 feet (60.96 cm.), while those in the upper rowstood 1 foot 6 inches (45.72 cm.) above the top of the bench. Thefloor of the pit was the cap rock of the mesa. The lower sectionwas 6 feet 6 inches (1.9812 m.) on one diameter and 6 feet 8 inches(2.032 m.) on another. It was impossible to obtain the diameter ofthe upper circle of slabs, but if the width of the bench remainedthe same, or approximately so, around the periphery there wouldhave been an increase of about 1 foot (30.48 cm.) over the diametersfor the lower part of the pit.One of the largest bin pits for the entire village was that for No.44. (PI. 1.) The latter was almost large enough to have servedin the capacity of a small dwelling, but inasmuch as all of thefeatures which have been taken to be characteristic of a domicilewere absent it was thought that it belonged in the group of storagebins. No stone had been used in its construction. The walls andfloor were composed entirely of plaster. Considerable sand haddrifted over the remains of this structure since the days when itwas in use. The original ground level, which was the level of occu-pation during the period of the structure's usefulness, was 1 foot6 inches (45.72 cm.) above the floor, but the present surface is 3 feet(91.44 cm.) higher. The pit measured 8 feet (2.4384 m.) by 8feet 6 inches (2.5908 m.) on two diameters. boberts] SHABIK'ESHCHEE VILLAGE 105Bin No. 45 was locaied at the extreme southern end of the villageand at no great distance from house P. Perhaps it had been thestorage place for that structure. (PI. 1.) It had been lined withstone slabs but the floor was of plaster. In size and general contourit was quite similar to the greatest number of bin pits in the village.The average height of the slabs used to line its walls was 1 foot 6inches (45.72 cm.). It quite closely approached the circular in form,except for the flattened northeast arc, and the north-and-south andeast-and-west diameters were the same, 5 feet 10 inches (1.778 m.).Refuse Mounds There were two small refuse mounds, in addition to the depositof waste material from the dwellings which filled the interior ofthe kiva, located at different points in the village. The largest onewas between houses A and D. It partially covered the remains ofhouse B and extended down the slopes of the mesa top south of Aand east of D. Its peak was about midway between the two houses,where the total depth of the refuse was 4 feet (1.2192 m.). Thesecond and smaller one was on the slope of the mesa east of houseQ and bin No. 33. It apparently had been the place where thesweepings and other waste material from the houses at that end ofthe village were deposited. Its greatest depth was 3 feet (91.44cm.). Its peak was about midway between the house and the storagebin but some distance east of a line connecting the two. (PI. 1.)In content, both of these mounds were characteristic of the wholesouthwestern area. They were built up of sand, ashes, charcoal,broken pottery, bones, fragments of bones, worn-out metates, brokenmanos, pecking stones used in dressing out the grooves of the metates,ordinary stone spalls, and the general debris resulting from the day-to-day life of such a community.The inhabitants were rather casual in the disposal of their refusematerial, as not all of it found its way to the mounds mentionedabove. Here and there throughout the village were traces of suchmaterial, indicating that many times the housewife in her haste orindifference dumped her sweepings and other refuse at the first con-venient place. A few of the bin pits suggested that they had beenused for such purposes after their original function had been fulfilled.Wishing to obtain, if possible, further confirmation of the breakin the occupancy of the village which the debris in the interior of thekiva indicated, trenches were run through the two main refuse de-posits and their strata studied. These revealed similar conditions.In the largest mound there was a streak of clean sand, varyingfrom a little more than 1 inch (2.54 cm.) in thickness at the periphery9215??Bull. 92?29 S 106 BUREAU OF AMERICAN ETHNOLOGY [boll. 92 of the heap to about 6 inches (15.24 cm.) at the center. There wassome irregularity in its position in the mound, as one would expectwhen deposits are built up in the way that dump heaps are, but onthe whole it averaged from 1 foot (30.48 cm.) to 1 foot 6 inches(45.72 cm.) beneath the top. This would indicate that quite a littlerefuse was deposited after the sand layer had been laid down.There was also a layer of clean sand in the second mound. Itwas not as thick as that in the first but was fully as distinct. Itvaried from 1 to 3 inches (2.54 cm. to 7.62 cm.) in thickness and wasfrom 6 inches (15.24 cm.) to 1 foot (30.48 cm.) below the top of themound. With this evidence from three different parts of the villagethere can be little question of the existence of an interval duringwhich it was unoccupied. The evidence showed that there was agreater deposition of refuse prior to the break, when the sandstratum was laid down, than there was afterwards. The kiva fillshowed about an equal amount in each division but it is possible thatthe lower section was more compactly settled because of the wateraction on it. At any rate there was clear indication of a period oftime when no refuse was being deposited at any of the three majorplaces of disposal. MATERIAL CULTURE It is impossible to give a complete picture of the material cultureof a group from the information gained by the excavation of a sitesuch as that in the Chaco Canyon. At best a knowledge of theirarts and industries, if their handicraft may be dignified to the extentof such an appellation, can be gained only in part. There was a vastquantity of objects used in the day-to-day life of the people whichby their nature had a transitory existence. Baskets, sandals, articlesof dress, wooden implements, and all things fashioned out of perish-able materials would soon decay and crumble into dust. This isespecially true when a village was located in the open and its ruinswere exposed to the weather. Water from the summer rains andslowly melting snows of winter seeps through into the interior ofstructures, into graves and refuse mounds, and hastens the disin-tegration until the only antiquities which can be recovered by exca-vation are those of an imperishable quality. Because of these condi-tions all that could be recovered, upon which an appreciation of theskill and development of the inhabitants of Shabik'eshchee villagecan be based, were specimens of pottery, bone and stone implements,and ornaments fashioned from the latter materials. From informa-tion gained in the excavation of cave sites, where dry sand and theabsence of all moisture has preserved most of the perishable objects Roberts] SHABIK-ESHCHEE VILLAGE 107 of the material culture of the Late Basket Makers, it is known thatthey were unusually adept in the manufacture of basketry, sandals,and various textiles. Consequently it is apparent that in the objectsdescribed in the following; pages there is a decidedly one-sidedrecord of the attainments of the people of that period. Pottery True pottery first made its appearance in southwestern culturesduring the Late Basket Maker period. As was pointed out in thegeneral introduction to this paper, unfired clay containers prob-nbly were known in the closing days of the main Basket Makerhorizon, that of Basket Maker II, but the ceramic idea did not becomefully developed until the following stage. At what point in theBasket Maker III period fired clay vessels became a component partof the cultural complex is not known, but the industry was well estab-lished ere the incoming Pueblo peoples brought the era to a close.Pottery of this period has been found in practically all sections of theSan Juan archeological area and, despite certain minor local varia-tions, its characteristic features remain so constant that there canbe no question as to its representing a distinct ware with a definitestyle of decoration.The pottery found at the Chaco village is quite crude when com-pared with that of the Pueblo periods which followed. It never-theless was a true pottery, and has many features which wereprophetic of developments which were to reach their culmination infollowing cultural stages. Only five complete vessels were secured,and one of these is a miniature bowl, but from the many fragmentsfound a fairly comprehensive discussion of the ceramic art is pos-sible. The fragments and vessels from this site show that althoughthe industry had passed its elementary stages when the village wasfirst occupied it had not attained to the higher degree of develop-ment apparent in the forms made just prior to the time when thepeople finally abandoned the mesa top. There was definite progressin this new occupation even during the life of this single community.The ceramics at Shabik'eshchee village constituted a single distinctware, but there are certain subclasses which may be made in thegroup as a whole. The most common classification for southwesternpottery, regardless of period, is one which is based primarily onfunction, a feature which is of essential consideration in that theform and general character of a vessel may be determined in largepart by its intended use. On this basis all of the clay containers fallunder one of two headings?culinary and nonculinary vessels. The 108 BUREAU OF AMERICAN ETHNOLOGY [boll. 92 cooking pots are of three varieties, depending upon the period towhich they belong. The earliest, those of the Late Basket Makers,were smooth surfaced ; then came the banded neck forms of Pueblo I ; and finally the vessels of the full Pueblo periods, II and III, whoseentire exterior surfaces were covered with finely indented corruga-tions. The outside of these vessels is almost always black, due totheir use over an open fire, but occasionally one is found which hadnot been put to its intended function and the color is a varying shadeof gray.The general characteristics of the nonculinary group in the SanJuan area, regardless of period, may be summed up in the statementthat the surfaces of the vessels are smooth and usually ornamentedwith some form of painted decoration. This class is made up ofone major and two minor groups. The first is the well-known andwidely distributed black-on-white pottery. The two minor formsare the red wares with black decorations and the vessels with ahighly polished black interior.Too hard-and-fast rules of definition should not be applied to theblack-on-white group. The light-colored surface may vary froma dull shadei of gray to a chalky white, while the pigment of thedecoration may be any hue in the range from a brownish-red tolampblack. The latter was due, in great part, to the manner inwhich the vessel was fired. It is not attributable to a difference inthe composition of the paint. On a vessel which had been properlyfired the decoration would be a good black, while a slight amount ofoverfiring would result in a brownish-black and a marked overfiringwould give a brownish-red hue.The red vessels with black decorations had certain amplificationsin the late prehistoric Pueblo period, when the black of the designwas outlined with a white pigment, but this feature does not, ofcourse, enter into the consideration of the pottery of the Late BasketMakers. There are other distinctions in the latter form which serveto differentiate it from the black-on-red of later periods. Thesefeatures will be considered in a following paragraph.The vessels with a polished black interior show two forms. One isthe group of vessels with a gray to grayish-brown exterior while theother has a brownish-red exterior. The one seems to be an outgrowthof the other.It must be borne in mind that where the manufacture of certainobjects was a general household practice and not a specialized indus-try concentrated in the hands of a few people certain individualvariations and off-forms occasionally will be found. These can notbe fitted into any classification and for this reason must not be con- boberts] SHABIK'ESHCHEE VILLAGE 109 sidered as vitiating the common attributes of the ceramics taken as awhole."8 ( Jonsidering the ceramics from the Chaco village as a unit, regard-less of shape, function, or decoration, there are certain features whichmay be regarded as characteristic. The clay used in their manufac-ture was comparatively