SI CIRC iiiii
111
Uiirib'riiiiiiiiiiwi:;:';!:





..JJLXM^tfU>^
SMITHSONIAN INSTITUTIONBUREAU OF AMERICAN ETHNOLOGYBULLETIN 121ARCHEOLOGICAL REMAINS IN THEWHITEWATER DISTRICTEASTERN ARIZONAPart I. HOUSE TYPES
By FRANK H. H. ROBERTS, Jr.
// ^^'

l,.s. ?
SMITHSONIAN INSTITUTIONBUREAU OF AMERICAN ETHNOLOGYBULLETIN 121ARCHEOLOGICAL REMAINS IN THEWHITEWATER DISTRICTEASTERN ARIZONAPart I. HOUSE TYPES
By FRANK H. H. ROBERTS, Jr.
UNITED STATESGOVERNMENT PRINTING OFFICEWASHINGTON: 1939
For sale by the Superinleiident of Documents, Washinston, D. C ........ Price 50 cents

LETTER OF TRANSMITTAL
Smithsonian Institution,Bureau of American Ethnology,Washington, D. C, May 9, 1938.Sir: I have the honor to submit the accompanying manuscript,entitled "Archeological Remains in the Whitewater District, EasternArizona, Part I : House Types," by Frank H. H. Roberts, Jr., and torecommend that it be published as a bulletin of the Bureau of Ameri-can Ethnology.Very respectfully yours, M. W. Stirling, Chief.Dr. C. G. Abbot,Secretary of the Smithsonian Institution. in

CONTENTS PageForeword xiIntroduction 1Remains of structures 17Group No. 1 21Structure 1 21Structure 2 33Structure 3 46Surface house and brush shelters near structures 1, 2, and 3 53Structure 4 58Structures 5a and 5b 62Structures 6, 7, and 8 67Structures 9, 10, and 11 81Surface house B and granaries 98Group No. 2 _ 102Structure 12 102Structures 13a and 13b 116Structure 14 and dance court 124Shelter and granaries 130Structure 15 and associated granaries 135Structure 16 and associated granaries 149Additional pit remains 160Structure 17__ 160Structure 18 166Unit-type structures 170Unit No. 1 171Unit No. 2 _ 194Unit No. 3 227Great Pueblo ruins 244Summary and discussion 253Literature cited 267Index 271T

ILLUSTRATIONS
Plates Pag?1. Map of Whitewater district 22. a. Looking across Whitewater Valley toward Puerco River, h, Arroyoin bottom of Whitewater Valley 2Interior of structure 1. h, Main framework for restoration ofstructure 1 28Brush and bark covering over framework, restoration of structure 1.h, Completed structure showing rounding dome and smoke hole-hatchway entrance 28Interior view of portion of restored structure. 6, Pillar left in ex-cavating pit, structure 1 28Masonry wall resting on slab foundation, h. Pilaster in structure 2__ 28Compartment side of structure 2. 6, Plaster pilaster behind woodenroof support 54Fire pit beneath wall of surface structure. 6, Opening into pit oven. . 54Stone ventilator aperture in structure 6. 6, Fire pit above floor infill of structure 6 54Ventilator side of structure 9. b, Portion of structure 11 54Pillar left from fill in pit for structure 10. 6, View across structures9, 10, and 11 -.. 104Floor in structure 12, showing ladder holes, deflector, ash and firepits, h, Plastered-over depression in floor of structure 12 104PiUar from fill, fire pit in fill, and logs on floor of structure 12. h.Pit for structure 13a 104Floor of structure 13b. 6, Portion of floor area structure 13b 104Structure 14. h, Upper end of ventilator shaft for structure 14 128Dance court as seen from end of Great Pueblo ruin. &, Portionof structure 14 underlying dance court and dais 128Burned roof timbers in pit of structure 15. 6, Series of slantingside poles for roof of structure 15 128Bowl below roof timbers in fill of structure 15. h, Jar containingcarbonized ears of corn lying on top of roof debris in structure 15. _ 128Pit and floor of structure 15. h, Compartment wall and ventilatorof structure 15 158Granaries and remains of shelter, structure 15 group. 6, Pit andfloor of structure 16 158Double burial beneath granary floor in structure 16 group. 6,View across Developmental Pueblo village 158Second unit in Developmental village. 6, Wall construction in por-tion of second unit 158Stone step and wall construction in room 8. h, Corner bin inroom 11 198Corner of building outside room 11 and bin. 6, Original floor inKivaB 198
VIII ILLUSTRATIONS Page25. a, Holes for slanting side poles at back of original bench top in Kiva B.b, Katcina niche in wall of Kiva B -- 19826. a, Subwall storage box in Kiva B. b, Ventilator openings in Kiva B. 19827. a, Surface structure in third unit. 6, Clearing debris from kiva forthird unit 23228. a, Wall construction in bench of third unit kiva. b, Holes for wain-scoting poles at back of bench in third unit kiva 23229. a, Floor in kiva for the third unit, b, Mound covering large buildingin Great Pueblo group of ruins 23230. o, Portion of Great Pueblo ruins, corner of building below and to leftof standing figure, b, Arroyo cut through refuse mound 232Text Figxjees
1. Group 1 pit and surface remains 222. Plan of structure 1 243. Diagram of pillar left from fill in structure 1 -- 284. Plan of structure 2 365. Plan of structure 2a 406. Plan of structure 3 477. Postulated reconstruction of cribbed roofing erected over pit of struc-ture 3 518. Plan of surface building A 549. Plan and section for pit oven 5610. Sections through surface remains A, pit oven, and structure 3, andshelter and structure 1 5811. Plan of structure 4 6012. Plan of structures 5a and 5b 6313. Surface remains B and associated pit remains 6714. Plan of structure 6 7015. Plan of structure 7 7416. Plan of structure 8 7917. Section through structures 6, 7, and 8 8018. Plan of structure 9 8219. Plan of structure 10 8620. Plan of structure 11 9021. Section through structures 9, 10, and 11 9222. Diagram of test pillar from fill in structure 10 9423. Postulated reconstruction of timber arrangement in superstructureerected over pits 9, 10, and 11 9924. Surface remains B in first group 10025. Group 2 pit and surface remains 10326. Plan of structure 12 10527. Diagram of positions of burned timbers on floor of structure 12 10728. Postulated reconstruction of superstructure erected over structure 12_ 10929. Diagram of pillar left from fill in structure 12 11130. Structure 13a, plan and section 11731. Plan of structure 13b 12132. Plan of structure 14 12533. Plan of dance court 12734. Plan of brush shelter 13135. Sections for structure 12, dance court and brush shelter 13436. Plan of structure 15 13637. Plan of structure 15 group of house, granaries, and shelters 146
ILLUSTRATIONS IXPage38. Plan of granaries A and B and shelter in no. 15 group 14839. Plan of structure 16 house and granary assemblage 15040. Plan of structure 16 15341. Plan of granaries A and B of the no. 16 group 15742. Plan of structure 17 16143. Plan of structure 18 16744. Plan of Developmental Pueblo village 17145. Plan of Kiva A 18146. Plan of brush shelter 20847. Plan of Kiva B, upper floor level 21048. Plan of Kiva B, original floor level 21349. Sections through portions of first and second units 21950. Plan of the third unit 22851. Plan of kiva for third unit 23452. Sections through portions of third unit 23853. Plan of large ruins of Great Pueblo period 246

FOKEWORDThe data furnishing the basis for the following report were ob-tained during the field seasons of 1931, 1932, and 1933. In thesummer of 1930, while the writer was engaged in investigationsat the Village of the Great Kivas on the Zuiii Reservation in westernNew Mexico,^ Mr. J. A. Grubbs, of Houck, Ariz., reported the ex-istence of a large group of ruins south of Allantown, Ariz. Uponcompletion of the project near Zuni in September of that year ageneral reconnaissance was conducted in the region bordering theArizona-New Mexico boundary line between the Zuiii Reservationand the Puerco River of the West. During the progress of this sur-vey Mr. Grubbs pointed out the ruins in question. In addition, manyscattered small village remains were noted in the immediate vicinityof the principal cluster. The possibilities for securing valuableinformation on the growth and development of one component of theprehistoric sedentary Indian culture pattern were so apparent thatpermission was obtained from Mr. Grubbs, who had acquired theland upon which they are located, to carry on a series of excavations.Work was started in May 1931 and continued through to the endof September. During the 1932 and 1933 seasons the investigationswere conducted in the months of June, July, and August.At the end of June 1931 the Laboratory of Anthropology at SantaFe, N. Mex., joined with the Bureau of American Ethnology andthroughout July and August the work was carried on as a jointproject. Four graduate students in anthropology assisted in the re-searches during the time the Laboratory cooperated in the under-taking. They were holders of fellowships granted by the Santa Feinstitution, for training in archeological field methods, and camefrom the departments of four universities. The men were: Carl F.Miller, University of Arizona; Solon T. Kimball, Harvard Uni-versity; Ralph D. Brown, University of Minnesota; and Dale S.King, University of Denver.The work of the 1932 season was carried on solely by the Bureauof American Ethnology, while that of 1933 was mainly under theauspices of the Laboratory of Anthropology, the writer being onleave of absence from the Bureau during the months of July andAugust. The personnel of the party for the third season consisted
1 Roberts, 1932.
XII FOREWORD
chiefly of graduate students, also holders of Laboratory fellowships.These men were : Erik K. Reed, Harvard University ; Joe Finkelstein,University of Oklahoma; Sidney J. Thomas, University of Texas;and Harold E. Cooley, University of Minnesota. Deric Nusbaum,Harvard University, joined the party during July and August as avolunteer worker.Linda B. Roberts supervised the cleaning and cataloging of speci-mens all three seasons. Secretarial work was performed by Ruth R.Butchart. Robert Kneipp, of Washington, D. C, assisted in theexcavations and around camp during the 1931 season. Similar dutieswere performed in 1932 by David Jones, Jr., of Gallup, N. Mex.Carl F. Miller was also a member of the 1932 party and assistedin the preparation of maps and plans of the ruins. He devotedconsiderable time to a study of the timbers used in the structures andby means of the Douglass tree-ring chart determined the cutting datesof many of the beams. From this evidence it is possible to give ap-proximate dates for the construction of the houses. Workmen em-ployed in the digging were Navajo and Zuhi Indians. The heartycooperation of all concerned aided the advancement of the researchesin no small degree.The main ruins have no name. The Navajos simply refer to themas "The Ruins." They were unknown to the Zunis and the whitesettlers in the vicinity do not have any particular designation forthem. Up to the time this report was written no work had beendone on the large Pueblo structures, consequently there is nothingabout these buildings to suggest a name.Information and materials accumulated during the three seasonsof investigations are so extensive that it is necessary to issue thereport in two parts. In this. Part I, house types and antiquities per-taining to that phase of the subject are discussed. Lesser objectsof the material culture, pottery, bone and stone implements, andother artifacts, are described in Part II, which will be published atsome subsequent date.The writer wishes to take this opportunity to express his per-sonal appreciation, as well as that of the institutions concerned, toMr. Grubbs for his generosity in permitting excavations on his landand for his cooperation in the promotion of the work.
ARCHEOLOGICAL REMAINS IN THE WHITE-WATER DISTRICT, EASTERN ARIZONA
Part I. House Types
By Frank H. H. Egberts, Jr.
INTRODUCTIONThe main site, where most of the investigations described and dis-cussed in the following pages were conducted, is located on the ridgeof mesas to the south of the Puerco Eiver of the West and west ofWhitewater Creek, an intermittent tributary of the Puerco. It is314 miles (5.633 k) south of AUantown, Ariz. (pi. 1). Specifically,it lies in the southwest quarter of sec. 34, T. 22 N., R. 30 E., Gila-Salt River meridian, Apache County, Ariz. The other excavatedruins are situated along the foot of the escarpment, below the prin-cipal group, at the west side of the valley.Topographically, the district is characterized by comparativelybroad valleys and flat-bottomed washes, and is cut by deep, narrowarroyos or ravines. The remains of horizontally laid sandstone beds,generally bordered by cliffs, form the high points.^ Inasmuch asthere is a variety of colors in them, these sandstone cliffs constituteone of the scenic attractions of the region. There are a number ofshades of red, ranging from dark to light, interspersed with layersof white. The brightness of the rocks contrasts with the diversehues of the valley bottoms and the slopes of the mesas.The region is a part of the major topographical division calledthe Colorado Plateau. According to one classification the archeo-logical site is located just south of the southwestern end of theManuelito Plateau subdivision, only a short distance east of themouth of Black Creek, called Bonito Valley and Defiance Creek onsome of the older maps, that forms the dividing line between theManuelito Plateau and Defiance Plateau provinces.' Another clas-
? Gregory, 1916, p. 26,
? Gregory, 1916, map, pL 1.