fine grained. It was not difficult for thepotters to secure this material because there are a number of depositsnot far from the village site. The clay was not sufficient in itself,however, as a binder or tempering material was necessary to hold ittogether and to prevent its cracking in the process of drying andlater firing. This temper apparently consisted of a fairly largeamount of white sand. It was necessary to take great care not to usetoo much of the tempering material, because under such conditionsthe cohesiveness of the clay would be lost and the vessel wouldcollapse.The large amount of sand in the Chaco vessels shows that theirmakers were still in the developmental stages of the industry andthat they had not yet learned that a smaller amount of temperingand a more careful kneading of the clay would give a harder, finer,and more compact texture to the paste. That a knowledge of thisfactor in pottery making was being attained, however, is shownby the difference between the potsherds from the lower strata of thevarious refuse mounds and those near the top. The former have acomparatively large content of sand and the paste is granular in crosssection and quite friable. The later forms show a marked reductionin the amount of temper used, an improvement in the paste texture,and a lesser tendency to flake off. It is in this same connection thatthere is a characteristic difference between culinary and nonculinaryvessels. The cooking pots, throughout the entire course of south-western pottery making, had a coarser, more granular paste than thenonculinary group, a fact which adds to the validity of the two-group functional classification.Under proper firing conditions the clay in the finished vesselbecame a light gray and even, on occasions, a rather good white incolor. When the vessels were underfired a distinct dark streak wasleft in the center of the paste. Overfiring produced a yellowish castwhich varied in intensity according to the amount of superfluousheat to which the vessel had been subjected. In fact, where the fir-ing had been greatly in excess of the amount needed, there was alight orange-red tone to the surface. The earliest group of the red 66 See Morris, The Beginnings of Pottery Making in the San Juan Area. Mr. Morrisand the writer, working independently, arrived at the same conclusions and observed thesame general features for the Late Basket Maker period. In order to simplify matters,however, Mr. Morris's classification is used because of the advantages accruing from analready published method of arrangement. 110 BUREAU OF AMERICAN ETHNOLOGY [bull. 92 pottery seems to belong in this category. The color was not due to ared slip, as in the later periods beginning with Pueblo I, but to anintentional overfiring.The surface finish on the vessels is rather rough. (PI. 11, a, b, d, e.)The loops of clay from which they were made were obliterated byrubbing the pot, while it was still moist, with some implement, suchas a corncob, piece of gourd, stick, bone, or stone, but the surfacewas not given the careful polishing which is so apparent in thewares of the later periods. An occasional piece gives evidence ofhaving been supported in a basket while being built up and perhapsfor decorative reasons the impression so made was not removed.(PL 12.)The practice of covering the surface of the rubbed-down vesselwith a thin coating of "liquid" clay, the so-called slip, had notdeveloped, and up to the present time no true specimen of LateBasket Maker pottery has been found with a slip. Vessels belong-ing to the group of painted wares occasionally give the appearanceof having been treated in this fashion, but careful examination showsthat they only have a pseudoslip. A probable explanation for thisfeature is that the paste of the vessel was quite moist when rubbeddown and given its slight polish, and as a result the finer, morefluidlike material was brought to the surface. A similar conditionmay be observed when a cement sidewalk is smoothed down duringthe process of its construction. Because of the tendency for thefiner material to work to the surface of the vessels the latter weregiven the appearance of having been treated with a very thin slip.The film of clay drawn to the surface in this manner did not com-pletely cover the coarse grains of the tempering material and conse-quently the protruding particles impart a slightly roughened qualityto it which is typical of the ceramics of the period. This featureshows quite clearly in the plates illustrating the pottery of theChaco village.Another characteristic feature of the surface treatment was theapplication of a red wash (" fugitive red " as it is known to thearcheologists of the Southwest) to- the exterior of vessels. Becauseit was not fired into the surface and made permanent it is notapparent, or else is very indistinct, on many of the fragments found.Potsherds picked up from the surface of the ground have lost itthrough the cleansing activities of snow and rain, but wheneverfragments or vessels are found in protected places this red is quitenoticeable. Many of the earlier investigators missed this featureentirely because they failed to study the vessels carefully before theywere washed and cleaned in the laboratory. Such treatment will, BUREAU OF AMERICAN ETHNOLOGY BULLETIN 92 PLATE 10 4 y^ - * *-*- '. V ** : a, Bin No. 11 6, Bin No. 5STORAGE BINS BUREAU OF AMERICAN ETHNOLOGY BULLETIN 92 PLATE 11 POTTERY FRAGMENTS BUREAU OF AMERICAN ETHNOLOGY BULLETIN 92 PLATE 12 V ?' s t FRAGMENTS FROM CULINARY VESSELS BUREAU OF AMERICAN ETHNOLOGY BULLETIN 92 PLATE 13 BOWLS a, Bowl with polished black interior b, Bowl with painted decoration ROBERTS] SHABIK'ESHCHEE VILLAGE 111 of course, remove the coloring matter. It is the impermanentcharacter of the pigment which led to its name.The " fugitive red " is observed most frequently on bowl exteriors,probably because there is a greater number of bowls and fragmentsfrom them available, but practically all forms had it. In manycases where the exteriors of cooking pots were not too heavily coatedwith soot from the fire traces of it have been found. In this respectthe customs of the Chaco village potters differed somewhat from thegeneral practice as noted by Mr. Morris. The latter reports it only /Fig. 29.?Outlines of culinary vessel shapes on the nonculinary vessels." 7 Chemical analysis of the red pigmentshows that its chief constituent was red iron ocher.There was considerable variation in shapes even at this early stagein the development of the ceramic industry. From the fragmentsfound it is clear that the potters made full-bodied jars with con-stricted necks; full-bodied vessels of an elongated spherical shapewith wide orifice ; globular or spherical pots with a slightly depressedtop and wide orifice; globular vessels with a small circular openingat the top; bowls, pitchers, ladles, and small vessels with lateralspouts.For culinary purposes the shapes most generally used seem to havebeen elongated spherical forms with slightly depressed tops and *7 Morris, Beginnings of Pottery Making in the San Juan Area, p. 176. 112 BUREAU OF AMERICAN ETHNOLOGY [bull. 92 wide orifices (fig. 29, a) ; the globular vessels with a depressed topand wide orifice (fig. 29, b) ; an egg-shaped pot with a small,slightly outcurving rim (fig. 29 c) ; vessels with a rather short, squatneck, wide orifice, and fairly globular body (fig. 29, d) ; some pitch-ers (fig. 29, e) ; and bowls (fig. 29, /). For some reason or otherthere seemed to have been a rather marked use of the latter at thisvillage. This feature has not been observed generally in the ves-sels from such sites.Absolutely accurate measurements of many of the forms can notbe given because of their fragmentary nature but a fairly closeapproximation of their size can be determined, and the followingfigures are given with the understanding that they will be consideredonly in such a light.The elongated spherical forms seem to have averaged about 10inches (25.4 cm.) in height and 8 to 9 inches (20.32 to 22.86 cm.) indiameter. The orifice averaged about 5 inches (12.7 cm.) in di-ameter. The average wall thickness was one-fourth of an inch(6 mm.).Globular vessels with a slightly depressed top and wide orificeaveraged approximately 8 inches (20.32 cm.) in height and 10 inches(25.4 cm.) in diameter. The orifice averaged slightly over 5 inches(12.7 cm.) in diameter. The average thickness of the walls of thevessels of this group was the same as in the group above.The egg-shaped pots with outcurving rims do not seem to have beenvery large. They ranged from approximately 6 to 9 inches (15.24to 22.86 cm.) in height and 5 to 8 inches (12.7 to 20.32 cm.) indiameter. Their orifices averaged 3 inches (7.62 cm.) in diameter andthe slight neck formed by the outcurved rim at the orifice was alittle less than 1 inch (2.54 cm.) high. The average wall thicknessof vessels of this group was three-sixteenths of an inch (4.5 mm.).The globular pots with short squat necks averaged about 9 inches(22.86 cm.) for body height and 8y2 inches (21.6 cm.) for bodydiameter. The average length of neck was approximately 2 inches(5.08 cm.) and the diameter of the orifice 5 inches (12.7 cm.).The walls averaged one-fourth of an inch (6 mm.) in thickness.Only two pitchers, or pitcherlike forms, giving evidence of usefor culinary purposes were found in the material from the village.They appeared to have been practically the same size. The averagebody height was 6 inches (15.24 cm.) and the body diameter wasapproximately the same. The necks averaged 2 inches (5.08 cm.) inheight and the diameter of the orifice was 4 inches (10.16 cm.). Thewalls of the vessels averaged three-sixteenths of an inch (4.5 mm.) inthickness, while the bottoms were one-fourth of an inch (6 mm.)thick. Roberts] SHABIK'ESHCHEE VILLAGE 113The bowls used for culinary purposes averaged G inches (15.24cm.) in diameter and were approximately 4 inches (10.16 cm.) indepth. Sufficient large fragments from these vessels were found tomake the figures here given quite accurate. These bowls had anaverage wall thickness of one-fourth of an inch (6 mm.). Therewas a slight tapering off of the wall near the rim which reduced thethickness at that point.Vessels comprising the second group, the nonculinary containers,were found in the following shapes, listed in the order of theirfrequency: Bowls, pitchers, large jars with constricted necks andsmall openings, globular pots with small orifices, ladles, and frag-ments from vessels with lateral spouts. None of the bird-shapedvessels so common in sites farther north was found at the Chacovillage.Bowls seem to have been chiefly of the hemispherical shape,although occasionally there were vessels which were a little too deepfor such a classification. (Fig. 30, /, g.) The rim was direct witha thin, rounded lip. In most cases it was rather uneven or undulat-ing. The walls of the vessel were tapered off for a short distancebelow the rim, so that the measurements at the latter point weresomewhat less than the average wall thickness. Bowl interiors werebetter finished than the exteriors. This was probably due to the factthat the decoration was confined in large measure to the interior ofthese vessels. In order that the maker might have a smoother sur-face upon which to paint her designs she undoubtedly gave thatportion of the vessel her greatest attention.In size the bowls ranged from those with a diameter of 4^2 inches(11.43 cm.) and a depth of 2% inches (6.04 cm.) to those with adiameter of 15 inches (38.1 cm.) and a depth of 9 inches (22.86 cm.).The main bulk of the bowls, however, ranged between 5 and 7 inches(12.7 and 17.78 cm.) in diameter and between 2y2 and 3^ inches(6.35 and 8.9 cm.) in depth. The average thickness of the bowlwalls was three-sixteenths of an inch (4.5 mm.). The bottoms weresomewhat thicker, measuring one-fourth of an inch (6 mm.), whilejust below the rim the average thickness was one-eighth of an inch(3 mm.).The material found during the progress of the excavations indi-cates that the people had two forms of pitchers. One had a globularbody with a very short neck, outcurved rim, and wide orifice. (Fig.30, c.) The other had much the same shape of body but a longerneck and more constricted opening. (Fig. 30, d.) In general ap-pearance the first was much like the culinary forms. On both typesthe handle extended from the lip to the shoulder. A rather curiousfeature of the handles was observed in the fact that those on the first 114 BUREAU OF AMERICAN ETHNOLOGY [bull. 92 type pitcher were broadly elliptical in cross section while those onthe pitchers with definite necks had a cross section which was ratherflat and quite suggestive of the so-called ribbon handles of laterperiods. One rather inexplicable quality is the general roughness g Fig. 30.?Nonculinary vessel shapes of the surfaces of these pitchers. This is even more marked than inthe general group of wares. Their outlines are fairly good, however.The average size of the pitchers of the group with the very squatnecks and wide orifices seems to have been about 7 inches (17.78 cm.)high, 7 inches (17.78 cm.) body diameter, iy2 inches (3.81 cm.) neckheight, and a 5-inch (12.7 cm.) diameter of the orifice. Roberts] SHABIK'ESHCHEE VILLAGE 115The second group averaged slightly more in height, having a totalof 7y2 inches (19.05 cm.), of which 2 inches (5.08 cm.) constituted theneck. The bod}^ diameter was 7y2 inches (19.05 cm.). The diam-eter of the orifice at the top of the neck was Sy2 inches (8.9 cm.).The average wall thickness for both groups was one-fourth of aninch (6 mm.). .Large jars, usually called water vessels, were represented by manyfragments in the potsherds collected from the various dwellings anddump heaps of the village, but complete restoration of any of themwas out of the question. A little more than half of one jar wassecured, and fortunately most of the pieces were from the same side,so that a good idea as to its form could be obtained. As far as canbe judged there was but one body shape for the vessels of thisclass. The lower portion, the section from the line of greatest diam-eter to the bottom, was between a half-oval and hemispherical form.The upper zone was generally slightly flattened and turned upwardto form the tapering neck. The neck was a part of the vessel wall inthis period but in later times it was made as a separate piece andfitted into the jar before the clay had completely dried. Enoughneck fragments were found to show that there were two generalforms. One was rather short and had a slight outcurve near therim (fig. 30, q.) while the other was long and tapered to a fairlysmall opening (fig. 30, b). There were no indications of handlesfor these vessels.As far as measurements for this group are concerned there is con-siderably less data than for preceding forms. The one jar which wassufficiently complete to warrant attempts at determining what itssize had been showed a total height of 16 inches (40.64 cm.), with abody diameter of 15 inches (38.1 cm.). The neck was 2y2 inches(6.35 cm.) high and its orifice was 3 inches (7.62 cm.) in diameter.The walls averaged five-sixteenths of an inch (8 mm.) thick.One of the forms which was quite characteristic of the Late BasketMaker period, although not numerically abundant, was the globularpot with small orifice. (Fig. 30, e.) In later periods the tops ofvessels of this class were depressed, and in this form they have beencalled "seed jars," chiefly because the first examples found had seedsin them. In many instances these vessels had two small horizontallugs with vertical perforations placed at opposite sides of the orifice.These could well have served as small handles. Thongs passedthrough the holes in the lugs would have enabled the owner to hangit up and thus give added protection to the contents. There is aninteresting feature of technique in manufacture apparent when thesevessels are compared with those of similar form made in the follow-ing periods. The earliest type, that belonging to Basket Maker III 116 BUREAU OF AMERICAN ETHNOLOGY [boll. 92 and Pueblo I, had a characteristic treatment of the lip of the orificewhich is in direct contrast to that of the later group. When the pot-ter was completing the vessel the opening was smoothed by rubbingupward and outward from the interior. This left a slightly percepti-ble ridge which in most cases does not seem to have been obliterated.Just the opposite procedure was followed in later times, when the rimwas smoothed from outside toward the interior.Vessels of the globular form with small orifice ranged from 4inches (10.16 cm.) to 8 inches (20.32 cm.) in body height and diam-eter. The commonest size seems to have been that which closely ap-proximated a body height and diameter of 6 inches (15.24 cm.). Theaverage thickness of the walls was three-sixteenths of an inch (4.5mm. ) . Representatives of two types of ladles were found in the frag-ments of vessels taken from the various houses and in the materialobtained during the work in the refuse mounds. The most abundantform seems to have been one which approached the half gourd shape.(Fig. 30, k.) This type certainly was suggested by the vegetalform with which the makers were unquestionably familiar. Thesecond type had a comparatively deep, oval bowl with a short, flat,solid handle. (Fig. 30, i.) The handle had a marked downwardslant to it with a slight upcurve near its outer extremity. It was,in fact, slightly crescent-shaped.As far as can be ascertained from rather scanty evidence theladles of the type comparable to a half gourd averaged about 6inches (15.24 cm.) in length. The handles were rather short andstubby with an average width of 2 inches (5.08 cm.) and a lengthof 2!/2 inches (6.35 cm.). The bowls of these ladles were fairlycircular in contour with a diameter of 3 to 3% inches (7.62 to 8.9 cm.) . The bowls were very shallow, ranging from one-half to three-fourths of an inch (1.27 to 1.9 cm.) in depth. In some instancesthere was an indication of a slight ridge between the handle andthe bowl. The average thickness of the vessels was three-sixteenthsof an inch (4.5 mm.).The second group had a larger average size, and while there isno definite evidence on the subject they seem to be of slightly laterdevelopment. The paste and general treatment suggest such a belief.The oval bowl of the dipper in most cases closely approximated 6inches (15.24 cm.) on the long diameter and 4 inches (10.16 cm.) onthe short. The average bowl depth was 2 inches (5.08 cm.). Theflat handle was attached to the bowl from one-half to three-fourthsof an inch (1.27 to 1.9 cm.) below the rim. The handles rangedfrom 3 to 4 inches (7.62 to 10.16 cm.) in length and from 1 to iy2inches (2.54 to 3.81 cm.) in width. They varied in thickness from Roberts] SHABIK'ESHCHEE VILLAGE 117 three-eighths to five-eighths of an inch (1 to 1.6 cm.). The wallsof the bowls averaged three-sixteenths of an inch (4.5 mm.) thick,while the bottoms had an average of one-fourth of an inch (6 mm.).The vessels with lateral spouts arc among the most interesting fea-tures in the ceramics of the Late Basket Maker period. (Fig. 31.)Portions of two vessels, including the spouts, were all that could befound at the Chaco village, but the type is well known from otherlocalities where complete specimens were obtained.08 In generalit may be said that the bodies of these vessels range from the globularto the oval shape, with slightly flattened upper zones. There is asmall circular orifice at the top of the body. The spout projectsfrom the side at about the point of greatest diameter and variesfrom a slightly downward to a slightly upward angle. None of thevessels was large. The globular portions range from 2V4 to 3%inches (5.71 to 8.9 cm.) in diameter. The sprouts range from 1*4 toiy2 inches (3.17 to 3.81 cm.) in length.The wall thickness of the two Chacofragments averaged one-eighth of aninch (3 mm.). What their use mayhave been is still a matter of question.Morris sums up the problem as follows : In the earlier description, it was suggestedthat the object in question might have been a FlG 3i._Vessel with lateral spoutlamp because the tip of the spout was burnedand disintegrated. Specimens subsequently exhumed do not confirm this belief ; moreover, it is not probable that such vessels were fat-bowled pipes because theyare neither caked nor blackened around the openings in the tops. Thus far,there is no evidence to clear up the question of function. Never in any laterperiod does the hollow spout appear in San Juan pottery. Because of its manypossible uses, it seems singular indeed that such a device, once originated, shouldnot have been retained.68The potters of Shabik'eshchee village made very little red pottery.Of all the vessels represented in the potsherds from the site therewere fragments from but three bowls and one pitcher belonging tothat group. It is impossible to tell from this material how largethe vessels had been, what decorations they had, if any, and to whatextent they differed from those of the so-called black-on-white seriesof the nonculinary wares. The potsherds do show, however, thattheir color was due to long firing and not to application of a slipcontaining coloring matter. The latter was used extensively inPueblo I and following periods.Conditions were more favorable with respect to the ware with ashiny black interior. (PI. 13, a.) The greater part of one bowl was 68 Morris, Beginnings of Pottery Making, p. 169.09 Ibid., p. 170. 118 BUREAU OF AMERICAN" ETHNOLOGY [BULL. 92found accompanying a burial, and large fragments, from a fourthto a half, from a dozen additional bowls are available for study. Thelatter indicates that as a group the vessels had a well-smoothedinterior with a fairly rough exterior. There was a glossy black finishover the entire interior surface, while the exterior was gray withlarge black blotches.Just how the potters of the period obtained the glossy black finishwill never be known, but it is highly probable that they pursuedmethods closely allied to those of the present Pueblo Indians of theEio Grande Valley in New Mexico. The latter make many blackvessels by means of a smothered fire. The smothering of the firecauses dense smoke to penetrate the paste of the vessel, leaving acarbon deposit which gives the desired black. If these vessels arereburned in an open flame this black disappears.The large splotches on the exterior of the old vessels were no doubtdue to a lack of control in the smoking process. Except at the rim,where the thinness of the wall permitted the smoke to penetratealmost completely through the paste, there is only a thin surface ofblack on the vessels.The exterior of one bowl found with one of the burials was some-what different from those described in preceding sentences. It hada decidedly brownish-gray color, showing a closer approximation tothe red vessels with a polished black interior which were to becomefairly abundant in later periods. There can be no question but that,as Morris has observed 7o and as the specimens under considerationshow, they represent the prototype of the later group of wares.The Chaco village bowls formed a fairly uniform group, as thediameters varied only from 5y2 to 8 inches (14 to 20.32 cm.). Themajority deviated only slightly from a diameter of 6 inches (15.24cm.). The depths ranged from 3 to 4y2 inches (7.62 to 11.43 cm.).The average was 3% inches (8.25 cm.). The average walls wereone-fourth of an inch (6 mm.) thick. DESIGNS ON THE PAINTED POTTERYThe decorations on vessels made during the Late Basket Makerperiod at the Chaco village were confined largely to the interior ofbowls and ladles, a practice common throughout the entire San Juanarea. Fragments from one globular pot showing exterior decora-tion and single sherds from two large jars with a trace of designon them were found, but on none of the latter was there sufficient toindicate what the ornamentation had been. The interior of the bowlsof the second type of ladle described, that of the bowl and handle 70 Morris, op. cit., p. 186. Roberts] SHABIK'ESHCHEE VILLAGE 119form, were decorated in the same fashion as the bowls. This isreadily understood, however, as they really were, in a strict sense,bowls with handles attached. In the other type of ladle both thebowl and the handle were treated with some form of design on theinterior. Occasional^ the decoration was only in the handle portion,or vice versa.The most outstanding characteristic in decoration seems to havebeen thatf of zigzag or stepped line elements. These appear -on theinteriors of bowls in combinations of 2, 3, and 4 parallel steppedlines. The figure generally bisects the bowl, although occasionaltripartite and quadrate forms are found. In a majority of