2 BUREAU OF AMERICAN ETHNOLOGY [Boll. 121sification places it in the southern part of the Navajo section of theColorado Plateau, an area characterized by young plateaus withsmaller relief than the canyon lands section to the north, a regionmarked by recent erosion and retreating rock scarps.* In general itmay be said that the topography in the immediate vicinity of theruins consists of gentle valley slopes and comparatively flat mesatops (pi. 2, ?, h). The valleys are dotted with some sagebrush,greasewood, rabbit bush, and many kinds of weeds. Extensivegrazing in recent years has thinned the grass and forage plants toa marked degree. The mesas are covered with cedar, pinyon, andsome yellow pine trees. The latter are more prevalent on the higherelevations north of the Puerco than they are along the southernmesas. In the open spaces on the mesa tops there is much more brushthan in the valley bottoms.Animals present in greatest numbers are the rabbit and prairie dog.There are some coyotes and porcupines, and a few fox and wildcat.Deer and antelope formerly roamed the countryside and the bearwas not unknown. There are not many kinds of birds in the vicinity,although the raven, nighthawk, hawk, a sporadic eagle, the pinyonjay, owl, dove, and swallow are observed. At one time the turkeywas plentiful, judging from the bones found in the refuse mounds.Insects are common and there are several kinds of snakes, althoughthe latter do not occur in large numbers.The soil in the valley bottoms and on the mesa tops, where therehas not been too great a sand drift, is capable of producing goodcrops when there is sufficient rainfall or where water is available forirrigation purposes. If the many ruins scattered throughout thedistrict may be considered as a criterion, it at one time supporteda rather numerous population. It is possible that prior to the cuttingof the deep arroyo channel in the bottom of Wliitewater Valley(pi. 2, h), as well as that of the larger Puerco, the whole flats couldhave been devoted to fields for corn, beans, and other agriculturalproducts such as the pre-Spanish occupants of the region raised. Therun-off from the higher slopes could have been used for flood-waterirrigation, thus providing, in addition to that normally falling onthe fields in the form of rain, sufficient moisture to insure good crops.There are no evidences of definite irrigation ditches or reservoirs.Tradition and the memory of Indians is at best doubtful evidencewith respect to conditions in the past, but a number of aged Navajoin this district insist that when they were boys there was no arroyoin Whitewater Valley and the Puerco flowed, during the periodswhen it carried water, in an ordinary stream bed and not in a deepchannel such as it occupies at the present time. If such was the case
*Fenneman, 1928, p. 342, map.
BUREAU OF AMERICAN ETHNOLOGY BULLETIN 121 PLATE 1
I- ?/'"^ V-? RUINS '/"'" '""'<-"^.^?;4.?-'?..-:y ' ,--5.%*~- <
: % sf , '?"??.,?(". 4n..,ii,. j?, ?^ ,u:^?"S!'?i'-"<S- ? - -.vr.f .-to'Hw ,- i .?..-. ,, ,.-, -.
'rt<_,,vW"j; i-rs,,,*
^w
(.h \ i^
^iv"^ 4w J *% 1/ C ~--u \ '4 K^ >^'"f /:^
Zufii
Map of the Whitewater district. Inset shows location of ruins with respect to general southwestern area.
BUREAU OF AMERICAN ETHNOLOGY BULLFTI^.i 121 PLATE 2
i/ioking across Whitewater Valley toward Puerco River.
h. Arroyo in bottom of Whitewater Valley.
ROBKBTS] ARCHEOLOGICAL REMAINS IN ARIZONA 3when the site under investigation was occupied, the flats below theruins would have provided an ample food supply. In this sameconnection it is interesting to note that reports by early explorersin the region do not mention deep-cut stream beds. It was only afterthe Whipple Railroad Survey party reached the district beyondNavajo Springs, 25 miles (40.234 k) west from the Allantown neigh-borhood, that reference is made to an arroyo of any depth.^ Thereis a notation that a bridge would be required at that place if a rail-road was built along the Puerco. Lieutenant Simpson crossed thePuerco of the West in 1849 en route from Canyon de Chelly toZuni. The crossing took place in the vicinity of the present settle-ment of Manuelito (pi. 1) and his diary contains a reference to thestream bed which was bordered by a few cottonwood trees.? Nomention is made of the height of the banks but it may be assumedthat it could not have been very great, since elsewhere in his journalLieutenant Simpson always takes care to state the amount of worknecessary to make deep gullies traversable for wagons and artillery.Had conditions at that time been similar to those existing today heundoubtedly would have made some mention of the fact, because inthe entry for that date he specifically describes the labor whichwould be involved in making other portions of the route accessible.The problem of the water supply for the various villages whoseremains occupy the immediate vicinity south of Allantown is onewhich has not yet been satisfactorily solved. There are a numberof springs along the course of Whitewater Creek, but these are at aconsiderable distance from the groups of ruined houses. If the sup-ply was no closer during the period of prehistoric habitation thesituation would have necessitated the transportation of water for adistance of about 2 miles (3.219 k). Some of the older maps indicatesprings and pools along the course of the Whitewater in locations muchnearer to the main archeological site than those now in existence.The filled-in remnants of old depressions suggestive of former pondsand places which probably were rather swampy may be seen alongthe banks of the arroyo. At the head of a small gully terminatingat the base of the low cliffs forming the edge of the mesa about one-quarter of a mile (0,402 k) from the main site are indications of theformer existence of a spring. There are large quantities of potsherdsscattered about at this place and in every instance the fragmentsnoted were from large jars such as were used for water and storagepurposes. This suggests that the onetime occupants of the nowruined houses obtained a portion of their water supply at that place,an occasional container being broken in the process. It is likely thatthe cutting of the deep arroyo in the bottom of the valley lowered
" Whipple, 1856, pt. II, sec. V, p. 28.? Simpson, 1850, p. 112.
4 BUREAU OF AMERICAN ETHNOLOGY [Boll. 121the water table of the region sufficiently to cause the drying up ofnumerous springs located along the bordering cliffs. Furthermore,drifting sand may have buried such sources as had not previouslydried up.At present the climate is marked by considerable variations. Thewinters are cold and the summers quite hot, but despite the heat ofthe summer sun the nights are as a rule chilly. Due, no doubt, to the0,300 feet (1,920.238 m) altitude. Some winters are accompanied byunusually heavy snows ; others are exceedingly dry. During the sum-mer season, particularly in the months of July and August, therenormally are heavy rams which frequently flood the country andfill the generally dry stream beds with raging torrents. A secondshort rainy spell usually occurs in September. Throughout the re-mainder of the year, it may be said, precipitation is deficient. Underaverage conditions, however, there is sufficient moisture for the grow-ing of such crops as are necessary to sustain life. Over a long periodof time the Indians of the uplands area developed types of corn, beans,pumpkins, and melons which germinate, flourish, and reach maturityduring the short growing season which the region provides, and ex-cept in periods of drought or when hordes of insects descend on thefields, the yield is ample. There is nothing to indicate that therehave been any pronomiced climatic changes between the time of thefirst occupation of this district by the Indian peoples and the presentand it seems logical to suppose that they coped with conditions notunlike those of today.The cultural remains present at the sites where the investigationswere carried on represent a number of stages in the development ofthe sedentary peoples of the plateau region. At the local of the maincluster of ruins there are vestiges of three or four distinct and sequentphases. In order that there may be a better understandmg of thesignificance of the various finds, a brief summary of the presentstatus of southwestern archeology is advisable. Continued researchin the ruins of the Southwest during the last 80 years has broughtto light many data concerning the history of the region. Investiga-tions since the termination of the World War have been especiallyproductive and the period since 1927 has been marked by a tremendousincrease in knowledge. At the present time archeologists group theremains located in the States of New Mexico, Arizona, southwesternColorado, Utah, eastern Nevada, western Texas, and northern Mexicointo two major divisions characterized by differences in materialculture, topogi-aphical environment, and geographical distribution.One division or province had its location in the uplands or plateauportion of the area; the other centered in the desert region. Theuplands group is called the Anasazi, a Navajo name for the peoples
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA 5who built and lived in the ruins which dot the plateau, and those inthe lowlands are referred to as the Hohokam, the name used by thePima to denote the ancient ones inhabiting the desert precincts.What may eventually prove to be a third and lesser province is onewhich hitherto has been thought to represent a regional variationof the Anasazi. There is a possibility that it may emerge as a sub-pattern resulting from a fusion of very early Hohokam and Anasaziin a peripheral area, but at present it is regarded tentatively as aseparate entity and is called the Mogollon. The name is taken froma prominent range of mountains in the district where the remains arefomid. There are peripheral precincts where the patterns are notclear-cut. In these outlying reaches many features common in thenuclear districts are missing and there is a progxessive fading of thebasic patterns in proportion to the distance from the central portionsof the provinces. These marginal regions are generally referred toas the "eastern, northern, western, and southern peripheries." Theymay in time be set apart as minor subdivisions, but at present theyare considered as local variants of the major divisions. Developmentsin the two major provinces followed somewhat parallel lines, butmitil comparatively late horizons were for the most part seeminglyindependent. In the lesser, tentative third province the imfolding ofthe cultural pattern progressed through a series of stages broadlysynchronous with those of the main two, but on the whole was moreclosely related to the Anasazi than to the Hohokam.The Anasazi remains are found in the regions of the San JuanRiver, the Eio Grande, the Upper Gila and Salt Elvers, the LittleColorado River, most of Utah, and a portion of eastern Nevada. Theuplands of this plateau area with its rugged mesas, narrow canyons,and broader valleys furnished the setting for an interesting culturalflorescence. From a very simple beginning the people who inhabitedthe region passed from a nomadic hunting existence to a relativelyhighly advanced mode of life. To simplify the study of this develop-ment the various stages through which the people passed have beengrouped under two main headings, names better known than thatof the main designation Anasazi, called Basket Maker and Pueblo.Each of these major divisions has been separated into a number ofminor groups. The Basket Maker, which evidence has shown to bethe older, has two, although until recently it was listed with three.The Pueblo, which followed and owed much to the Basket Maker,has five. The growth and development of the various stages did notfollow a smooth and ordered progression, but advanced intermittentlywith periods of quiescence during which there was little change. Itis the material from the intervals when conditions were static thatfurnishes the picture for each typical horizon. The boundary lines68764?39 2
g BUREAU OF AMERICAN ETHNOLOGY [Bull. 121between periods are often vague and there is an overlap of charac-teristics which sometimes tends to be confusing. These occurrencesgenerally may be explained through careful consideration of all ofthe factors involved. Wliile the progression of stages in the culturesequence implies a certain degree of contemporaneity between sites ofthe same horizon, it does not imply necessarily that they will fallwitliin identical chronological dates. There may have been, and nodoubt frequently was, a difference in actual years in which objectswere in vogue in various districts. For example, during a certainstage in one portion of the area a particular style in ceramic art mayhave been very popular, so popular, in fact, that it spread into othersections. Like groups of objects found at two widely separatedruins would indicate a degree of similarity in age between two sites.In a broad sense they belong to the same stage of progression andyet in actual years may have been separated by an appreciable inter-val. This tendency to a chronological discrepancy in direct propor-tion to the distance from a center of origin is well illustrated in ourown culture. During the era prior to the development and improve-ment of the motion picture and radio, the styles of the day in NewYork were the styles of two and three years later in the more remotesections of the country. The popular music of Broadway did notreach the farthest corners until long after it had been forgotten in themusic halls where it was introduced. This lag in cultural traits waseven more marked in earlier times. Failure to consider this tendencyhas in the past caused some confusion in the minds of southwesternworkers and students and for that reason is emphasized here.During the decade following the first conference of southwesternarcheologists and workers in related fields, held at Pecos, N. Mex.,in the latter part of August 1927, the classification adopted by thatgroup was widely used by writers on the subject. The Pecos Classi-fication listed three divisions for the Basket Maker, denoted by thename and numerals I, II, III, and five for the Pueblo, again using thename and numerals I to V. Because of some confusion arising fromthe implications of chronology embodied in the numerical qualifiers,the writer suggested a new terminology in an article in the AmericanAnthropologist ^ and added one additional name in a subsequentreview of southwestern archeology appearing in American Antiquity.*Inasmuch as this later classification has been accepted and used bya number of investigators it will be used in this report. It was de-veloped by the substitution of descriptive names, terms suggested bythe characteristics of the period, for the numerical designations-Some of these names were provided as alternative titles at Pecos,others were used by certain workers prior to the first conference, and
'Roberts, 1935.* Roberts, 1937.