thedesigns there are tips of various kinds at the corners of the steppedlines and frequently the space between the lines is filled with a seriesof dots. (PI. 14.) In a large number of the decorations there is acircle in the center of the design, at the bottom of the bowl. Thiscircle and the parallel zigzag lines are without doubt the carryingover of elements found in the decorations on baskets made duringthe preceding cultural period. Baskets of this type were found atWhite Dog Cave by Guernsey and Kidder, and by Morris in theCanyon del Muerto."Plate 15, a, 6, c, are simple forms of the stepped line, tipped cornerelements, and are quite characteristic of the earlier decorations onbowls of the Basket Maker III period. This style of design wasquite widespread in its distribution. Morris found bowls at con-temporary sites in the La Plata Valley of southwestern Coloradowith quite comparable designs on them. The chief difference be-tween the one represented in Plate 15, a, and a somewhat similardecoration on one of Morris's vessels 72 is in the absence of secondaryfigures. On the La Plata bowl there is a further element in additionto and quite independent of the figure which bisects the field fordecoration. Also, on Morris's bowl the tips at the angles of the zig-zag lines are small triangles and not the T-shaped figures shown inthis design. The second decoration (pi. 15, b) varies from its LaPlata Valley counterpart in a minor detail also. 73 The tips at thecorners of the stepped lines are further augmented by simple lines,whereas in the other vessel these embellishments take the form ofsmall triangles suggesting pennants.A typical example of the 3-line form of stepped parallel line de-sign is illustrated in Plate 15, c. Another example is illustrated inPlate 11, d. A more complicated and less common form is shown 71 Guernsey and Kidder, Basket-Maker Caves, pi. 24, a, b, f, h, i. Morris, op. cit., figs.42 and 43.n Morris, op. cit., fig. 39, b. Morris, Antiquities of Southwestern Colorado, pi. 65, e.73 Morris, Beginnings of Pottery Making, p. 191, fig. 38, b. Morris, Antiquities of South- ' western Colorado, p. 200, fig. 7. 120 BUREAU OF AMERICAN ETHNOLOGY [bull. 92 in the next design. (PI. 15, d.) Here the tips at the angles of thezigzags were extended in such a way as to interlock with those ofthe other lines, thus forming what at first appearance seems to bea rather involved decoration but one which is in reality fairly simplein its basic elements and one which runs quite true to form.Characteristic examples of the stepped-line figures with dots areshown in Plate 15, e, /. The former is a common decoration, al-though there are slight differences in each design found. As anexample of the tripartite method of ornamentation / is decidedlytypical. A slightly different form of this style of decoration is alsoillustrated by one of the vessels which Morris has pictured. 74A different use of dots in the make-up of a design is to be notedin Plate 16, cl The painter of this decoration used two straight-linepanels meeting the circle at the center of the bowl. The dots wereplaced inside of these framing lines, as in the previous examples,but instead of the rather hit-or-miss fashion they were so arrangedas to form a zigzag line. The ticks jutting out from the framinglines of the panels are quite often observed in the designs of thisperiod. Still another use of dots is depicted by the next design.(PI. 16, ft.) Here they were used to outline figures composed ofstepped lines embellished by triangular elements. This seems tohave been a favorite form of decoration at the Chaco village, aspotsherds from many different bowls were found bearing portionsof such designs. (PL 14.) It was fortunate that enough frag-ments were recovered from this particular bowl to make possible thecomplete restoration of the design.The quadrate form of decoration is illustrated in Plate 16, c.Fragments from several different bowls were found bearing this typeof design. There seems to have been considerable variation in thenumber of small figures placed inside the framing lines. Morrisfound one example with a single row of them,75 while fragments fromthe Chaco village show two and four rows as well as the triple-rowexample pictured in the drawing. Potsherds from two differentbowls show that these rows of small figures were occasionally usedwithout accompanying framing or border lines.A decoration which, up to the present time, stands as a uniqueexample in the designs of this period is depicted in Plate 16, d. Itwas necessary partially to restore this design, but there can be noquestion as to its correctness, as fragments representing a considerableportion of the bowl were recovered. The pairs of zigzag lines em-bellished with triangular figures are quite suggestive of the somewhatsimilar element in the decoration above. (PL 16, ft.) The other fig- 74 Morris, Beginnings of Pottery Making, p. 175, fig. 25, d.76 Ibid., fig. 25, a. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 92 PLATE 14 .; :>,.v> -Mm if / POTSHERDS WITH PAINTED DESIGNS, BLACK-ON-WHITE SERIES BUREAU OF AMERICAN ETHNOLOGY BULLETIN 92 PLATE 15 /DESIGNS FROM BOWL INTERIORS BUREAU OF AMERICAN ETHNOLOGY BULLETIN 92 PLATE 16 /DESIGNS FROM BOWL INTERIORS BUREAU OF AMERICAN ETHNOLOGY BULLETIN 92 PLATE 17 BOWLS a, Undecorated. 6, With painted design. noBERTs] SHABIK'ESHCHEE VILLAGE 121 ures, on opposite sides of the field of decoration, are quite interestingin form and may possibly have been painted there to represent centi-pedes. It is not unusual to find this life form represented on waresof later periods.The last two designs figured, Plate 16, e, f (see also pi. 13, b), aregood examples of the triple-panel style of decoration in which thepanels, pendent from the rim of the vessel, tend to cut across thefield of decoration in an oblique direction. In outline the panels tendsomewhat to the curvilinear form. Both of these designs illustrateadditional combinations of the zigzag lines, dots, and triangularelements. In / there is a suggestion of the so-called lightning sym-bol. As a matter of fact it may be considered in the light of a vari-ation of the zigzag-line element. But, in this case it takes a negativeform, owing its presence not to the actual lines but to the bordersformed by triangular figures. Neither of the latter designs repre-sents a form common in the decorations of this period, but the stylebecame quite popular, in slightly modified forms, in the periodsimmediately following.Only one example of a true band decoration was found on potteryfrom this site. This design consists of a series of triangular figurespendent from the rim of the bowl. (PI. 17, &.) The latter wasfound in the protokiva house excavated during the work of the pre-ceding summer. Band designs are rare in the earlier stages of theLate Basket Maker ceramic phase but occur occasionally on speci-mens made toward the close of the period.Although the decorations on the vessels of the Late Basket Makersare predominantly of the geometrical type, occasional realistic de-signs are to be observed. Except for the centipedelike figures pre-viously mentioned, the specimens from Shabik'eshchee village donot contain a single complete example of the realistic art. There aretantalizing fragments on which may be seen the feet of a bird, por-tions of some animal, and even suggestions of human beings, butin not a single instance is there enough to make restoration possible.Half of one bowl bearing two humanlike figures was found a numberof years ago in a rincon across the canyon from Pueblo Bonito byMr. Morris. The remains of this vessel show that it belongs un-questionably to the period under discussion and in order that someidea may be had of the nature of the life forms in decoration adrawing of it is included in this report. (Fig. 32. ) 70 The square-shouldered, triangular-bodied figures are very characteristic of theBasket Makers. At many sites quite similar forms have been noted 78 University of Colorado Museum, Boulder, Specimen No. 744.9215??Bull. 92?29 9 122 BUREAU OF AMERICAN ETHNOLOGY [BULL. 92 painted on the walls of caves and on near-by cliffs. 77 This bowl alsoillustrates another variety of the quadrate form of decoration. Thepanels which marked off the field for decoration into four segmentswere composed of elements which are quite characteristic of bothtextile and ceramic designs. This element is observed frequentlyand many of the fragments of pottery from the Chaco villageshowed portions of designs in which it had been used. The char-acteristic circle drawn in the bottom of the bowl is very apparent inthis specimen. It is quite likely that the other half of the vesselcontained similar figures. That the missing portion had been di-vided by a panel is shown in the traces of the latter which are presentnear the central circle. Fig. 32.?Painted design from half of a bowlThe use of life figures in the decoration of bowls is not at allsurprising because such representations frequently are found onbaskets of the preceding period. 78 It is quite apparent that fromthe beginning of the manufacture of painted vessels both geometricand life forms were used in decoration and that the two formsexisted side by side. That one did not grow out of the other orresult from it in the early development of southwestern designsis certain. Another interesting feature is that practically all of thelife figures on the pottery of the early prehistoric periods arealmost identical with the drawings on the cliffs throughout the area.In very late times, it is true, life forms became highly convention-alized, and there are the many so-called bird symbols on the Rio 77 Kidder and Guernsey, Archeological Explorations, pp. 197-198, figs. 100, 101, pi. 96,a, b. The writer has observed such figures in several places along Montezuma Creek insoutheastern Utah. The latter were pecked into the rock, however, and not painted.They were close to Basket Maker III sites in each instance.78 Pepper, Ancient Basket Makers of Southeastern Utah, [)p. 13, 15. Guernsey andKidder, Basket-Maker Caves, Plate) 26, c. Roberts! SHABIK'ESHCHEE VILLAGE 123Grande wares, the butterflies and moths, and birds of the Hopi, andother involved symbols, but they are all long subsequent to the formsat present being considered.There is one feature in the decorated wares of the period which israther marked, and that is the painted rims of the bowls. In agreat majority of cases the lips of the bowls were so treated, produc-ing what may be considered as a frame for the entire design. Unfor-tunately the rims have been so rubbed and chipped that it is notpossible to tell whether or not the break, which is so apparent onvessels of later date, was present. The line break consists of asmall unpainted space on the rim. It is explained on the basis ofthe belief among some of the modern Pueblo Indians that if the linewas made a complete circle the spirit of the vessel could not pass inor out.In the foregoing discussion no attempt was made to read myste-rious and profound meanings into the so-called symbolism of thepottery decorations. Inasmuch as such a treatment of prehistoricdecorations would be largely guesswork, the writer has neither theinclination nor the qualifications necessary for such a study. Mysticnumbers apparently played but a small part in the decorative combi-nations. Series of elements from 1 to 10 may be found on thesame piece of pottery. It is true that certain numbers or combina-tions of numbers are found more frequently than others, but as totheir having symbolical significance there is a question. One mightfind an infinite variety of reasons to explain why 1, 2, 3, and 4 figureswere used in a decoration and none of them would really tell whatthe maker had in mind. It was probably discovered that certaincombinations fitted into a certain design better than any other num-ber and they were used for that reason and not because they were feltto symbolize some esoteric factor in the religious beliefs of the people.Such explanations, like the reasons for certain features in the kivas,probably were thought of long after the symbol or design had comeinto use. It is the opinion of the writer that there has been too greata tendency on the part of the dilettante and amateur in archeologyto read deep and mystic meanings into things which should not be soconsidered and the field of pottery designs has been one in whichtheir imaginations have run riot.As an example of what the Indian thinks of the elements in adesign and their symbolic nature, the findings of one investigatoramong the Pueblos of the Rio Grande are interesting. 79 Whenquestioned as to the meanings of certain figures the replies wereoften evasive and the Indians were frankly puzzled as to what theinterrogator wanted. When meanings were obtained for a symbol w Guthe, Tueblo Pottery Making, pp. 85-88. 124 BUREAU OF AMERICAN ETHNOLOGY [boll. 92from more than one woman they were generally different. Themeanings obtained are usually the result of insistent questioning.The potters had never definitely given an interpretation to theirdesigns until asked many times to do so. Then they strove to pleasethe investigator with some sort of an explanation. There are somedefinite representations but no questions need be asked concerningthem. The clouds, the lightning, life forms, all speak for them-selves. Their meaning is so plain that there can be no doubt withregard to them. It is in the geometrical figures where attempts aremost frequently made to read what in the majority of cases is notpresent.The characteristics of the pottery as a whole may be summarizedbriefly as follows: The vessels are gray, brownish-white, white, ororange-red in color. They have slightly pebbled surfaces, due toprojecting particles of the tempering material used in the paste. Thedecorated vessels have a thin pseudoslip which resulted from addi-tional rubbing of the unfinished vessel. There was a marked use ofdots, tipped, stepped lines and solid triangular figures in the designs,together with a small circle at the bottom of the bowl. The majorityof the forms represented at the Chaco village belong in whatMr. Morris has termed the " Standard complex," which is character-istic of the normal phase of the ceramics of the Late Basket Makers.80Additional Clay Objects Pottery containers were not the only fired-clay objects uncoveredduring the excavations. There was a second group consisting ofclay pipes or cloud blowers. (PI. 18.) The latter name has beengiven to the objects of this type because the modern Pueblo Indiansuse similar ones for ceremonial purposes. During the progress ofcertain religious observances small puffs of smoke, supposedly repre-senting clouds, are blown to the cardinal directions by priests usingpipes of this form. It can not be ascertained, of course, whether theancients used them for the same purpose or not, but it is quite logicalto suppose that they did. They certainly would be an inconvenientform of pipe to smoke for pleasure. They tend to the tubular orelongated funnel shape. A short hollow reed was placed in thesmall end to serve as a stem. The bowl portion, if such it may becalled, in the present specimens is very shallow ; in fact, it constitutesonly about one-fourth of the total length. The remainder of theobject is solid, except for the small hole running through it from thebottom of the bowl to the small end. Two complete specimens and athird which is practically complete give a fair idea as to the generalsize of these objects. ?0 Morris, Beginnings of Pottery Making, p. 161. KOBERTS] SHABIK'ESHCHEE VILLAGE 125The cloud blower figured in Plate 18, &, measures 2^4 inches (5.7cm.) in length. Its diameter at the bowl is 1 inch (2.54 cm.), while thedepth of the latter is one-half inch (1.27 cm.). The second one, o,is somewhat smaller than the other two. Its total length is 1% inches(4.44 cm.). The diameter at the bowl is seven-eighths of an inch(2.22 cm.). The latter's depth is one-half inch (1.27 cm.), whichis somewhat greater in proportion to the length than in the other twospecimens. The third one, d, is 2 inches (5.08 cm.) in length. Thebowl end has a cliameter of 1T3? inches (3 cm.) and a depth of three-eighths of an inch (1 cm.).A unique form of cloud blower is pictured in Plate 18, a. In thisinstance there was a single bowl with two stems. The writer hasnever seen or heard of such a form coming from the Southwest.There is no question but what it had been used as a pipe, because theinterior of the bowl is burned. In fact, when it was dug out of thesoutheastern corner of the compartment in house D there was stilla thin crust of residue, the cake, lining its interior. Both stems areperforated. The stems are so nearly the same length that thereis no appreciable difference between them. The total length of theobject is 1% inches (4.76 cm.). The bowl is elliptical in contour.The long diameter is seven-eighths of an inch (2.22 cm.) and theshort is five-eighths of an inch (1.58 cm.). The bowl depth is one-half of an inch (1.27 cm.).A portion of another cloud blower, about one-half of the stemend, is interesting because it was decorated with a zigzag line formedby a series of dots. The latter were not painted on but had beenpunched into the surface before the object was fired, when the claywas still moist. (PI. 18, e.)There was another class of objects which bore the same type ofdecoration. (PL 18, /.) They were quite like the cloud blowers intheir general form but had no lengthwise perforation through thestem. The larger end was cupped, resembling somewhat the bowlportion of the other objects. This " bowl " showed no traces of fire,however, and the objects certainly could never have been used forthe same purpose as the cloud blowers. Mr. Morris reports similarobjects, but made from unfired clay, which he found in the Canyondel Muerto in northeastern Arizona, and one unfired specimen fromthe Mimbres Valley in southern New Mexico. He sums up the ques-tion of these objects as follows : No definite function can be assigned to these nipple-shaped objects. It is plainenough, however, that they had no utilitarian value. Notwithstanding that theyare purely speculative, a few suggestions may be worth presenting. The con-clusion seems unavoidable that the female figurines described in a previous sec-tion were cult objects. The nose and breasts are in high relief, being, in fact,the most conspicuous portions of the figures. Presumably wheu cult personages 126 BUREAU OF AMERICAN ETHNOLOGY [bull. 92 appeared in ceremonial regalia or were represented in altar fittings, these sameanatomical features would have been stressed. The four specimens of the pres-ent series, which were unquestionably completed, seem to represent the type,and are perforated in a manner that would permit them to be attached to avertical surface so that they would rest with their longitudinal axes in ahorizontal plane. Thus, lashed to masks, they would have served very well todepict noses, or sewn to clothing, mammary glands. Again, when held tip down-ward, they are suggestive in shape of the large carrying baskets in use at thetime of their manufacture. To represent these tbey may have been placed onthe backs of manikins, perhaps attached directly to the substances of which thelatter were made, by means of the perforations which are in a comparableposition to the loops on the baskets themselves, or fastened to the ends ofminiature tump lines passing across the foreheads of the figures."The writer is inclined to believe that the last suggestion of Morrishas a closer bearing on the problem than the preceding ones. In thenext period, Pueblo I, and those following, objects which are verysimilar in shape and outline were made. The latter show quite dis-tinctly that they were patterned after the baskets referred to in theabove quotation. In view of this fact it may well be that the cruderobjects here considered were the prototype for the later forms. Thismust be considered, however, as Mr. Morris points out, in the lightof pure speculation. Our present evidence is not sufficient for thedrawing of any definite conclusions.BoneworkBone implements from an archeological excavation are interestingnot only for a study of the various types of tools made from this mate- rial and their probable function but also for the record which theyfurnish of the fauna present in the region at the time the site wasoccupied. The implements from the Chaco village were not ex-tremely numerous, but even in the small series available, for study anumber of different animals and birds are represented. It is certainthat the mule deer {Odocoileus henuonws)^ the prong horn antelope(Antilocarpa amiericaruv) , the wildcat {Lynx i*ufus), and jackrabbit{Lejyus californiem texiamos) were in the neighborhood, becausethere are implements made from bones of these animals in the col-lection. The process of tool making is such, however, that in manycases the distinguishing features of the bone are removed and it isimpossible to identify the animal or bird represented. This is espe-cially true of the latter bones from the Chaco site. It is possible toidentify them as such, but further than that no definite statementscan be made. It is quite possible that they are from turkeys, as thebones are large enough to have come from such a fowl, and it isknown that there were wild turkeys in the region. The elk (C&rvus 81 Morris, Beginnings of Pottery Making, p. 158. Roberts] SHABIK'ESHCHEE VILLAGE 127 canadensis Occidentalls) is represented by one flaking tool which wasmade from one of the prongs of an antler.82It is readily apparent that bone furnishes an ideal material fromwhich to make certain kinds of implements. The ease with whichit may be fashioned and the slight amount of labor involved inshaping many tools from it are attributes of no little value. Addedto this is the factor of a fairly constant supply of the raw materialresulting as a by-product from the ever-important industry of thesatisfaction of hunger. Many times it was necessary only to sharpen(lie end of a broken bone to make a serviceable punch or awl. Thecondyle or articular end was left intact and from its nature furnishedan excellent handle for the implement. At other times the splitbones and fortuitous splinters remaining after a feast were takenand given a sharp point. No other work was necessary to make themusable. In other instances the maker was more particular and care-fully smoothed and polished the entire implement.Bone splinters seem to have been the commonest source of mate-rial as far as the boneworkers of the Chaco village were concerned.A majority of their implements were made from such fragments.However, complete bones and specially cut pieces were by no meansneglected. A general classification which might be made wouldgroup the implements into two divisions, one consisting of the simplepointed splinters, the other comprising tools which had been care-fully worked and polished in addition to being shaped at the ends.AWLS Typical specimens of awls which belong in the sharpened-splintergroup are illustrated in the upper row of Plate 19. The majority ofthem are fragments of long bones. Some indicate a partial polishwhich is the result of use but they do not show the definite and in-tentional smoothing which is apparent on those of the middle andlowest rows of the same photograph. Some of the latter were madefrom rib fragments as well as from fortuitous splinters from longbones.The long bones from the deer, antelope, and smaller mammalsseem to have been favored especially for the making of awls. (PI. 20.)The common practice, apparently, was to split the longer bones,sharpen one end to a good point, leaving the condyles at the other,and then to polish the entire implement. Constant resharpeningof the point would, in the course of time, so wear down the bone thatpractically nothing but the articular surface which formed the handle 82 The writer wishes to express his appreciation for the assistance given him in theidentification of these bones by Dr. A. Wetmore and Mr. H. H. Shamel, of the U. S.National Museum. 