ROBBBTS] ARCHEOLOGICAL REMAINS IN ARIZONAthree were the writer's own choice. They were applied to a modifiedform of the Pecos Classification published in a number of the writer'sprevious reports in the bulletin series of the Bureau of AmericanEthnology. Thus at present the period designations and their earliercounterparts are as follows:
Pecos Classification
o BUREAU OF AMERICAN ETHNOLOGY [Boll. 121gion was thinly populated by a widely distributed, simple nomadichunting people. Their food consisted of game and such wild fruitsand vegetable products as could be found. On occasions they mayhave erected flimsy brush shelters against the vagaries of the weather,but more probably relied on caves for such temporary protectionfrom the elements as they deemed necessary. In the course of timethey took up agriculture, having obtained corn (maize) and a knowl-edge of how to plant and cultivate it. Whence this product came orhow it reached them is not known, although it is supposed that it wasintroduced from the Mexican area to the south. Eventually theircrops became more abundant and as a consequence the peopleadopted more sedentary habits. Bountiful harvests mean a surplusof food which can be stored against future needs. Well-stockedgranaries require a certain amount of protection which can not beprovided if the owners wander too far afield. Hence small communi-ties assembled around the storage-bin nucleus. It is at this point thatthere is first definite knowledge of the people. The hypotheticalseries of events leading up to this stage constitute the Basket Maker Iof the Pecos Classification, the conjectural stage omitted from thepresent revised nomenclature.Basket Maker granaries were small pits dug in the floors of cavesand were lined with large stone slabs. They were covered with fun-nel or domelike superstructures of poles, brush, bark, and mudplaster. Little is known of the dwellings of this period, as only afew ruins, possibly attributable to it,^ have been noted and informa-tion obtained by excavation is not available. Numerous caves contain-ing granaries and other remains have been investigated and nohouses found. Impermanent brush shelters wliich have left no tracesmay have been used in some cases. There is evidence that slabcists lined with grass and bark were sometimes used for sleepingplaces, but these hardly could be called houses. The main objects ofthe handicrafts of the period which have been preserved are excel-lent baskets, sandals, ropes, nets, twined and woven bags, and robesmade from fur cloth. There was no true pottery, although at aboutthe end of this stage unfired clay vessels with grass or cedar barkbinder made their appearance. The idea of making clay vesselsprobably came from the south, but the development was entirelylocal. The main weapons were an S -shaped club and a short javelin,the latter thrown by means of an atlatl or spear thrower.Modified Basket Maker is characterized by some additional fea-tures. True pottery with painted decoration was developed. Semi-subterranean dwellings of a comparatively permanent nature wereerected. Several kinds of corn were grown where previously there
? Hargrave, 1935, p. 42.
ROBEETS] ARCHEOLOGICAL REMAINS IN ARIZONA Qhad been but one. Beans appeared in the list of agricultural prod-uce. Kobes made from feathers were occasionally employed. San-dal types were changed and baskets became less important in thecomplex. The bow and arrow began to replace the spear throweras a weapon. Tliis feature probably correlates with another factor,one which had a marked influence on subsequent developments.Namely, new peoples were drifting into the area and it is possiblethat they were the bearers of this different kind of weapon, althoughit seems to have preceded them as a diffused cultural trait. This newgroup did not sweep over the plateau as an invading horde, butmade a gradual penetration in successive small bands. Their arrivalis recorded by differences in skeletal remains. The Basket Makerswere long-headed or medium long-headed, wMle the newcomers werepossessed of broad skulls. The latter are found only in burials madetoward the end of the Basket Maker era. These immigrants seem-ingly brought little with them beyond the bow and arrow, possiblythe grooved ax, and a capacity for cultural development. They tookover, changed, and adapted to their own needs the material cultureof the older inhabitants and launched the second major stage in thehistory of the Anasazi, that of the Pueblos.There were many changes during the course of the DevelopmentalPueblo stage. These were not brought about immediately, as theearliest phase was one of transition and instability. Eventually newfeatures appeared and the foundations were laid for future develop-ments. Cotton and materials made from it became one of the im-portant items in the industry of the people. Pottery was moreextensively made and the ceramics took on definite features typicalof the period. Fur-string robes or blankets were replaced to aneven greater degree by those made from feathers. The wild turkeywas domesticated; previously the dog was the only tamed creature.The grooved ax became an important tool. Apparently cradlingpractices were changed as crania from the graves of this and subse-quent periods show an occipital flattening. Broad skulls found inassociation with INIodified Basket Maker remains do not have sucha deformation. In the nuclear parts of the province the single-room,semisubterranean dwellings gave way to structures with only slightlydepressed floors instead of pits. The major portion of the house wasabove ground and had several contiguous rooms. In some cases thewalls were of pole and plaster (jacal) construction, and in othershorizontally laid stones were employed. These were replaced bymasonry houses erected entirely above ground. Buildings of thisnature were a single story in height, had flat roofs, and containedfour, six, eight, sometimes even more, rooms. The rooms were placedin a long single row, a shorter double row, an L-shape, or in themore highly developed forms in the shape of a rectangular U the
jQ BUREAU OF AMERICAN ETHNOLOGY [Boll. 121wings forming an enclosed court at one side of the building. Thiskind of dwelling is called the single-clan or unit type. In the pe-ripheral districts semisubterranean houses continued in use for sometime, although in modified form. Pits for the underground portionwere dug deeper and entrance to the chamber was by means of aladder through the smoke hole in the roof instead of by an entrywayat one side. The latter was retained, however, in reduced and alteredform and functioned as a ventilator. These structures eventuallywere abandoned in favor of others approximating the unit type.These new dwellings did not reach as high a degree of excellence asthose in the nuclear districts. In the south and west, pole-and-mudhouses and buildings consisting of irregular agglomerations of roomswhose walls were formed from large quantities of adobe mud andunworked boulders constituted the living quarters. In the Flagstaff,Ariz., area rectangular semisubterranean houses prevailed to theend of this horizon.The new-type dwellings necessitated some provision for a place inwhich to hold the religious rites and ceremonies that had previouslybeen performed in the old circular domiciles. Instead of changingthe rituals to meet the requirements of the new houses, each groupprovided one of the old style structures. It generally was placed atthe south or southeast side some distance from the dwelling. Eventu-ally these special chambers became more formalized and developedinto what is called the kiva. They did not become as highly special-ized in the peripheral districts and in some sections the circular formis entirely missing, yet its purpose was served by analogous rectangu-lar structures.During the period when these changes were taking place the peoplelived in small villages scattered throughout the province. Then anew trend set in. The population began to concentrate in large urbancenters and the Great Pueblo period was at hand. Extensive ter-raced houses, several stories in height and containing many hundredsof rooms, were built on canyon floors, in the caverns in cliffs, andon mesa tops. The ceremonial chambers were included in the mainblock of the building and their former subterranean character wassimulated by filling with earth the spaces between their circular wallsand the rectangular ones enclosing them. Pottery forms and stylesof decorations crystallized and became so highly specialized that eachcenter had its own characteristic wares. The rise of these urbancenters was more the result of an ever-growing tendency on the partof the people to abandon the outlying small houses and concentratein various communities than to a sudden and marked increase inpopulation. Wliat caused this movement is not known. It probablywas induced by a combination of factors. Severe droughts occurperiodically in the Southwest and such an occurrence may have made
ROBKKTS] ARCHEOLOGICAL REMAINS IN ARIZONA Hsome districts so unproductive that the occupants were forced tojoin their kinsmen in sections where conditions were less rigorous.In order not to encroach too much on tillable land the dwellings wereconsolidated and evolved into the typical pueblo or apartment houseof the period. At the same time there was a constantly increasingpressure from the nomadic tribes of the borderlands. Droughtconditions would make marked inroads on the supplies of seeds, nuts,wild fruits, even on game, and small unprotected villages with re-serves of corn and beans would be tempting plunder. That raidsdid occasionally take place is indicated around the peripheries of theprovince by the remains of unit structures which give evidence thatthey were pillaged and their inhabitants slain. For this reason thedesire to gather in populous centers as a means of defense may havebeen an important element in the trend of affairs.The large towns thrived for a time and then began to decline.The Regressive period was setting in. In some of the centers thepeople, although reduced in numbers, carried on. Other communi-ties were abandoned. There was a pronounced withdrawal from themore northern districts and a definite drift toward southern portionsof the area. Causes for this phenomenon have not been determined.Unquestionably there were several. Encroaching and plunderingnomads evidently became troublesome during this period, as manylarge ruins show that attempts were made to fortify the villages.Internal discord and factional strife may have arisen and as a resultlarge groups m.oved out to settle elsewhere, a thing that has hap-pened in historic pueblos and no doubt did in earlier times as well.Then there were recurring droughts. A series of prolonged cropfailures may have reduced the resistance of the people to such an ex-tent that they gave up the struggle and started off in search ofbetter locations. Irrespective of what the actual causes were, therewas a marked recession from the former cultural peak and the erawas one of instability and migration. The northern part of theprovince, that traversed by the San Juan River and its tributaries,ultimately was deserted. New villages and communities developedalong the Rio Grande and the Little Colorado River. After thesewere established the arts and industries took on new life and theRenaissance was under way.Outstanding in the Renaissance period was the growth of variouscenters along the Rio Grande and the Little Colorado River. Thehighest stage of development in these two regions was attained duringthis stage. It, like preceding horizons, had certain typical featuresin pottery types and other objects of the material culture that pro-vide criteria for identification. The arrival of the Spaniards in thelatter part of the period, 1540, served as a check on the progress ofthe native culture, but the full impact of European influence was not
^2 BUREAU OP AMERICAN ETHNOLOGY IBulu 121felt until after the revolt, reconquest, and complete capitulation ofthe Pueblos in 1700. The Historic period begins then and continuesthrough to the present. It is the era of the modern villages. Thestage in general shows the gradual replacement of the Pueblo culturepattern by traits introduced by the white man, a process continuingwith marked acceleration at the present time.^?Several diagnostic traits are used in identifying the various stagesin the Anasazi pattern. The same traits apply whether the PecosClassification or the revised one discussed in preceding pages is used.Skeletal material is thought significant for the two major groups,Basket Maker and Pueblo, although elements in the material cultureare so well differentiated that they probably furnish a more reliablecriterion. Indicative components in the complex are : House and vil-lage types, architectural features, textiles, baskets, sandals, picto-^graphs, stone and bone implements, kinds and styles of ornaments,and pottery. Pottery is considered to furnish the most abundant, con-venient, and reliable criterion, with the culinary or utilitarian waresthe simplest for sequence determinations. The reason that potteryplays so important a part is that it is characterized by easily noteddifferences in style and form and it was an exceedingly sensitive ele-ment from the standpoint of variations in both time and place. Theseare factors, however, that will be discussed in greater detail in thesecond part of this report, that dealing with the artifacts, and neednot be considered further here.The sequence of the several stages outlined in the foregoing pagesis based on evidence obtained mainly from excavations. The mostimportant type of proof was that of the stratigraphic relationship be-tween different forms in the complex. In places undisturbed layersof deposits containing objects of a particular form and style werefound overlying other layers in which similar objects exhibited differ-ent characteristics. There was no question but that those from theupper level were more recent and the variation in objects furnished adefinite standard by which to gauge their relative ages. In addition,changes and developments in articles of the same type were demon-strated. Modified Basket Maker remains have been found above thoseof Basket Maker in so many sites that there can be no doubt of theirsequence. Elsewhere ruins of Pueblo structures have been notedresting on mounds covering the remains of Modified Basket Makerhouses, and Developmental Pueblo dwellings have been discoveredbeneath the foundations of Great Pueblo buildings. Correlationsbetween sites and districts were worked out through study of tradeMFor a more detailed discussion of the evolution and growth of the Anasazi (BasketMaker-Pueblo) cultural pattern and a review of the archeological progress culminating inthe Pecos Classification see Roberts, 1935. For references and examples of ruins and sitesbelonging to the various periods see Roberts, 1932.
ROBBBTs] ARCHEOLOGIGAL REMAINS IN ARIZONA 13objects and their positions in local sequences. The entire structureof Anasazi history was erected on this foundation. That it is essen-tially correct has been shown by the development of a new methodfor determining not only relative age but actual dates. This new aidto the archeologist is dendrochronology, the tree-ring calendar dis-covered and perfected by Dr. A. E. Douglass. Dates obtainedthrough this medium demonstrate that the stratigrapliically deter-mined sequence is true in its main outline and that the relative agesof numerous large ruins and village sites, even of remains in differentdistricts, were correctly deduced by the archeological method.Dr. Douglass, of the University of Arizona, in studying sun spotsand their effects on climatic conditions, turned to the growth ringsof trees in an effort to obtain evidence on the occurrence of droughtperiods and the intervals of moisture. In doing this he discoveredthat definite ring patterns, as distinct as human fingerprints, re-corded specific year groups and as a consequence developed a systemwhereby he can tell the year when a log was cut from a living tree.Beginning with trees whose actual cutting date was known, he hasbeen able to devise a type ring chart going back to the beginningof the Christian era. To secure evidence to substantiate his owntheories he was forced to resort to timbers from ruins for materialantedating living trees and thus furnished the archeologists with avaluable time scale.^^ Wlien beams are found in ruins it is possibleto check their rings against the type chart and, provided the outersurfaces have not been damaged or removed, tell the year of theircutting. The timber may not have been placed in a house immedi-ately after it was cut and occasionally a log was reused, but suchfactors can be checked by the archeological aspects of the site and adate is assured which approximates closely the year or years whenthe dwellings were erected.Explanations for certain features noted in the cultural patternare also furnished by dendrochronological studies through informa-tion on the occurrence of drought and periods when conditions weremore favorable. Dates for a number of the droughts correspondto definite phenomena in Anasazi history. One dry period correlateswith the trend toward concentration in urban communities at theend of the Developmental period. Another was undoubtedly animportant factor in the abandomnent of some of the centers at theend of the Great Pueblo stage. The greatest expansion and growthin one of the leading centers of the Great Pueblo era took place ina 20-year period when conditions were favorable. Wlien more datesare available and additional information has been obtained fromseveral districts that are not too well known at present, an inter-
im Douglass, 1932, 1935.
2^4 BUREAU OF AMERICAN ETHNOLOGY [Boll. 121esting study of the effects of weather on the movements of peoplescan be developed. Present evidence only hints at the possibilities.There are no Hohokam remains in the Whitewater district, henceonly a summary mention of the general character of the desert cul-tural pattern is necessary. According to present information therewere six main stages in the Hohokam. They are called Pioneer,Colonial, Sedentary, Classic, Kecent, and Modern. Although thetwo patterns overlapped to some extent along the hazy boundaryline between the two provinces, the Hohokam and Anasazi were quitedistinct until about the year 1000 A. D. After that date Anasazifrom southern parts of the plateau began to drift into the Hohokamprovince and establish communities there. The two seemingly livedside by side for a time and then the northern group withdrew andreturned to the uplands. During the Developmental Pueblo periodthere was a northward thrust of Hohokam into the periphery of theplateau province in the Flagstaff, Ariz., region.There are no dendrochronological dates for the Hohokam. Timbersused in house construction were of varieties of wood not adapted totree-ring studies. The determination of periods depends in largemeasure on pottery types. Some dates can be suggested on the basisof trade objects, as in the case of the Anasazi penetration, but on thewhole the status is much the same as that of the Anasazi prior to theperfection of the tree-ring calendar. The time factor is largely postu-lated. Present evidence is that the Pioneer stage was roughly con-temporaneous with Modified Basket Maker and the beginnings ofDevelopmental Pueblo, Colonial with Developmental and early GreatPueblo, Classic with late Great Pueblo and Regressive, Recent withRenaissance, and Modern with Historic.The main differences between the Anasazi and Hohokam may besummarized briefly. The Hohokam practiced cremation ; the Anasazi,in all stages, buried their dead. The Hohokam built rectangularsingle-unit houses of pole, brush, and plaster construction during allstages; while the Anasazi progressed from circular or rectangularsingle-unit dwellings of poles, brush, and plaster to multi-storiedcommunal houses built of stone. The Hohokam progressed fromfloodwater irrigation to extensive canal systems for supplying theircrops with water ; the Anasazi depended mainly on floodwaters but insome sections did employ series of small ditches. The Hohokam hadextensive carving in shell, which was rare in the Anasazi pattern.The Anasazi domesticated the turkey; the Hohokam apparently didnot. Pottery made by the two groups differed in certain respects.The coiling method was used in both cases but the finishing processesvaried. The Anasazi smoothed the surfaces of their vessels withscrapers and polishmg stones; the Hohokam completed theirs by em-
ROBBBTs] ARCHEOLOGICAL REMAINS IN ARIZONA 15ploying a paddle and anvil. The Anasazi painted wares over a longperiod were of the type with white background and black designs,followed by a series of polychrome forms; Hohokam painted vesselshad red designs on a buff background, a factor that was responsible forthe name first used to designate the group. Earlier papers call thepattern the Eed-on-Buff Culture.^^The Mogollon pattern seems to center in the San Francisco andMimbres River valleys in southwestern New Mexico, although tracesof it are found extending in all directions for considerable distances.The basic feature is that of a sedentary agricultural-hunting com-plex, the hunting aspect being more pronounced than in either theAnasazi or Hohokam. Several stages are recognized in the pattern.These are called Georgetown, San Francisco, Three Circle, Mimbres,and Animas. Chief characteristics of the Mogollon are that it hadrounded semisubterranean houses followed by rectangular semisubter-ranean structures, then Pueblo type dwellings. Pottery was made bythe coiling process with the scraper and polishing stone as finishingimplements. The dead were buried as a rule, but there was somecremation. The atlatl and bow and arrow were used for weapons. Inthe beginning the people were a roundheaded group with low skullvault. They did not practice cranial deformation. Later a round-headed group with high skull vault and deformed occiputs appearedin the area. As is to be expected in peripheral districts, there werelocal adaptations of borrowed elements, developments peculiar tothe region, and some lag in the appearance of certain features.Because of the abrupt change in house types, as well as markeddifferences in some of the minor elements in the material culture,it is thought that there may be a break between the Georgetown andSan Francisco stages. Perhaps there is a period which thus farhas not been recognized or for which evidence has not been found.The true Mogollon pattern is included in the Georgetown and SanFrancisco. The beginning of Anasazi and Hohokam influences arenoted in the Three Circle. These culminated in the Mimbres whenthe pattern became a coalescence of Mogollon-Anasazi-Hohokamtraits. At the end of the Mimbres stage the people deserted theprovince, seemingly moving south into northern Chihuahua where theyfused with other groups. Subsequently another group appeared inthe province. They apparently came in from the north and west andwere fully developed at the time of their arrival. They were Anasaziand this stage is the one called the Animas.Timbers from a number of Mogollon houses have been datedand this makes correlation with the Anasazi easier than with theHohokam. It has been suggested that basic Mogollon, as yet not
^ For detailed information on the Hohokam (the Red-on-BnflE) consult the reports of theGladwins and E. W. Haury.
^Q BUREAU OF AMERICAN ETHNOLOGY [Boll. 121clearly defined, was contemporaneous with Modified Basket Maker,Georgetown and San Francisco with Developmental Pueblo in itsearlier stages and Three Circle with its later phase, Mimbres withGreat Pueblo, and Animas with Kenaissance. On the whole theMogoUon shows closer affinities to the Anasazi than to the Hohokamand numerous Mogollon traits penetrated into some sections of theAnasazi province.^*At the main archeological site south of Allantown three periods inthe growth of Anasazi pattern are represented. There are traces ofModified Basket Maker, Developmental Pueblo remains are nu-merous, and the large ruins belong to the Great Pueblo era. Inview of this association the location is an important one from thestandpoint of information on the transitions between the severalperiods. Fortunately all of the occupants of the site did not buildon the same spot. A series of trenches dug in various places alongthe ridge where many of the remains are located demonstrated thatin the early stages the people lived in straggling settlements of singlefamily houses. As time went on the domiciles were placed in morecompact groups and a certain system of arrangement developed.The first structures were semisubterranean m form. Later this typewas given up and small buildings, with several contiguous rooms,were erected above ground. Ultimately two large communal houses,sheltering many families, two and three stories in height and con-taining more than 100 rooms evolved. These last structures occu-pied only a small portion of the area covered by the older horizons.As a consequence it is possible to obtain representative remains ofthe separate stages without first excavating those of more recentdate. There is definite stratification in a number of places, how-ever, which establishes the proper sequence for the different forms ofstructures and types of associated objects. Complete excavation ofthe site should give valuable information concerning the changeswhich took place during the closing days of Modified Basket Makerand the beginnings of Developmental Pueblo, together with data onthe growth of the cultural pattern through Developmental and GreatPueblo. The data covering the remains to the end of the Develop-mental stage are helpful and interesting. As no work was done inthe Great Pueblo ruins knowledge of their character is meager. Forsome reason as yet undetermined the location was abandoned dur-ing the Great Pueblo stage and never reoccupied.The small ruins scattered along the valley bottom represent dif-ferent periods. Inasmuch as a number belong solely to one singlehorizon they furnish opportunity for valuable checks on the dataobtained at the main site. The unavoidable mixing of objects over
" Detailed discussion of the Mogollon will be found in Haury, 1936, a, b.
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA X7a long period of occupation sometimes produces a certain amount ofconfusion in the evidence and clear-cut material from representativesof a single phase serves to clarify such complications.KEMAINS OF STKUCTUKESActivities during the summer of 1931 and most of 1932 were con-fined in the main to the outer fringes of the main site, to the excava-tion of pit remains and accompanying surface ruins, and the trenchingof nearby trash mounds. Toward the close of the 1932 season atten-tion was turned to a small unit-house of the fully established De-velopmental Pueblo type located 1 mile (1.609 k) up canyon, to theeast, from the main site. This was done for the purpose of singlingout and establishing definitely the characteristic Developmentaltraits in the district so that they could be recognized more quicklyin the material from the main site. The 1933 work consisted of acomplete and thorough investigation of a small Developmental villagesituated on the talus at the foot of the escarpment just below themain ruins. This place was chosen for two reasons. First and fore-most was the desire to give the graduate students as wide an experienceas possible in the several phases of an excavation. Second, and notinconsequential, was the fact that available funds were not sufficientto warrant an attempt to excavate the large ruins on the main site.The work of the two previous years carried the investigations tothe point where that was the next logical step. Complete excavationof the large stone remains would require an extensive budget andrather than start something that could not be continued through to aproper conclusion it was deemed advisable to leave the mounds im-disturbed. At the main site the subterranean portions of 18 pitstructures were cleared of the debris that accumulated in themduring the interval since they were abandoned and fell into ruin,and 2 more were excavated at the nearby places where additionalwork was done. Fire had destroyed several of these structures andthe charred remnants of their superstructure timbers were lying onthe floors in positions clearly indicating the type of roof construction.This evidence, in addition to that furnished by the pits, makes itpossible to reconstruct an accurate picture of the semisubterraneanstructures prevailing during the early stages in the occupation of thesite. The general term of structure is used in preference to that ofdwelling because in one or two examples there is a question as towhether or not the function was ceremonial or domiciliary. One inparticular suggests that it might have served first as a dwelling andthen, after a lapse of time during which it was untenanted, wasremodeled for a religious chamber.
Jg BUREAU OF AMERICAN ETHNOLOGY l^vvu 121The remains probably represent places of habitation in a majorityof cases. The structures were rather crude, but they no doubt servedtheir purpose to a degree highly satisfactory to their builders. Theyhad consisted of a circular, oval, or rectangular excavation roofedover with a pole, brush, bark, and plaster superstructure. The earthwalls of the pits were faced with a coating of plaster made fromadobe mud in which there was a slight admixture of ashes. Two ex-amples varied from the others in that the walls were lined with largestone slabs set up on edge. Resting upon and rising above the topslabs in one were several courses of masonry formed from roughstones. The stone facing of the walls in these cases was covered withplaster and to the casual observer the interior would not have differedin appearance from the other structures. The pits varied in depthfrom 2 feet (60.96 cm) to 6 feet (1.828 m) and from 10 to 20 feet(3.048 to 6.096 m) in diameter.Four upright posts set in the floor close to, or a short distancefrom, the walls supported the superstructure over the pits in all buttwo of the groups of remains. In one there were four posts andone masonry pillar, all placed against the wall, and the otherexception had five posts. The upper ends of the main supports wereforked and sustained cross beams. Tree trunks with suitable crotcheswere secured for the purpose. The latter formed a rectangularframework against which rested the upper ends of small timbers,their butts placed on the ground some distance back from the edgesof the pit. These slanting poles constituted a major part of theroof, in the form of a sloping ceiling. The rectangular space at thecenter was flat. Near the middle was an opening that served bothas smoke hole and entrance. The entire framework was coveredwith reeds or grass, then strips of cedar bark, brush, and leaves.A tliick layer of adobe plaster was spread on top of this and overall was a thin coating of earth. The structure with four posts anda masonry pillar had a different type of covering, a cribbed roof.This superstructure was like those generally found in later-day kivasor ceremonial chambers. The timbers were laid from post to post,each succeeding row above being placed nearer the center of theroom, the logs cutting across the corner of the series below, etc.,until the framework rose into a kind of dome. The area near thecenter, bordering the smoke hole and entrance, was usually flat.The framework was covered with bark, leaves, and plaster. Thetops of the roofs, both types, were only elevated above the groundlevel sufficiently to provide for drainage. In most cases the struc-tures were placed along the top of a ridge or on the side of a hill sothat the run-off of surface water would be facilitated. A villageconsisting of houses of this type would not be striking in appear-ance. All that would be visible would be a series of low, rounding
ROBBRTS] ARCHEOLOGICAL REMAINS IN ARIZONA 19mounds with the ends of ladders projecting through rectangularopenings in their tops.The interior features were usually simple, although an occasionalexample was characterized by a large variety of basins or cachepockets in the floor. Near the center of each chamber, directly underthe opening in the roof, was a fire pit. A second depression, closeto the fire pit and generally at its southeast side, served either asa rest for the lower end of the ladder used in gaining access to theroom, or as a depository for ashes. In one instance entrance had notbeen through the roof and there was no central ladder, althoughone was employed to reach a short passageway leading into thechamber. Various small holes or pockets scattered about in thefloors probably served as depositories for implements of stone andbone or other minor objects. There were some examples of stor-age recesses in the walls. Where these were present they wereon or extended slightly below the floor level; none were placedabove it.An aperture in the wall at the east or southeast side of the roomopened into a small tunnel. The latter led to a vertical shaft whoseoutlet was on the ground level some distance from the edge of theroof mound. This feature has been termed the ventilator because aconstant supply of fresh air was drawn down through the shaft andthe tunnel by the circulation of air in the main chamber. An up-right slab of stone was generally, although not always, set in thefloor at the base of the ladder, between the opening in the wall andthe fire pit. This stone has been termed the deflector, since it pre-vented inrushing air from blowing directly on the fire, and divertedit around the walls of the room. The draft through the ventilatorseems at times to have been so strong that it was necessary to stopthe opening completely and well-worked, oval-shaped slabs of stonewere provided for the purpose. Wlien a number of the pits wereexcavated these cover stones were in position in front of the openings.The structures in general showed a marked similarity to the wide-spread semisubterranean type of house built in many sections of theSouthwest. They are particularly comparable to pit dwellings inthe Chaco Canyon, in northwestern New Mexico,^* and at the oldLong H Ranch in eastern Arizona, about 30 miles (48.280 k) south-west from the present location.^^ There were individual differencesand variations such as may be expected in any group of ruins, butin their main essentials the structures exhibited a marked similarity.One feature brought to light by the excavations was peculiar tothis site. The structures in two different groups were connected.