128 BUREAU OF AMERICAN ETHNOLOGY [bull. 92 would be left. Several examples of this are to be seen in Plate 20,&, c, d, h. The cannon bone of the mule deer was especially adaptedfor this type of implement because there is a natural groove downthe center of the anterior surface which, with the division of the ar-ticular surfaces at the end, gave a natural line of cleavage, so to speak,along which the cut was made. Four of the implements picturedshow this very clearly. All but two of the bones in this group arefrom the mule deer and antelope. The implement lettered g was madefrom the tibia of a jack rabbit while / is from the ulna of a wildcat.The latter bone was only slightly modified in the process of makingthe desired implement.Other examples of awls which were made with but little modifi-cation of the original bone are illustrated in Plate 21. As a matterof fact all that was done was to rub one end down to a sharp pointand give the bone some additional polishing. Even this was notalways done. The commonest sources of material for tools of thistype appear to have been the radii, tibiae, and ulnae of jack rabbits(a, b, c, d, e) and the fibulae and radii of wildcats (/, g).SPATULATE AWLS There is still another class of bone tools which have the charac-teristic sharp points of the awls but another feature in addition.These implements are the so-called spatulate awls. (PI. 22, upperrow.) One end has the customary sharp point of the awl while theother is squared or rounded and has been smoothed down until theedge is quite thin. Frequently an additional feature of this edge isits beveling. The increased usefulness of such an awl over thesimple type is quite apparent. With such a tool it would be possibleto employ one end for smoothing, rubbing, or scraping and theother for making perforations. The spatulate awls of the upperrow, Plate 22, are fairly simple in form. The one at the left endof the lower row, e, suggests a little better workmanship in its manu-facture. The point is not as sharp as in the other specimens but thespatulate feature was better before the object was broken. Thebone is fairly thin and had a very good edge for smoothing andrubbing purposes. It also might have served as a shuttle in weaving.The second implement from the left in the lower row, Plate 22, /,is of the spatulate form at one end but was not pointed at the other.The latter was slightly rounded instead. This implement moreclosely approaches what is generally considered the true spatulatetype. It was the only specimen of its kind found in the village.While it is not common in the collections from any of the south-western ruins this form of implement is found more frequently atsites belonging to later periods. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 92 PLATE 18 CLOUD BLOWERS AND ADDITIONAL CLAY OBJECT BUREAU OF AMERICAN ETHNOLOGY BULLETIN 92 PLATE 19 BONE IMPLEMENTS, AWLS a measures 4J6 inches (1238 cm.) in length BUREAU OF AMERICAN ETHNOLOGY BULLETIN 92 PLATE 20 BONE IMPLEMENTS, AWLS a measures 7x?ia inches (19.72 cm.) in length BUREAU OF AMERICAN ETHNOLOGY BULLETIN 92 PLATE 21 BONE IMPLEMENTS, AWLS g measures b}\(, inches (13.9 cm.) in length BUREAU OF AMERICAN ETHNOLOGY BULLETIN 92 PLATE 22 BONE IMPLEMENTS, SPATULATE AWLSg measures 5?? inches (13.7 cm.) in length BUREAU OF AMERICAN ETHNOLOGY BULLETIN 92 PLATE 23 BONE IMPLEMENTS, PUNCHES a measures 5}i inches (13.97 cm.) in length BUREAU OF AMERICAN ETHNOLOGY BULLETIN 92 PLATE 24 S^ I CZ BONE IMPLEMENTS a, Spoon spatula, 10 inches (25.4 cm.) in length, b, Scraper or dressing tool, e, Flaker BUREAU OF AMERICAN ETHNOLOGY BULLETIN 92 PLATE 25 BONE IMPLEMENTS, NEEDLES, BODKINS, WHISTLES, AND BEADSLong whistle measures 4 inches (10.16 cm.) in length hoberts] SHABIK'ESHCHEE VILLAGE 129The third implement in the lower row, Plate 22, g, is what mightbe termed a knifelike spatula. It was made from a large splinterfrom the shaft of a long bono of a large mammal, probably adeer. One side and end was worked down to a slight curve and afairly thin edge. Such a tool, in spite of its lack of finish in allother respects, would be very serviceable in the skinning of animals.The last implement in the lower row of this same plate, h, is dif-ficult to elassif}'. A fragment from a long bone was taken, its endsrounded off and one edge beveled to make a fairly sharp blade. Theopposite edge was rubbed down to a comparatively blunt surface andthe sides were well polished. The most appropriate name for suchan implement, if it may be permissible to borrow from the nomencla-ture of stone tools, is side scraper. A small bone side scraper ofthis nature would find many uses.PUNCHES Differing from the awls only in the degree of the sharpness ofthe pointed ends is the group of so-called punches. The latter, likethe awls, were made from fortuitous splinters, from especially cutlong bones, and occasionally from antler prongs. They were pol-ished or not according to the whim of the maker. (PI. 23.) Aswill be noted from the illustration, the points on these implementsare considerably blunter and more rounded than those of the awls.The implement made from an antler tip (pi. 23, c) is of addi-tional interest because of the fact that it had been hollowed out atthe base end and the edges smoothed down. This suggests that itmight possibly have been a handle for a knife rather than a punchas its tip end would indicate. The only objection to the former isthat it is almost too small to have held a knife blade of any greatusefulness. It is one of those forms which do not readily fall intoany classification. FLAKING TOOLS There is another form of tool which frequently is classed withthe punches but which, strictly speaking, is a different implement.These are the flakers. They were made generally from antlers andwere finished with a fairly rounded point, one which could hardly beused for penetrating purposes. (PI. 24, c.) Their function wasessentially different in that they wTere used in the fashioning ofstone implements. In the flaking of spear points and arrowheadsthe later Indians used this type of implement. To give a stone flakeits desired shape small fragments were removed by sudden pressureexerted through the blunt points of the flaking tools. The pictured 130 BUREAU OF AMERICAN ETHNOLOGY [BULL. 92 specimen was the only good one secured. Pieces from others of thesame type and material were uncovered, but they were too frag-mentary to serve for illustrative purposes. SCRAPER OR DRESSING TOOLAn unusual form of scraper or dressing tool was found. (PI.24, h.) It appears to have been made from a scapula or shoulderblade, possibly that of a deer, but the distinguishing features wereso obliterated in its manufacture that it is impossible definitely toidentify it. All of the edges were carefully smoothed down, but theconvex one shows a high degree of polish which could have come onlythrough long use. The small, slightly concave section at one end ofthe generally convex edge shows an additional feature in that oneside is sharply beveled, giving it an almost knifelike character. Justwhat the function of this implement may have been is not known.There is no doubt but that it had several uses, as did all of the tools,but it would seem that it was preeminently adapted for use in thedressing and tanning of skins. SPOONLIKE SPATULAThe first implement illustrated in Plate 24, a, is also of a ratherunusual character. It was made from one of the long bones of alarge mammal, quite likely a deer, but the work on it had been doneso carefully that no distinguishing marks are left. The bone wassplit down the center, the interior carefully scraped, the spoon orpaddlelike end given the desired shape, and the whole implementcarefully polished. Such an object certainly was adapted to a mul-tiplicity of uses. It could well serve in the nature of a spoon. Thebeveled edge at the end of the paddle portion would function nicelyas a spatula or as a scraper. One corner of this section of the im-plement is slightly rounded and shows a high degree of polish whichis clearly indicative of use. An almost identical specimen was foundaccompanying a burial discovered while excavating on the knollacross from the main village the preceding summer. NEEDLES AND BODKINSAnother group of implements which might be considered as avariation of the awl type but which are far superior in functionalpossibilities comprises the needles and bodkins. (PI. 25, a, b, c, d.)Implements of this form are comparatively rare and unfortunatelynot one complete specimen was secured during the excavations. Aperfect needle with characteristic " eye " was found in the remainsof the structure, referred to in a preceding paragraph, called Arroyo Roberts] SHABIK'ESHCHEE VILLAGE 131 House, located on the floor of the canyon just below the village, anda complete bodkin was recovered during the excavation of the proto-kiva house, but fragmentary specimens were all that came to lightduring the 1927 work.The inclusion of the first implement (pi. 25, a) in this groupmight be questioned. The large head would be a decided hindrancein any sewing operations but the tool would be very good formaking baskets. It might be considered as one form of bodkin,although it is quite possible that the perforation was for purposesof suspension only. Hung by a cord about the neck of the owner,it would not be lost as easily as the ordinary type of awl or punchand would be ready to hand on all occasions.The second implement (pi. 25, h) shows clearly that it was atone time perforated at the blunt end. During its use, however,the portion weakened by the " eye " broke off. The tool was not dis-carded, though, and the broken end was smoothed down, the imple-ment then serving as a very fine awl. The smoothing process didnot obliterate completely the "eye," traces of which are still quiteapparent.The next two specimens pictured (pi. 25, c, d) are fragmentsfrom the bodkin type of implement. They were made from broad,flat bones, probably ribs. The " eye " feature is quite apparent butunfortunately the points are missing and could not be found in thedigging. Such tools would be of value in weaving and the makingof baskets but certainly would not be as serviceable in true sewingas the characteristic needle form previously described. It is ofinterest to note that even in the early stages of southwestern develop-ment the people appreciated the advantages of an implement withan " eye " in it. WHISTLES AND BEADSThe last type of bonework to be considered is that of the objectsmade from the shaft portions of the long bones of small mammalsand birds. These are the bone tubes, including the whistles andbeads. Their manufacture was quite a simple process. All that wasnecessary was to cut off the ends of the bones and then to polish theobjects. When a whistle was desired it was necessary to do someadditional work and to make a vent or stop in the side.Plate 25, e, /, are two examples of the whistles or, as they some-times are designated, birdcalls made from such bones. The flutetype, tubes with a series of vents, does not seem to have been madehere during this period. Many fragments from the single-holedwhistles were found, but no evidence of the flute form was found.As in so many of the preceding examples, it is impossible to tell 132 BUREAU OF AMERICAN" ETHNOLOGY I bull. 92 what animal is represented. It is certain that the whistles weremade from mammal bones, but beyond that fact nothing more canbe said.A different form of whistle is represented by a small tube withtwo notches cut on opposite sides at one of the ends. (PI. 25, g.)Just how this was used is not clear, but two Zuhi workmen, CharliePinto and Bill Poncho, were insistent that it was a whistle. Theystated that a small cord was placed across the opening in the notchesand that the breath was blown across it, thus making the desired sound.They illustrated the method and obtained a sharp, hisslike note. Itis possible that this type of whistle is the prototype of, or bears somerelation to, the so-called Bitsitsi whistles used in one of the Shalakoceremonies at the present pueblo of ZufiL Mr. Hodge found boneforms of this present-day "vegetable-matter" whistle at the old Zuniruin of Hawikuh.83 The main point of difference, however, is thatthe bone examples from Hawikuh were in two pieces, fitted togetherand tied, while the specimen from the Chaco village is a completetube. This may be a sufficient difference to invalidate such a com-parison, but nevertheless both have the small notches at one end.The remaining examples of bonework belong to the so-called beadgroup. The latter includes all short bone tubes, cut and polished,which might have served for such a purpose. Plate 25, k, was madefrom the humerus of an unidentifiable bird. The two specimens,h and i, were not completed. Traces of the grooves made in cuttingthem are still apparent. The polishing and finishing process hadnot been resorted to. While rather indefinite in character these frag-ments are quite suggestive of rabbit bones. A finished specimen isshown in j. The last three specimens, ?, m, and n, were probablybeads. It is impossible to tell from what animal or bird they weretaken. Objects of StoneThere were not many objects fashioned from stone in the materialwhich came to light during the progress of the excavations, butthere are a number of different tools and implements represented inthe collection. Stone material seems to have been used chiefly formetates or milling stones, manos, mauls, knife blades, spear andarrow points, although some odd objects fashioned out of it wereobtained. METATESThe metates or milling stones were all of the same general typeand quite characteristic in form. (PL 26.) They were made fromcomparatively thin slabs of sandstone, and except for the groove or 83 Hodge, Hawikuh Bonework, p. 130. IDCM < 3CD oo _lozIHLU , 2 feet (60.96 cm.) at its greatest length; 1 foot6 inches (45.72 cm.) wide; 3 inches (7.62 cm.) thick; the groovemeasured 1 foot 4 inches (40.64 cm.) in length, 8 inches (20.32 cm.)in width, and 2 inches (5.08 cm.) in depth. &, 2 feet 1 inch (63.5cm.) long; 1 foot 5y2 inches (44.45 cm.) wide; 4 inches (10.16 cm.)thick; the groove was 1 foot 3y2 inches (39.37 cm.) long, 9 inches(22.86 cm.) wide, and 2 inches (5.08 cm.) deep, c, 1 foot 8V2 inches(52.07 cm,) long; 1 foot 2% inches (36.83 cm.) wide; 5 inches (12.7cm.) thick; the groove measured 1 foot 2y2 inches (36.83 cm.) long,9 inches (22.86 cm.) wide and l 1/^ inches (3,81 cm.) in depth.MAULS OR HAMMERSNo stone ax heads were found in the material from the houses anddump heaps, but the one-time occupants of the village seem to havemade considerable use of grooved mauls or hammers. A numberof good specimens of this type of stone implement were found andthere were many fragments from similar tools in the debris. Allindications point to two general types of mauls. One tended to thecylindrical form while the other was flat. Both forms had a groovearound the center for the purpose of hafting the head to its handle.The shape was no doubt due, in part at least, to the material avail-able for the makers. Rounded bowlders from the bed of a water-course would be excellent for making the cylindrical forms, whileany small block of stone would serve for the flat type.These hammers, and others like them, were no doubt used forquarrying the large slabs which were used to line the pit or excavatedportions of the storage bins and dwellings, for metate stones andother material needed in the day-to-day life of the people. Thesemauls may have served on occasions for driving pegs and posts intothe earth and even, possibly, for killing. The smaller forms arefrequently referred to as club heads. None of the Chaco villagemauls was as small, however, as those generally included in thelatter group. All of the specimens found were made from a fairlyhard sandstone. The material is soft enough to be easily workedand the heads were shaped, where the stone was at all altered, by thepecking and rubbing process. w Guernsey and Kidder, Basket-Maker Caves, p. 93. Roberts] SHAEIK'ESHCHEE VILLAGE 135Two characteristic examples of the flattened type of hammer ormaul are illustrated in Plate 27, a and d. Some effort was made toshape <7, but there was apparently no intentional alteration or modi-fication of a except the central groove. The former was peckedalong the edges with the evident intent of tapering the stone towardthe ends. They may have been quite pointed originally, but throughuse have become decidedly blunt. One side of the implement is veryflat and smooth, presumably the result of considerable rubbing, whilethe other is rather rough and uneven. This feature suggests thatperhaps the implement was made from a discarded mano or handstone used in the grinding of corn. The grooves around the centersof both of these implements are well made. They are deeper at theedges, the top, and bottom of the implement than at the sides.Maul a (pi. 27) is the smallest of the group. It has an over-alllength of 5 inches (12.7 cm.), a breadth of 3y2 inches (8.9 cm.),and a thickness of l2y? inches (5.71 cm.). Maul d measures 6y8inches (15.55 cm.) in length, 4y8 inches (10.5 cm.) broad, and 2y8inches (5.4 cm.) thick.One of the examples pictured (pi. 27, b) shows very few signsof use. There is practically no battering on the ends. It is wellshaped with one conical end and the other quite rounded. It is agood example of one form of the cylindrical type. The total lengthof this specimen is 5T\ inches (13.15 cm.), its breadth is 4T8e inches(10.61 cm.), and its thickness 3y2 inches (8.9 cm.).The second example of the cylindrical type (pi. 27, c) appearsoriginally to have been of the doubly conical form, but one end hasbeen so battered and broken through use that the shape is somewhataltered. It is a larger implement than the one described above andgreat care seems to have been taken in its manufacture. The surface,where not roughened by use, indicates careful rubbing and therecan be little question but that when the tool was first made it was aAvell-smoothed object. Its present size is: length 6y> inches (16.51cm.) ; breadth, 4T3e inches (10.61 cm.) ; thickness, 3 T3ff inches(8.07 cm.).The absence of the ax-head type of stone implement seems to berather characteristic of the earlier phases of the sedentary culturesin the Southwest. Mauls of the types described seem to occur quiteregularly but the edged tools are comparatively rare. Morris reportedfinding fragments and also a number of complete ax heads in his LaPlata work, but from the information available at the present writ-ing this seems to be a rather isolated case. The axes appear to havecome from a structure which was unquestionably more closely relatedto the Pueblo I period than it was to the characteristic phase ofBasket Maker III. The house type and other objects found indicate 136 BUREAU OF AMERICAN ETHNOLOGY [bull. 92 that the site was approximately on the border line separating thetwo stages.86At 17 Late Basket Maker sites in the region lying between thePiedra and upper San Juan Rivers, some distance east of the LaPlata district, the writer found many cylindrical and flat mauls butno axes. Guernsey and Kidder report neither from their true BasketMaker investigations,87 although three axes are figured in theirearlier paper dealing with the Basket Maker and Pueblo cultures.These specimens were definitely shown, however, to have belonged tothe later rather than the earlier phase.88 Mr. N. M. Judd found buttwo of the grooved axes during six seasons of investigations in theregion north of the Colorado River in southeastern and southernUtah. These two specimens were found in a canyon in which someof the old houses were located but which also includes in its ruinsthe remains of dwellings of the Pueblo periods. The axes seem tohave belonged to the latter rather than to the former cultural level.89Future work may bring more specimens to light, but in the presentstate of our knowledge it would appear that the ax was a little-usedimplement even in the Basket Maker III period. Under such cir-cumstances it is quite probable that the timbers necessary in the con-struction of the houses were felled and the required lengths obtainedby means of fire. The butt ends of some of the timbers which were still in place in some of the ruins excavated suggested that such amethod had been used. CHIPPED IMPLEMENTSThe group of chipped implements includes scrapers, knife blades,spearheads, and arrowheads. The materials from which these objectswere made vary somewhat but there is not a great number of differentkinds of stone represented. There is considerable variation in thequality of workmanship exhibited by the different specimens. Thisis in some measure attributable to the purpose for which the objectwas intended and in no small degree to the material itself. Oneform of stone chips differently from another. Some kinds are capableof taking a fine edge and of being worked to very thin blades whileothers will produce but a thick, coarse, rough-edge tool at best.Then, too, the human equation with its differing degrees of abilitymust always be borne in mind. Even in the most primitive groupscertain individuals are more skilled in some lines of work than theirfellows. 86 Morris, Antiquities of Southwestern Colorado, p. 201.87 Guernsey and Kidder, Basket-Maker Caves.88 Kidder and Guernsey, Archeological Explorations, p. 125, pi. 52, i, j, k.80 Judd, Archeological Observations, p. 146, pi. 50, d, h. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 92 PLATE 28 CHIPPED STONE IMPLEMENTS a measures 2J.2 inches by 1)6 inches (6.35 cm. by 4.13 cm.) BUREAU OF AMERICAN ETHNOLOGY BULLETIN 92 PLATE 29 MISCELLANEOUS STONE OBJECTS g measures 3J4 inches (8.3 cm.) in length robebts] SHABIK'ESHCHEE VILLAGE 137The scrapers (pi. 28, ?, b, d, f) are all rather crude in form butare sufficiently finished to serve their intended function admirably.A more perfectly shaped implement would do the work no betterthan these rough forms. Although they are of a nature which, inthe minds of many, might relegate them to the all-inclusive and non-descript class termed " rejects," they should not be so considered.All were found in houses in association with other objects of dailyuse and all show signs of wear. The chipping around the edges iscarefully done. The primary beveling of the edge by the removalof large flakes was improved upon by a secondary, more minutechipping which greatly sharpened the cutting or scraping portionsof the implements.Plate 28, ?, shows a flake of indurated shale which was struck oiffrom a larger block of the material, as is shown by the bulb of per-cussion at one end. The stone was carefully chipped all around itsedges and would be of considerable service in a variety of uses. Itcould be used for severing bone, for cutting and scraping purposesin the preparation of skins, for scraping poles, and even for shapingwooden objects.The second implement (pi. 28, b) is of the same material as thefirst. It is of an entirely different shape, however, and was chippedalong one edge only. The stone is an elongated triangle in crosssection, the apex forming the cutting edge. The latter received asecondary sharpening through the medium of the flaking tool. Theimplement would have done good service in the capacity of anunmounted knife blade or a side scraper. It apparently was quitepointed originally, but was broken off, presumably during use. Thisdid not lead to its immediate discardment, as the broken end showssigns of rubbing which could have resulted only from continued use.A good side scraper was made from a chip off a quartzite bowlder.