"Roberts, 1929.
"Roberts, 1931.
2Q BUREAU OF AMERICAN ETHNOLOGY [Bull. 121There were no partitions at the sides where they joined and long,narrow houses resulted. One of the examples had two connectingchambers, the other three. Each of the rooms was complete in itself,but because of the absence of a separating wall, became an integralpart of a larger structure. Another of the groups had three adjacentchambers, although they were not connected. The significance ofgrouping of structures by threes is not known. A comparable situa-tion was found at the Long H Kanch where three pit houses werebuilt close together. Each was distinct in itself, although there wasa small opening between two of them which had a definite purpose,that of a ventilator.i^ Another example of a row of three placed injuxtaposition was reported from southern Colorado by Martin." Inthis case there seemed to have been no opening between any of thechambers. Perhaps the best example of pit dwellings placed in acontiguous row, although not in series of threes, is in Harrington'sPueblo Grande in Nevada. The houses at that site, however, differconsiderably from the pit structures at tliis location.^^The pit remains at the main site were associated with some surfaceruins and all features were investigated as the work progressed.The pits were numbered consecutively, although they occurred intwo main groups. The first cluster was located south of the moundsenclosing the Great Pueblo ruins and consisted of 12 pits, a pit oven,4 brush shelters, and the remains of 2 surface masonry structures(fig. 1). The latter appeared at a later stage than some of the pits,but the oven and several of the brush shelters apparently were partof the older horizon. The second main group was located northof the large stone ruins and included 6 pits, some surface remains,and a dance court (fig. 25).The various trash mounds or refuse heaps were trenched for thepurpose of obtaining stratigraphic evidence on ceramic changes andalso for the purpose of locating burials. They were situated at thesoutheast side of the house groups. Their relation to the latter isshown on the plans.The 2 additional pit remains were located on the floor of thevalley. One was at the site of the fully developed unit-type struc-ture of the late Developmental period and the other was included inthe early Developmental village. To preserve unity in considerationof the pit type of structure, they are numbered as continuations ofthe main sequence and discussed in that order.? Roberts, 1931, fig. 1, pp. 19-43." Martin, 1930, pi. vii, fig. 2, pp. 27-33."Harrington, 1927, p. 264.
robbets] archeological remains in arizona 21Group No. 1struotuee 1The first of the pit remains uncovered consisted of a roughly rec-tangular excavation with rounded corners. The structure seemed tohave been one of the oldest in the first group. This was evidencedboth by the potsherds found on its floor and by the rings in thecharred beams from its burned roof. The floor had a great varietyof pockets and basins in it, but in the main the interior features werequite comparable to the general type of pit house found throughoutdie plateau region. The pit had been dug into the native earth andthe face of the excavation covered with a mud plaster to make thewalls of the room. There was no encircling bench of the type fre-quently found in such cases, but the equivalent of such a feature hadbeen obtained by setting the lower ends of the slanting poles of thesuperstructure back some distance from the top of the wall. Theground level at the time when the house was built was a sandy sur-face which could have kept the sloping timbers from sliding withoutthe necessity of a bench. As a consequence the builders may havedecided that the actual construction of such a feature was unnecessary.The roof was supported by four posts and the charred timbers lyingon the floor demonstrated that the superstructure had been of thegeneral type described in a preceding paragxaph.Near the center of the room was a circular fire pit and adjoining ita pit like those in the majority of houses where the lower end of theladder rested. In this particular example it must have had someother purpose because the ends of the upright pieces of the ladder hadbeen embedded in the floor a few inches from the edge of the secondpit. It is possible that the occupants first made use of the pit andthen discarded it in favor of a second form of ladder base. Therewas nothing to show that this had been the case, however. This pitwas practically filled with wood ashes when the debris was clearedfrom the structure. There were no indications that a fire had burnedin it and the ashes must have been removed from the true pit andplaced there. The secondary pit was possibly a special provision forthe depositing of ashes. Several of the houses at the site had similarfeatures and it may be that in some of the structures in this localitythere was a definite shift in function from ladder box to ash pit as acorrelative of embedding the lower ends of the ladder posts in thefloor.The ladder provided for gaining access to this structure was un-doubtedly of the two-pole-and-rung type. The lower ends of the sidetimbers were still in place in the floor between the basin, that wouldnormally be considered the ladder rest, and the deflector stone. The68764?39 3
22 BUREAU OF AMERICAN ETHNOLOGY [Buli.. 121remains of the posts sloped at just the angle which the ladder wouldhave required in order to pass over the fire pit to the smoke hole andentrance in the roof above. Indications in early forms of houses inthe Southwest are that the runged ladder was employed practically
IS7 reerTO GREAT PUEBLOKUINS
A^5OSTRUCT3 N0.4URt ty)
"^ RETUSE ANDBURJAL MOUNO
STRUCTURES
_ , NO. 5.FOl NO. 5k
-o
1V. REFUSE ANDBURIAL MOUNDFiGUEE 1.?Group 1 pit and surface remains. Numbered dots indicate location of burials.from the beginning of the use of pit structures. Traces of their pres-ence were found in pit dwellings at the Long H Ranch " and a num-ber of the structures excavated in 1931 and 1932, in addition to No. 1,gave ummstakable evidence of their use. It is not possible, of course,? Roberts, 1931, pp. 33-34, 56.
ROBEBTS] ARCHEOLOGICAL REMAINS IN ARIZONA 23to determine just liow these ladders were made. Examples foundin Pueblo ruins of later horizons have consisted of two upright poleswith the rungs lashed to them. In other cases the long timbers werenotched and the rimgs fastened in the notches with cords made fromhide or yucca fibers. Somewhat similar methods of construction wereno doubt used in the ladders in earlier times but until one is foundintact their exact nature will not be laiown.The large number and variety of holes or pockets in the floor ofNo. 1 are indicated on the plan (fig. 2). It is not possible to deter-mine what all were used for, although most of them no doubt servedas depositories for small objects employed in household industries.Bone and stone implements were recovered from a number of them.Objects of this type would easily be misplaced unless special effortswere made to provide them with a container. A simple solution ofthis problem would be to make a pocket in the floor or wall. Forsome reason or other the latter seems rarely to have had such a featureat this site. The floor furnished all the space needed for suchpurposes.The larger basin-like depressions (fig. 2, c, d, w) possibly servedas places in which baskets or jars were set, although there was noindication that such had been the case. What function the curiouslyshaped depression e may have been intended to fulfill is a questionfor which no answer has been found. At the time when the remainswere excavated a small pitcher of the culinary variety of pottery wassitting in the end toward the ventilator side of the room. Severalwho saw the house suggested that loom sticks may have been placedacross the depression in the arm-like projections at each side of thecentral portion. In view of weaving practices in general, as theyare known throughout the Southwest, this does not seem to be a goodexplanation. There was nothing about the depression itself to furnisha clue.The hole marked A, figure 2, corresponds in its general size and posi-tion to similar ones found in most pit houses and in many present-daykivas. In the kivas or ceremonial chambers the feature is called thesipapu, because the modern Pueblos use that name or a close varia-tion of it to designate the hole. According to them it symbolizes themythical place of emergence through which their ancestors are sup-posed to have passed in their journey from the inner portions of theearth, where they were created, to the surface of the world upon whichthey now live. Whether a similar interpretation was made in thepast or not is a matter of conjecture. The belief is so firmly implantedin Pueblo mythology, however, that it may with considerable justifica-tion be assumed that the prehistoric people regarded the feature inthe same light. A further support for this theory is derived from
24 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 121the clear evidence that the later kivas represent definite survivals ofthe old subterranean house.At the base of the ladder an upright stone was set in the floor mid-way between the fire pit and the ventilator opening in the wall (fig.
FiGUBE 2.?structure 1. a, holes for support posts ; 6, c, and d, storage basins ; e, basinin floor ; f, wall forming compartment ; g, basin in floor ; h, sipapu ; i, fire pit ; /, ash pit
;
K ends of ladder poles; l, deflector; m, stone to protect ventilator opening; n, pole place-ment; 0, small shaft in ventilator; p, main ventilator shaft; q, pot rest or storage hole;level "' ^' ^' '^' ^""^ ^' ^*?''''^? ^?^^^ ^?*^ ^^^'^^' "' ^^?<^ accumulation above old ground2, 1) . This is an example of the deflector. It was so placed to preventair, coming through the ventilator aperture, from blowing directlyon the fire.One interesting feature in this structure consisted of a row of
EOBEBTS] ARCHEOLOGICAL REMAINS IN ARIZONA 25seven stones forming a bin-like enclosure at the southeast corner ofthe room (fig. 2, // pi. 3, a). One of the typical interior elementsin Modified Basket Maker houses consists of a low wall of stone slabsseparating the ventilator or, as was the case in many of the dwellingsof that period, the entrance side of the chamber from the rest of theroom. Sometimes the partition wall extended entirely across fromsupport post to support post. Then again it was broken by a doorwayin front of the passage opening into the wall. Occasionally a houseis found that only has corner bins at that side of the room.^*^ Theenclosure in one corner of No. 1 is very suggestive of the more com-plete examples found in the Chaco Canyon houses and possibly repre-sents a survival in the houses of a later day of a characteristic com-mon during a previous stage in house development. The purposeof such compartments is not known. Two metates or grinding stoneswere found in the present example. This would suggest that theenclosure may have served as a mealing bin for the house. On theother hand it is quite possible that the stones were placed there duringintervals when they were not in use, since one was leaning against thepartition and the other against the wall. The bin may have servedsolely in the capacity of a storage place. Firewood, jars containingsurplus food and other items connected with the daily life of the peoplecould have been placed there where they w^ould not interfere withthe general trend of affairs in the main part of the structure. Some-what comparable enclosures in the earth lodges of some of the PlainsIndians, dwellings which in many ways bear a strong resemblance tothe pit houses of the Southwest, despite the fact that they werelargely surface structures, were employed for storing the supply offirewood in winter.^^In a previous discussion of the partitioned-off portion of the pithouse, the theory was advanced that the compartment may have hadsome ceremonial significance in addition to its utilitarian purposes.Persons not taking part in such religious observances as were per-formed in the main portion of the room may have gathered in thecompartment where they could witness but would not intrude uponthe rites. An analogous situation is described in a report on a councilheld in an earth lodge at the mouth of the Platte Eiver in 1833. Thewriter, J. T. Irving, observed that the passage was completely crowdedwith women and children watching the progress of events within thestructure.^^ When it is recalled that the ventilator unquestionablywas derived from and represents a modified survival of an olderpassageway entrance, the parallelism becomes more apparent. In the
?> Roberts, 1929, flgs. 4, 5, 21, 22 ; pis. 4, 6, 7 a.?Bushnell, 1922, flg. 8. Coues, 1897, vol. 1, p. 337. Maximllien, 1843, vol. 3, fig. 4,facing p. 36.? Bushnell, 1922, p. 117. Irving, J. T., 1835, vol. 1, p. 234.