(PI. 28, d.) This implement was chipped along its convex edgeonly. The opposite edge has been worn and smoothed considerablyand has the peculiar polish which comes only through use. This toolcould have served a variety of uses.One of the finest scrapers of the group (pi. 28, /) was made froma chert flake. The object is considerably thinner than the ones justdescribed and, except for a small section of the periphery at thebulb of percussion where it was struck off a larger bowlder, waschipped all around the edge. This chipping is too finely and care-fully done, as was true in the case of the other scrapers, to considerthe object other than a specific tool, although generalized as far asuse was concerned. As it is at the present time it would make aserviceable implement.9215??Bull. 92?29 10 138 BUREAU OF AMERICAN ETHNOLOGY [boll. 92A group of flaked implements which shows more careful shapingand which represents a more specialized form is that of the knifeblades. The largest of these (pi. 28, e) is a well-shaped blade ofchert. The edges are not as finely chipped as in some of the otherspecimens, a fact no doubt due to the material itself, but neverthe-less it would have made a very useful implement when mounted in awooden or antler handle. The second one illustrated (pi. 28, k) wasoriginally an excellent specimen, but through use or accident thepoint had been broken off. The blade was made from a thin chipof chalcedony. The cutting edges show unusually careful flakingand the surfaces were well worked down by the removal of large crossflakes. The point was broken off of the third specimen also. (PI.28, h.) This blade was not as finely shaped as the preceding onebut no doubt was just as effective in use. The cutting edges showcareful chipping but one side has a rather pronounced convex ridgealong the center of the blade. The grain of the material is suchthat this projection could not have been removed without breakingthe tool.Two types of material seem to have been popular for the makingof projectile points. This was no doubt due to the fact that bothare easily and readily chipped. One group of points was madefrom obsidian flakes, the other from chalcedony. Implements of thelatter group were found in slightly greater numbers, although theredoes not seem to have been an especially marked preference foreither. Caches of the raw material were discovered at several pointsin the village and there was about an equal amount of the two kindsof stone in each.The chief distinction between spear points and arrowheads is basedlargely on their size. The larger specimens are considered underthe former heading and the smaller are placed in the latter group.Although there were fragments from many broken spearheads onlytwo practically complete examples were found. The largest (pi.28, c) is longer than some of the knife blades, and while it might beconsidered by some as belonging with the latter group there are cer-tain features about it which show that it should be considered asa spearhead. Notched knife blades with tangs for mounting areoccasionally found, but experience has shown that in the earliercultures, at least, the notches of this group are set at an acute angleto the long axis of the blade. Spear points, on the other hand,have the notches set at right angles to the long line of the imple-ment.90 This spearhead was made from a large flake of chalcedonyand shows careful workmanship. The blade is thin, the bevel is not 80 Guernsey and Kidder, Basket-Maker Caves, p. 87, pi. 35. Roberts] SHABIK'ESHCHEE VILLAGE 139 pronounced, and the edges are comparatively sharp. The shoulders,at the notches, are square to rounded in form.The second specimen of a spearhead might be considered eitherunder this classification or as a large arrowhead. (PL 28, g.) Itwas made from obsidian, and although the edges are well chippedthe blade is quite thick. The tang is rather large but, as was notedin the example described above, the notches are at right angles tothe long axis of the blade.As will be observed from the points illustrated, the characteristicand predominant form of arrowhead was one with long, sharp barbs.The notches are set at an acute angle to the long axis of the bladeand are quite deep. The neck of the tang, because of the style ofnotch, is narrower than its base, so much so in many cases that thetang appears to have been set into the base of the arrowhead. Thesquare or rounded shoulder forms with the. notch at right angles(pi. 28, j, n, q) are decidedly rare. They became the predominanttype, however, in later periods in this region.The edges of all the points are well chipped, the blade is flat andthey no doubt had good qualities of penetration. All of the pointsare small and must have been intended primarily for small game,although a number of those pictured would have been large enoughto inflict a severe wound on a human being.All of the points illustrated were found ,in the houses or at thebottoms of the refuse mounds. None came from the surface. A greatmany specimens were picked up from the surface, but they are notincluded in this study because they can not be considered as definitelyrepresenting any period. A curious and perhaps significant fact isthat the great majority of the latter show the square shoulder andstraight tang instead of the form noted for those from the excava-tions. It seems rather certain that at the time Shabik'eshchee villagewas occupied the bow and arrow had gained a foothold and weredefinitely replacing the short spear and atlatl, the chief weaponsthus far known from the Basket Maker II period. Until quiterecently it was thought that the bow and arrow were unknown inBasket Maker III times because in some of the earlier sites belongingto this cultural stage there were no indications of the weapon.OTHER OBJECTS OF STONE There are several additional objects made from stone which do notfall into any general classification. In some instances no definitepurpose or function may be ascribed to them and in other cases anyuse suggested would be largely conjectural. Consequently they mustbe considered as being for the most part indeterminate. 140 BUREAU OF AMERICAN ETHNOLOGY [bull. 92 Plate 29, ?, is a block of calcite, which shows considerable smooth-ing and shaping. There is a cone-shaped cavity in one end whichwas carefully made. The depth of the latter is approximately halfof the total length of the stone. The bottom of the cavity is quitepointed and the sides show no signs of rubbing, so that it could nothave been used as a holder for a revolving implement, such as a firestick or drill. Its function is not known.Two objects which may have been fetishes, stones believed to havesome magic properties, were found in two of the houses. One ofthem is a piece of selenite, which might be considered as an elongatedform of the conventional heart shape. (PI. 29, b.) This stone showsno rubbing or smoothing, except what little natural shaping it mayhave received along its edges. It is practically transparent, but whenheld in certain lights it is filled with colors which have a rainbowlikequality which must have attracted an Indian and would no doubtsuggest to his mind certain attributes of significance. The second isan egg-shaped quartz crystal. (PI. 29, c.) Its form is natural, butseveral places on the surface indicate that slight projections or roughspots have been broken off, perhaps accidentally or possibly with someintent to improve the original shape.One of the objects (pi. 29, d) was made from a piece of pink-and-white indurated talc. From its general shape and characterthis stone might well have been used in the smoothing of pottery.The sides and ends show polish as from use. Such an implementwould have been serviceable in obliterating the bands of clay fromwhich a pot was built up and in shaping the interior walls of thevessel. The stone shows quite clearly where it was cut and brokenoff from a larger piece of material. The cut was not made com-pletely through the stone but about one-third of the way on one sideand approximately one-fourth of the distance on the other. Thiswas sufficient, however, to make it possible to break it off along afairly straight line. The rough spots remaining were then rubbeddown.One of the indeterminate objects is a block of jasper which hasone smooth side and one end partially rounded. (PI. 29, e.) Itmay or may not have any special significance. It was not found inany of the houses but was recovered from the lower part of one ofthe refuse mounds. Although smooth and slightly polished its sur-faces are too rough for it to have been used as a rubbing stone. Itmay simply have been picked up and carried into the village by oneof the children or its unusual form may have had a temporaryattraction for one of the older inhabitants, who took it home withhim only to discard it. ROBERTS! SHABIK'ESHCHEE VILLAGE 141 Another object, which might be included in the fetish group butwhich certainly is indeterminate in its character was made from apiece of lignite. (PI. 29, /.) It tends to the regular prism form,although the corners are rounded and one end is conical. All of itssurfaces are smooth and show a fairly high degree of polish. Whatits function may have been is an unanswerable question. It prob-ably belongs to that all-inclusive, frequently unsatisfactory, groupof objects in which, as a last resort, the archeologist places suchspecimens, called the ceremonial.One of the objects, made from red sandstone, has the general shapeand characteristics of the clay cloud blowers discussed in precedingparagraphs. It is conical in shape and there is a bowl in the largerend. (PI. 29, g.) The customary lengthwise perforation from thebottom of the bowl to the tip of the object is missing, however.There seems to be little question but that it represents an unfinishedsandstone cloud blower. Such objects are not unknown in the South-west, although they are by no means common.91There was another group of stones which could not be consideredunder any of the classifications mentioned above. It was found in abowl accompanying one of the burials and is composed of threeazurite crystals which show attempts at polishing and six pieces ofred ocher which also show rubbing. The latter no doubt furnishedthe raw material for some of the red paint used, perhaps, in thefacial and body ornamentation of the owner. This material is a verycommon source for the red paint used by the Indians of the South-west. A large fragment from a bowl was found in one of the houses,and in it was a cake of this red material which had been worked up,probably to use as paint. A chemical analysis was made of it by Dr.E. V. Shannon, assistant curator of the Division of Physical andChemical Geology of the United States National Museum, who kindlyfurnished the following memorandum : The material seems to consist of about equal parts of red iron ocher (oxide), silica, sand, and organic matter. A part of the latter is easily combustible andmay represent a grease binder of some sort. The balance is more refractory andmay be ordinary humus. There also seems to be present a considerable amountof fine-cellular bony matter.The present Pueblo Indians in the Southwest use grease in workingup their body paints and it is interesting to note that those whooccupied the area so long ago had the same custom. The sand andhumus in the material can be accounted for on the basis of mixturesince the structure went into ruin and was covered up. The bonymatter might possibly have been in the grease binder. It may have 61 Dr. G. P. Merrill, head curator of geology, U. S. National Museum, kindly assisted thewriter in the identification of the stones used in the manufacture of the various objects. 142 BUREAU OF AMERICAN ETHNOLOGY [bull. 92been that some of the marrow from split bones was mixed with thepaint and the particles included in that fashion.ORNAMENTS Stone in various forms furnished some of the material used in themanufacture of objects for personal adornment. Included in thisgroup are pendants, beads, and objects for use in inlays or mosaics.Pendants were made from hardened gypsum, turquoise, and hema-tite. (PI. 30, e, f,i.) A rather curious feature of the small col-lection of ornaments secured is that there are only two definitependants of stone material. The small hematite chip, /, shows care-ful rubbing and polishing but no perforation. That it had beenintended to serve for a pendant is shown, however, by an incompleteperforation. The maker started to drill a hole through the sub-stance, working from each side, but never finished it. Of the largenumber of turquoise flakes found in the excavations only one has aperforation. The others have been worked and polished and probablywere used as a part of a mosaic.The mosaic pieces seem to have been attached to some perishablematerial, perhaps wood or basketry, because in several instances theworked stones were found lying in a position which indicated thatthey had been attached to some object which had disintegrated. Itis known from other sites that the base for mosaics occasionally con-sisted of basketry and in other instances of wood.The stone beads were of two kinds, flat and tubular. The flatbeads were made from alabaster, while the tubular ones seem to havebeen lignite. (PI. 30, g, k.) Only one complete specimen, an un-finished bead, of the latter was found, although many broken frag-ments were observed. Shell Ornaments Shell was also used for the making of ornaments. One fragmentfrom a shell bracelet (Glycymeris) ; two pendants made from fresh-water mussel shells or river clams; one shell disk, unperforated,from a fresh-water mussel shell; and fragments from many olivellashell beads were found. (PI. 30, ?i, 5, c, d), h.) There is but onecomplete specimen of the latter. All of the others were broken orhad reached such a stage of decay that they crumbled to dust soonafter being removed from the ground.92 02 Dr. W. B. Marshall, assistant curator of mollusks, U. S. National Museum, identifiedthese specimens for the writer. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 92 PLATE 30 ORNAMENTS, STONE AND SHELL a measures ljs inches (3.5 cm.) in length x- ,*W I-< CM 3OQ CO >oo IHLUz