2Q BUREAU OF AMERICAN ETHNOLOGY tBoi^- l^isame connection it has been suggested that the so-called spectators'bench in the regular kivas of the Hopi Indians may represent a modi-fied survival of the compartment.^^ It should be pointed out, however,that there is no chronological hiatus between the modern and earlyforms, since many of the circular kivas of the Developmental and GreatPueblo periods have deep recesses or niches above the ventilatorwhich apparently fit into the compartment-spectators' bench sequence.There is no reason why the pit-dwelling compartments could not havefunctioned in all of the ways outlined. They undoubtedly hadutilitarian value in the beginning, or they would not have been in-corporated in the structure. In the later ceremonial chambers themodified survival may well have served only a ritualistic function.The ventilator at the southeast side of the chamber consisted ofa passage and shaft. The horizontal portion was dug through theearth as a tunnel. In some cases it was the custom to run a trenchfor the passage and then cover it over with poles, brush, and earth.In this particular instance the more difficult method was followed.The aperture from the room into the tunnel was oval in shape andwas finished with a coating of plaster. Around the edges of theopening was an offset in wliich a cover slab could be placed. Thiswas in position when the remains were excavated. The sill of theopening was some distance above the floor level and consisted of athin slab of stone. The ventilator passage was practically hori-zontal ; there was no upward slant at the outer end as is frequentlyfound in structures of this tyi^e. The shaft was roughly circular in.contour and was faced with large stone slabs. The ventilator forNo. 1 was not large enough to have functioned as an actual entrance,hence it must have served purely as a means for bringing fresh airinto the chamber.The ventilator had one feature not commonly found. This con-sisted of a small vertical shaft which rose from the ceiling of thepassage, just back of the main aperture into the chamber, and openedin the top of the pseudo bench just in front of the sloping super-structure poles (fig. 2, o). Just what reason motivated the con-struction of this shaft is not clear. It may have functioned forventilating purposes by distributing incoming air at a higher levelthan the ordinary opening in the wall did. Experiments with thetwo openings showed that one or the other had to be closed beforethe air would circulate properly. The aperture in the wall had acover stone for that purpose and there was a thin slab over the topof the shaft. With one of these in place and the other removedthe air passed through the uncovered opening in an appreciabledraft. When both openmgs were left uncovered no movement of air
?? Roberts, 1929, p. 89.
ROBBBTS] ARCHEOLOGICAL REMAINS IN ARIZONA 27could be felt. Since this was the only structure in the entire groupwhich had such a feature it is possible that some individual pecu-liarity in the house necessitated the arrangement. No other reasoncan be advanced for its presence in No. 1 when similar constructionwas absent in the other structures.The structural details of No. 1 were so clearly shown that thesuperstructure was restored in order that visitors to the site mightsee just what domiciles of that type were like. In doing this it wasnecessary to use new timbers because the original posts, beams, andpoles were too greatly damaged by the fire which destroyed the struc-ture to permit their reuse. Every effort was made to follow the plansof the first builders, although in two factors it was necessary todeviate from them. One of these was in the matter of the use ofreeds on top of the slanting wall poles. At the present time thereis no source for such material in the general vicinity. Where the pre-historic inhabitants secured their supply is not known. It is possiblethat the occasional small ponds and marshy places in the valleybottom, referred to in the introduction, furnished places where reedsgrew. The second variation consisted in the placing of two heavybeams in such a position that they rested on the earth at each sideof the pit rather than on the support posts. This was done as aprecautionary measure so that there would be no likelihood of theroof caving in when a number of visitors stood on it or when domesticanimals belonging to the nearby Navajo chanced to cross it, factorswhich the early people did not have to consider.The preliminary framework of upright posts and beams for theflat portion of the roof is shown in plate 3, h. The slopingside poles of the superstructure were then put in place and coveredwith brush and cedar bark as illustrated by plate 4, a. The hole wasthen covered with earth and took on the appearance of a low, flatmound (pi. 4, &). A view of one corner of the completed in-terior is illustrated by plate 5, a. The weathering qualities of such aroof were well tested by the winter of 1931-32 when unusually heavyrains and snows fell throughout the district. For a period of sev-eral weeks more than 3 feet (91.44 cm) of snow covered the house.Only a small amount of moisture penetrated to the interior duringthe melting period and a slight amount of repair was all that wasnecessary in the late spring of 1932. Had the structure been occu-pied and the usual amount of upkeep provided by the tenants, itwould have come through the winter with practically no damage.Wlien the debris of accumulation was removed from the interiorof No. 1 a 3-foot (91.44 cm) square pillar was left near the center ofthe pit (pi. 6, h). This was done wherever practicable in orderthat a careful study might be made of the fill and that a stratigraphic
28 BUREAU OF AMERICAN ETHNOLOGY [Buii. 121test could be made of the potsherd material contained within suchsections. Considerable information of interest was obtained in thisway. The pillar for No. 1 was 6 feet 6 inches (1.981 m) from top tobottom. In making the study of the material which it containedthe earth was removed stratum by stratum and put through a screen.The top 6 inches (15.24 cm) consisted of a recent accumulation ofsand with a slight admixture of ashes (fig. 3, a). There was not agreat amount of material, such as chips of stone and bone, in thislayer and there were no potsherds. The contents indicated a winddeposit which probably was subsequent to the Indian occupation.
A'
BUREAU OF AMERICAN ETHNOLOGY BULLETIN 121 PLATE 3
a. Interior of struc-tim
b. ^Tain framework for restoration of structure 1.
BUREAU OF AMERICAN ETHNOLOGY BULLETIN 121 PLATE 4
0. Brush-and-bark covering over framework, restoration of structure 1.
Coiupleied structure shovvini; rounding dome and smoke hole-hatclnv.ay entrance.
BUREAU OF AMERICAN ETHNOLOGY BULLETIN 121 PLATE 5
a. Interior view of portion of restored structure.
b. Pillar left in excavating pit, structure 1.
BUREAU OF AMERICAN ETHNOLOGY BULLETIN 121 PLATES
(/. Miisdurv wall restiiii; on slali fnuiHlalinii
fi. Pilaster in structure 2.
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA 29The tliird layer (fig. 3, c) was 8 inches (20.32 cm) in depth andconsisted of sand, ash, large fragments of charcoal, rocks, and brokenstone implements. This stratum unquestionably recorded the deposi-tion of debris and waste material from a habitation, since there was alarge showing of potsherds and broken animal bones.The layer immediately below (fig. 3, d) was also 8 inches (20.32cm) tliick. It was separated from the one above by a fine streak ofwater-washed sand. Although insignificant in itself this mere lineof deposit was of importance because it marked the old ground level.Everything above this line, both in the fill and beyond the bordersof the house, represents an accumulation subsequent to the occupa-tion level when the pit was dug. The relation of this accumulationto the ground level around the pit is illustrated by the section throughthe house (fig. 2, z). The layer just below the sand streak consistedof ashes, charcoal, sand, broken bones, stone chips, and potsherds,typical refuse-mound content.Next below this deposit in the pillar (fig. 3, e) was a 2-inch (5.08cm) streak of water-washed sand. It contained no objects of stone,bone, shattered pottery, or charcoal. The deposit recorded a heavyamount of precipitation with the attendant washing of material intoa depression.The sixth layer from the top of the pillar ranged from 6 inches(15.24 cm) to 1 foot 3 inches (38.1 cm) in thickness (fig. 3, /). Thecontent consisted of sand and charcoal, which again suggested a typ-ical refuse deposit. There were large numbers of stone chips, spalls,broken bones, and potsherds. The next two layers were sterile asfar as any traces of human occupation were concerned and havebeen combined in the diagram (fig. 3, g). The first of the two was1 inch (2.54 cm) thick and was composed entirely of water-washedsand. This again indicates either a heavy rain or a deep fall ofsnow. The melting of the latter could easily carry such materialinto a pit. The 4 inches (10.16 cm) immediately below consisted ofclean wind-blown sand. The 5 inches (12.7 cm) of these combinedstrata evidence a comparatively long interval in which there waslittle activity around this portion of the site. It is quite possiblethat it may even have been abandoned for a time.The next layer was the bottom of the pillar (fig. 3, A). It meas-ured 2 feet 1 inch (63.5 cm) in thickness and comprised burnedroofing timbers, plaster from the roof, broken stones, ashes, and thegeneral debris which follows a conflagration. The greatest numberof potsherds in the entire pillar came from this level. Some of themwere on the floor, but the majority were high enough in the fill tosuggest that their presence was due to the use of the pit as a dumpimmediately following the fire.
oQ BUREAU OF AMERICAN ETHNOLOGY [Bull. 121The general significance of the fill in pit No. 1 may be summarizedbriefly as follows: The structure was destroyed by fire and im-mediately thereafter was used as a dumping place for refuse fromother nearby habitations. Then for some unknown reason the sitewas abandoned for an interval sufficiently long to permit an ac-cumulation of 4 inches (10.16 cm) of clean sand. On top of thislayer an additional inch (2.54 cm) of sand was deposited by water.People again began to move about and dwell in the immediate vicin-ity and evidence of their presence is to be found in the charcoal,ashes, and potsherds deposited there. This record is broken by theevidence of another interval of marked precipitation when only cleansand was washed into the pit. Following this there was a con-siderable period when refuse was deposited in the pit. The ac-cumulation settled and became compressed into a deposit 8 inches(20.32 cm) thick and the pit was completely filled. A thin sheetof sand was then swept across the top by water, possibly the run-offfrom a heavy rain. Everything above this line probably representsdeposits that accumulated through a combination of factors. Ma-terial no doubt drifted down from the ridge above and mixed withthat collecting as a result of human activities. Ultimately the entiresite was abandoned and the topmost layer resulted from action ofthe elements, wind and surface water depositing material at the spotwhere the house once stood. The significance of the potsherds fromthis section will be considered in detail in the second part of thisreport under the subject of pottery.Despite the large amount of charred roofing material obtainedfrom the floor of No. 1, it has not been possible to determine thecutting date of the timbers by the Douglass method.^* A number ofbeam fragments from this house gave a good series of rings, butwhen this report was written they could not be fitted into the majorchronological chart. The work on the beam material was done byMr. Carl F. Miller under the supervision of Dr. Douglass.The diameter of the pit portion of No. 1 was 13 feet 5 inches(4.089 m) on the sipapu, fire pit, ventilator line. At right angles tothis measurement it was 13 feet 8I/2 inches (4.177 m) from wall towall. There was no true bench to the structure but the poles form-ing the sloping part of the superstructure were set back from the edgeof the wall an average of 1 foot 6 inches (45.72 cm), forming to allintents and purposes what might be called a bench. The depth ofthe pit varied somewhat from side to side. The wall opposite theventilator was 4 feet II/2 inches (1.257 m) above the floor. At theventilator side of the chamber the top of the wall was only 3 feet8I/2 inches (1.130 m) above the floor. This difference is due to the
^ Douglass, 1932, 1935. The method of obtaining dates from the growth rings in thebeams found in prehistoric houses is described by Dr. Dous;lass in these articles.
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA 3]^fact that No. 1 was placed on the side of a hill so that there was acertain amount of slope in the top of the wall.The holes for the support posts varied somewhat in size and intheir distances from the wall. That at the north corner of thechamber was oval in shape with diameters of 11 inches (27.94 cm)and 9 inches (22.86 cm). It was 9 inches (22.86 cm) from the wallat one side and 1 foot 3i^ inches (39.37 cm) at the other. Its depthwas 2 feet (60.96 cm). The hole at the east corner had diametersof 9 inches (22.86 cm) and 10% inches (26.67 cm). It was 4i/^ inches(11.43 cm) from one wall and lli/^ inches (29.21 cm) from the other.It measured 2 feet 2 inches (66.04 cm) in depth. The hole at thesouth corner of the room was irregular in outline because of thefact that it had been formed by the use of two supports. It is likelythat the original post placed at that corner may have needed rein-forcing, so a second one was placed there. Due to the irregularity,it is not possible to give the exact measurements of each segment.It will suffice to give the total length and greatest breadth. Acrossthe long diameter it measured 1 foot 3I/2 inches (39.37 cm). At itswidest section it measured 9 inches (22.86 cm). The wall at thispart of the room had more of a curve than for the two previouscorners, so that the edges of the hole were about equidistant fromit. The average distance was 7 inches (17.78 cm). One segment ofthe hole had a depth of 1 foot 10 inches (55.88 cm) and the other1 foot 7 inches (48.26 cm). The west post placement was morenearly circular than the others with diameters of 8^/2 inches (21.59cm) and 9 inches (22.86 cm). It was 1 foot (30.48 cm) from thewall and had a depth of 2 feet (60.96 cm).The fire pit was roughly circular in form with diameters of 2feet (60.96 cm) and 2 feet 1% inches (64.77 cm). Its depth aver-aged 10 inches (25.4 cm). The oval depression which functioned asa depository for ashes was 2% inches (6.35 cm) from the fire pit.The depression had diameters of 1 foot 3 inches (38.1 cm) and 1foot 10y2 inches (57.15 cm). Its average depth was 7 inches (17.78cm).The deflector stone stood 7 inches (17.78 cm) from the oval de-pression adjacent to the fire pit. The stone was 1 foot 3% inches(39.37 cm) long and ranged in thickness from 3 inches (7.62 cm) to2 inches (5.08 cm). It stood 1 foot 3 inches (38.1 cm) above thefloor. The second stone placed upright in the floor 3l^ inches (8.89cm) from the ventilator opening suggested a secondary deflector.It was not high enough, however, to have functioned in such acapacity and probably was placed there to protect the sill of theopening. This stone had a length of 1 foot 1 inch (33.02 cm) anda thickness of II/2 inches (3.81 cm). It rose 5 inches (12.7 cm) abovethe floor. At each end of this stone was a small hole in the floor
Qo BUEEAU OF AMEKICAN ETHNOLOGY [Bull. 121which had contained a post, possibly so placed to aid in closing theventilator opening. These holes were 1 and 2 inches (2.54 and 5.08cm) from the ends of the stone. Their diameters ranged from 21/2to 4 inches (6.35 to 10.16 cm). They were 4 inches (10.16 cm) deep.The sipapu was 3 feet 6 inches (1.066 m) from the fire pit and 3feet 51/2 inches (1.054 m) from the wall at the side opposite the ven-tilator." The sipapu had diameters of 31/2 and 4 inches (8.89 and10.16 cm). The hole was not cylindrical but jug-shaped (fig. 2, h).As a consequence the diameter below the floor level was 8l^ inches(21.59 cm). The depth of the sipapu was 6 inches (15.24 cm).The various holes and pockets in the floor had a considerable rangein size and depth. The hole h, figure 2, had diameters of 1 foot 1/2inch (31.75 cm) and 9 inches (22.86 cm). The average depth was 10inches (25.4 cm). The odd-shaped depression e had two parts. Thelarger was 2 feet 11/2 inches (64.77 cm) long and 1 foot 3 inches(38.1 cm) wide. The second had a length of 1 foot 9 inches (53.34cm) and a width of 1 foot 1 inch (33.02 cm). The average depthof the entire depression was 4 inches (10.16 cm). The basin d, figure2, had a length of 1 foot 8 inches (50.8 cm) and a width of 1 foot214 inches (36.19 cm).The curiously shaped pit with arms (fig. 2, e) had a length of 3feet 7 inches (1.092 m) and a width of 1 foot 5i^ inches (44.45 cm).The arms ranged in length from 5 inches (12.7 cm) to 10 inches(25.4 cm). The average depth of the main portion was 10 inches(25.4 cm) and of the arms 2 inches (5.08 cm).The five-sided depression in the floor within the compartmentformed by the row of upright slabs set between the east wall andthe ladder had a width of 9 inches (22.86 cm) and a length of 1foot 1/2 inch (31.75 cm). Its depth was 3 inches (7.62 cm). Theholes where the ladder posts were placed had diameters of 3 to4 inches (7.62 to 10.16 cm) and 314 inches to 4 inches (8.89 to 10.16cm). One had a depth of 3 inches (7.62 cm) and the other 4 inches(10.16 cm). The holes at either end of the stone in front of theventilator opening, n, figure 2, had diameters of 2I/2 inches to 4l^inches (6.35 to 10.79 cm). Their depths were 3 inches (7.62 cm)and 21/^ inches (6.35 cm)
.
The oval-shaped hole q, figure 2, had diameters of IO1/2 inches(26.67 cm) and 61/2 inches (16.51 cm). Its average depth was 31/2inches (8.89 cm). The hole r had diameters of 6 and 7 inches (15.24and 17.78 cm) and a depth of 5 inches (12.7 cm). The one imme-diately next to it, s, measured 9 inches (22.86 cm) by 81/2 inches (21.59cm) on two diameters and was 10 inches (25.4 cm) deep. The irreg-ular holet measured 51/2 inches (13.97 cm) and 6I/2 inches (16.51 cm)on two diameters. It was rather shallow, having a depth of but 3inches (7.62 cm) . The hole marked u on the diagram was covered
KOBEETs] ARCHEOLOGICAL REMAINS IN ARIZONA 33with a small, thin stone slab when found. The cover was squarewith rounded corners and had been carefully worked. The hole waspractically circular with diameters of 3^ and 4 inches (8.89 and10.16 cm). Its depth was 5 inches (12.7 cm).The hole v had diameters of 3I/2 and 4 inches (8.89 and 10.16 cm)
.
The depth was 6 inches (15.24 cm). The oval depression w near thesouth support post had a long measurement of 1 foot 10l^ inches(57.15 cm) and a short one of 1 foot 2 inches (35.56 cm). Theaverage depth was 4i/^ inches (11.43 cm). The hole x had diametersof 5 and 7 inches (12.7 and 17.78 cm). Its depth was 2 feet 2 inches(66.04 cm). The oval hole y measured 5i/4 inches (13.34 cm) on itslong diameter and 2% inches (6.98 cm) across the short way. Itsdepth was 4 inches (10.16 cm).The enclosure formed by the row of stones / measured 5 feet 1 inch(1.549 m) by 3 feet (91.44 cm). The stones used to form the com-partment varied in size from one measuring 5 inches (12.7 cm) longto one which was 1 foot 1 inch (33.02 cm) in length. Their averagethickness, from which there was only a slight variation, was I14inches (3.81 cm). There was also a difference in height. The tall-est was 1 foot 11 inches (58.42 cm) and the shortest 11 inches(27.94 cm). At the time when the house was in use, however, suchdiscrepancies in height would not have been noticeable as a heavycoating of plaster extended along the tops of the stones and madea level rim.The Oldening into the ventilator measured 1 foot 2 inches (35.56 cm)wide at the sill and 1 foot (30.48 cm) at the top. The opening was1 foot 3 inches (38.1 cm) high. The sill was 3 inches (7.62 cm)above the floor. The oifset into which the cover slab fitted had anaverage width of 3 inches (7.62 cm) and a depth of 2 inches(5.08 cm). The small shaft rising from the ceiling of the tunnel andopening just inside the sloping roof poles was 8i/^ inches (21.59 cm)from the aperture in the wall. This shaft had a diameter of 10inches (25.4 cm). The tunnel of the ventilator from the opening inthe room to the main shaft at its outer end measured 6 feet 2i/^ inches(1.892 m). The tunnel had a width of 1 foot 1 inch (33.02 cm)where it opened into the shaft and was 1 foot 3 inches (38.1 cm) high.The shaft was slightly oval in contour with diameters of 1 foot 91/2inches (54.61 cm) and 1 foot 7 inches (48.26 cm). Its depth belowthe ground level at the time of occupation was 3 feet 3 inches (99.06cm). It is 7 feet 8 inches (2.336 m) below the present ground level.STKUCTUBE 2Structure No. 2 was one of the few in the entire group that had apit lined with large stone slabs. The use of slabs to form a facing
3^ BUREAU OF AMERICAN ETHNOLOGY [Boll. 121for the walls of the excavated portion of pit houses was a widespreadpractice during the Modified Basket Maker and early DevelopmentalPueblo periods. In some sections, however, the builders were contentto cover the earth walls with a heavy coating of adobe plaster. Thistreatment would have been entirely satisfactory where the qualityof the ground was such that the sides of a pit would hold their form.Where there was considerable sand and the likelihood of crumbling,some reinforcement was required and this generally took the form oflarge stones. In some districts material that could be used in thismanner was not available and it was necessary to place a wainscotingof poles around the walls before the plaster was applied, or else todepend entirely upon the retaining capacity of the plaster itself. Justwhy there was so little use of slabs in this group of pit structuresis not known. The low cliffs bordering Wliitewater Valley a fewhundred feet from the house remains would supply a limitless num-ber of such stones. Furthermore, the soil has a heavy sand contentand walls of pits are prone to slump unless held in place by someother material. The plaster used on the walls of most of the houseswas unusually thick and it is quite possible that the builders foundthat it was sufficient.The pit portion of No. 2 was roughly oval in form. One wall wasalmost straight but the others had a slight curve to them and thecorners were rounded. The slabs used to line the pit varied in height.Some were low while others extended to the ground level. To com-pensate for the discrepancy, the spaces above the shorter slabs werefilled in with a rough form of masonry (pi. 6, a). The use of hori-zontal masonry in this manner had a fairly early inception in theSouthwest. Guernsey found examples in some of his Modified Bas-ket Maker houses,^^ and a number of those in the Chaco Canyonexhibited the feature.^^ It is quite possible that the idea of buildingstructures of coursed stones, which became one of the outstandingcharacteristics of Pueblo architecture, had its beginning in this way.With improved methods of laying stone and the development of themason's technique it became progressively a simple matter to erectstructures entirely above ground and eventually to increase the num-ber of stories in the buildings.Structure 2 was of unusual interest because it gave evidence of twodifferent occupations separated by an interval of some length of time.There were two distinct floors with a fill 1 foot (30.48 cm) in depthlying between them. This fill for the most part suggested materialrequiring an appreciable period to accumulate. Lying on the origi-nal floor was a deposit of charcoal several inches deep. This layer? Guernsey, 1931, pi. 8, lower; p. 11.? Roberts, 1929, pi. 2, a; p. 16.
ROBBBTS] ARCHEOLOGICAL REMAINS IN ARIZONA 35also included ashes, potsherds, broken stones, bones, and other refuse.Above this the fill consisted of clean sand. The second floor was laidon top of this material. The upper surface of the sand layer had notbeen level, as was shown by the presence at several places of lensesof refuse representing intentional fills of dump-heap rubbish soplaced that a smooth, flat base was provided for the plaster. As apart of the work of remodeling and to compensate for the lesseneddepth of the pit resulting from the new floor level, it was necessary toincrease the height of the wall. This was accomplished by addingseveral courses of rough masonry to the top of the original wall, asis shown in plate 6, a, beginning at the 4-feet 2-inch mark on themeasuring rod. The more recent courses did not exhibit as carefulworkmanship as the original ones.The floor plans were somewhat different. The upper was moreelaborate than the lower. The two levels will be discussed separately.In excavating the house remains the second floor, or in other words,the more recent, was first uncovered. The various features of thislevel are shown by the diagram (fig. 4). The superstructure hadbeen supported by four upright posts as shown by the holes in thefloor (fig. 4, a). The position of these holes suggests that the roofv^'as of the truncated type with flat center and sloping sides.Associated with the support at the south corner of the room was afeature of marked significance insofar as it offers a plausible expla-nation for the development of a structural element common to thekiva or ceremonial chamber of later cultural horizons, namely, thepilaster or stone pillar used to support the roof. A pillar of stoneand adobe mud had been placed between the main support post andthe wall at that side of the chamber, presumably as a brace for thetimber (fig. 4, ^/ pi. 6, &). This was possibly necessitated by the roofshifting out of alinement at that side of the room. The stones andplaster were fitted around the pole and its impression was still to beseen in the face of the pillar. It would be an easy step from a pieceof construction of this kind to one slightly more elaborate in whichthe masonry could support the main roof beams and the uprighttimber be omitted entirely.The evolution of the pilaster, on the basis of the foregoing evidence,might be postulated as follows: In the earliest of the semisubterra-nean structures the main support posts for the roof were set in thefloor some distance from the walls of the chambers. As time wenton, the posts were placed almost, but not quite, against the wall. Atendency for the weight of the superstructure to push one of theposts out of alinement, toward the wall, led to reinforcing it withblocks of stone and mud plaster. From the first attempts to con-struct such a brace a fairly good pillar was evolved and then it wasdiscovered that the pillar itself was fully capable of supporting the
36 BUREAU OF AMERICAN ETHNOLOGY [BCLL. 121roof and that the wooden upright posts were not necessary. Whencircular dwellings of this type were no longer occupied as habita-tions but were present in the village only in the form of a survival
5FT
FiGUBB 4.?Structure 2. a, holes for superstructure posts; ft, stone embedded in floor;c, compartment walls ; d, stone posts ; e, subfloor slabs reinforcing fire pit ; f, fire pit
;
g, ladder box and ash pit ; h, storage hole ; i, pilaster behind support post ; /, ventilatoropening; k, ventUator shaft; m, fill over ventilator passage; n, plaster used to reducepassage size.in the ceremonial chamber, the roof supports were made of stone.Had No. 2 been the only structure in which this feature was presentit might be attributed to an individual effort to remedy a weakness
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA 37in one particular pit house. Another one, however, gave evidence ofa similar piece of construction and in addition had a jpHaster withno upright timber in front. This will be considered in greater de-tail in the discussion of that structure. No. 3, and further commentis not necessary in connection with No. 2.The fire pit in the upper floor was roughly oval in shape and itslining was mainly of plaster. A small section of the periphery atthe north side had a stone slab placed in it and there was one smallstone along the western arc, but the remainder was entirely of plas-ter. The pit was reinforced, or it might be said was built up, bya series of slabs set in the fill wliich rested upon the original floor.The tops of most of these stones were covered with plaster and werenot in evidence until the work of removing the upper floor wasstarted. The stones are indicated in the drawing, figure 4, e. Thefire pit on tliis floor level had not been used throughout the entireoccupation of the chamber. It had been allowed to fill with ashesand charcoal. This accumulation was then overlaid with a sheet ofclean sand upon wliich two small stones rested. The entire fill wasthen hidden by a coating of plaster. The fires lighted in the struc-ture during the last days in which it was inliabited were laid in therectangular box that is generally considered as the ladder pit ordepository for ashes. The secondary pit in tliis instance was facedon three sides with large stone slabs. The fourth, that toward theventilator, was finished with plaster. When the debris was cleanedfrom this structure the rectangular box was partially filled withwood ashes containing some bits of charcoal. The stone and plastersides showed clearly the marks of burning from the fires that hadbeen kindled there at various times.This floor was not broken by many holes or cache pockets. Therewas no sipapu and only one hole that might have functioned forstorage purposes. This single example was located near the southcorner of the chamber (fig. 4, K).At the ventilator side of the room there was a suggestion of acompartment such as those described in the discussion of No. 1. Theinclosure in this case, however, was not formed through the erectionof a wall of upright slabs, but consisted of horizontally laid blocksof stone set in large quantities of adobe mortar. The plaster por-tion of the partition was not well preserved and in places largechunks had been broken out (pi. 7, a). The relation of these lowwalls to the support posts and fire and ladder pits is illustrated bythe diagram, figure 4, c, and needs no further description. The wallsthemselves rested upon the floor and showed clearly that they wereerected after it had been laid down.68764?39 1
2g BUREAU OF AMERICAN ETHNOLOGY [Boll. 121Close to the low walls forming the compartment were two rec-tangular-shaped blocks of stone placed in upright positions about mid-way of each partition. The stones were set in the floor in the mainpart of the chamber a few inches from the inclosure (fig. 4, d; pi. 7,a). Wliat their purpose may have been is not known. Such fea-tures were rare in structures of this type previously excavated andhow extensively they occurred cannot be determined until more re-mains have been uncovered. The suggestion that two pieces ofwood were used in similar positions was noted in No. 9, discussed inlater pages of this report. None of the other structures, however,seem to have had them.At the east side of the pit a large irregular slab of stone was placedflat in the floor. Its upper surface was flush with the floor leveland it constituted the only piece of paving in the entire chamber (fig.4, h; pi. 7, a). There was nothing to indicate what its significancemay have been. A similar stone was embedded in the original floorin practically the same position and there seems little question butwhat they had a definite purpose. Other struct\ires of this generaltype at various locations have also shown comparable stones in thefloor. Two suggestions may be offered as to their function. Onespecifically concerns the ceremonial side of the people's life; theother, based on some evidence, concerns a more prosaic utilitarianphase. The suggestion from the ceremonial standpoint is that whenthe chamber was employed for religious observances a drum or basketfunctioning as a drum may have been placed there. Some of thekivas in the modern pueblos have such a feature. The other explana-tion is that a milling stone or metate may have rested upon the flatslab. The stone was incorporated in the floor to prevent such wearand tear as a heavy object, like a metate, would cause. In one struc-ture located elsewhere " and in Nos. 5a and 15 at tliis site, grindingstones were found on the floor in the same relative positions as thisslab occupied. In one case the metate rested upon such a stone, butthe latter was not embedded in the floor.During the remodeling process in No. 2 the ventilator opening wasreduced in size and the passage made correspondingly smaller. Themethod employed in restricting the aperture was that of placing aframe of stones inside the original opening. Two blocks were setupright along the sides. A new lintel was placed on them and a sillwas set in at the bottom. All were covered with plaster to completethe frame of the opening. A new floor level for the aperture andpassage was made necessary by the fill above the first level of occupa-tion. This is shown by the drawing of a section through the chamberand ventilator in figure 4. The reduction in the size of the passage2^ Roberts, 1931, p. 47.
ROBBETSl ARCHEOLOGICAL REMAINS IN ARIZONA 39was accomplished through the use of thick layers of plaster. Therelation of the remodeled passage to the original one is also illustratedby the drawing. Figure 4, w, shows the extent to which it was foundnecessary to apply the adobe plaster.Very little change was required as far as the shaft at the outer endwas concerned. The only alteration made in it was the addition ofa row of stones around the northern arc. This was essential, as sandhad drifted over the old ground level to such an extent that somemeans for preventing its falling into the shaft had to be provided.This was accomplished by setting additional stones on top of theslabs that had lined the original shaft. The passage for the ventilatorwas first constructed by means of the trench-and-cover method. Itwas not tunneled as in the case of No. 1. To make it, a deep trenchwas dug from the side of the house to the shaft. The excavation wascovered with poles and slabs of stone and brush, and the space to theground level was filled in with rubbish from a refuse mound.The original floor in No. 2 w^as similar in most respects to the onejust described. There were some differences, however, as the plan(fig. 5) shows. The same holes were used for the main support postsfor the structure ; as a matter of fact it is not at all improbable thatthe same posts may have been in service in both occupations. Thepresence of the lower level was discovered when the post molds werebeing cleaned out. The old floor line was so plainly marked that theremoval of the upper level was decided upon.The fire and ladder pits occupied practically the same position ason the later level. As previously noted, a stone was embedded inthe floor at the east side of the fire pit. The bin or compartmentfeature at the ventilator side of the chamber was not as pronouncedin the earlier structure as it was in the remodeled form. Instead ofa low wall separating the ventilator portion of the room from theremainder of the chamber there was a ridge of adobe running fromthe base of each post at that side of the room to the ladder pit.Ridges in the floor in this position are quite common in ModifiedBasket Maker houses and also are present in structures belonging tothe Developmental Pueblo phase. Where an actual compartment isabsent, one frequently is indicated in this way. It was a not uncom-mon practice, in certain sections, to incorporate logs of wood in theseridges.^* There was no evidence of timber having been used in sucha way in No. 2; the plaster alone sufficed. In this the ridges corre-sponded to the type present in the Chaco Canyon.^?The fire pit in the first floor tended to an oval shape with one sideslightly flattened. No stones were used in it and the lining was
?Judd, 1926, p. 113, fig. 29.2? Roberts, 1929, pp. 51. 57-58. 59.
40 BUREAU OF AMERICAN ETHNOLOGY [Boll. 121entirely of plaster. The ladder box was a long rectangle in shape.The ends were abraded in such a fashion as to indicate clearly thatthe base of a runged ladder had rested there. In this floor one
SFT
FIGDBB 5.-Structure 2a. a, holes for main support posts; 6, stone embedded in floor;c, adobe ridges forming compartment ; d, sipapu ; e, fire pit ; f. ladder box ; g, storage hole
;
' ^T''! '" .'/' P?'*' "i^l'^; ^' storage pit; h, ventilator opening; I, ventilator shaft;m, nil above ventilator passage.feature was present that was lacking in the later level, namely, asipapu (fig. 5, d). In addition there was a second trench, sunilarto the ladder pit, in the floor just inside the compartment (fig. 5, h).
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA 4XAt the bottom of this groove, near one end, was a hole whose sidesbore the imprint of a pole. The only explanation which suggestsitself for this feature is that there may have been a plaster deflectorreinforced by a single upright bit of timber. Pole and mud, orwattle and plaster deflectors have been found at various places inthe Southwest and it is possible that this structure may have hadone. The floor around the edges of the depression indicated anupward slope as though plaster had extended up over some object.It is possible that the remains of such a deflector were removedwhen the second floor was placed in the pit. There were two holesin the floor at this place in the chamber (fig. 5, g^ j)- Both werecarefully plastered and possibly served for storage purposes. Theventilator during this stage of occupancy was larger than in thesubsequent one. Its features were sufficiently described in the dis-cussion of the upper floor to obviate need for further conmient.No. 2 was located at the end of the ridge upon which the firstgroup of structures was built. It was not as deep as the others andas a consequence the pillar left in the center of the room when thedebris was removed from it was not as tall. This condition wasaugmented by the fact that the pillar rested upon the upper or sec-ond floor level. It was 4 feet 6 inches (1.371 m) high and like theone discussed for structure 1, 3 feet (91.44 cm) square. The stratawere not as clearly marked as they were in No. 1 and for that reasonan arbitrary 6-inch (15.24-cm) level was adopted as the best meansfor studying the material. The earth and ashes were removed every6 inches (15.24 cm), put through a screen, and all of the objectsfound therein saved and tabulated.The general character of the fill from top to bottom in the pillarwas sand with a large ash and charcoal content. The first layer or6-inch (15.24-cm) level was composed in the main of sand and ashes.No charcoal was present. Other material, however, included pot-sherds, broken bones, and fragments of implements made from thatmaterial. In addition, there was some broken stone in the form ofspalls and chips.The second stratum contained a large proportion of charcoal,numerous chalcedony chips, and other fragments of stone. Therewas not as much bone as in the first, but quite a few fragments wererecovered. The showing of potsherds was not as marked in thislevel ; in fact there was a decided diminution in the number present.On the other hand, this layer yielded material which was absentfrom the first stratum, namely, obsidian flakes.More charcoal was present in the third layer than in the second.There was approximately the same amount of broken bones, a slightlysmaller number of stone chips, although the latter contained a good
^2 BUREAU OF AMERICAN ETHNOLOGY [Bull. 121percentage of obsidian flakes, and practically the same quantity ofpotsherds. The fourth level showed little change in the content ofcharcoal and bone in comparison with the third. It differed fromthe latter, however, in that the number of stone chips decreased.There was a sharp upturn in the amount of potsherds in the depositat this point. In fact they doubled in number in the fourth stratum.The fifth level again remained consistent in the amount of charcoalpresent. There was an increase in the number of broken animalbones and also the quantity of stone chips. The potsherds dimin-ished, however, and there were only half as many in this layer asin the one above.In the sixth level from the top there was an increase in the char-coal content accompanied by a decrease in the amount of both bonefragments and stone chips. The potsherds were more numerous thanthose found in the level above but did not reach the peak of thosein the fourth stratum. The sixth layer also contained large quanti-ties of building stone and spalls from such material. This was thefirst stratum in the fill where objects of this nature were present.Layer No. 7 contained an amount of charcoal equal to that presentin the sixth. Furthermore, it showed a marked increase in thenumber of building stones and ordinary chips. The bone contentwas the same as that in 6 but the quantity of potsherds was sharplydiminished.A somewhat smaller percentage of charcoal was observed in theeighth level than in the two above it. There was an increase in bonefragments and stone chips. The number of potsherds was also largerin this layer than in the one immediately above.The charcoal content in the ninth stratum was more plentiful thanin the eighth ; as a matter of fact this layer corresponded to the sixthand seventh in that respect. The amount of bone material in theninth was the equivalent of that in the eighth. This was also truefor the bone fragments and stone chips. There was a marked decreasein the number of potsherds.There are several points of significance in the data from the strati-graphic section. The greatest number of potsherds occurred in thefirst and fourth levels. The peaks in the quantity of bone fragmentsoccur in the first, third, fifth, and last levels. The flint chips werefound to occur in greatest numbers in the second, fifth, and eighthlevels. The last two peaks coincide with those of the bone fragments.The presence of obsidian was confined to the second and third layers.The only shell obtained came from the seventh layer. The greatestamount of charcoal appeared in the sixth, seventh, and ninth strata.The first had none at all and the remaining levels were about equalin their content.
EOBEHTs] ARCHEOLOGICAL REMAINS IN ARIZONA 43The material thus outlined may be considered to indicate two dis-tinct intervals when considerable refuse was deposited in the pit.One of these occurred near the top of the pillar and the other abouthalfway between the top and the upper floor. There was a slight in-crease just above the floor, but the amount of refuse did not corre-spond to the larger deposits at higher levels. The gap between themoderate showing of refuse at the lowest level, on top of the laterfloor, and midway of the pillar was quite pronoimced. That betweenthe midsection and top of the pillar was not so evident. The data ingeneral may be interpreted as indicating that for a short time afterthe house was abandoned it was used as a dumping place for refuse.Then for a considerable interval there was only a natural accumula-tion of material. Following this, refuse was again deposited in thedepression. After a considerable amount of this rubbish had beendumped there, another interim of natural accumulation occurred.The duration of this process was comparatively short and it was im-mediately followed by another round of active deposition of refusethat was even greater in proportion than the preceding one had been.This stage was succeeded by a gradual tapering off in the amount ofrefuse thus disposed of and the top layer again seemed to be oneresulting from natural accumulation. Just what the intervals whenonly slight amounts of refuse entered into the fill may indicate isproblematical. They possibly may be interpreted as representingperiods when there was little activity about the site. Perhaps, as wassuggested in the case of structure No. 1, the inhabitants may havemoved away for a time. In the first of the intervals this was oflonger duration than in the second. The evidence furnished by thepotsherds collected from the different levels has some bearing on theproblem, inasmuch as a certain progression in form is noted. Thiswill be discussed at some length, however, in connection with thesubject of pottery and need not be brought forward at this place.The charcoal fragments from the fill above the second floor in No.2 gave the dates 814 and 815, placing the second occupancy wellalong in the development of the Pueblo Pattern. Material from thefill between the original and upper floor levels could not definitelybe dated. It apparently belongs to the cycle occurring about themiddle of the eighth century. This would indicate that considerabletime elapsed between the abandonment of the structure in its originalform and the remodeling and reoccupation which took place in theninth century.The pit for No. 2 had a diameter of 13 feet 31/2 inches (4.051 m)on the fire pit, ladder, ventilator line. At right angles to tliisdiameter the chamber measured 13 feet 11 inches (4.241 m) from wallto wall. The depth varied at different points around the chamber.
.. BUREAU OF AMERICAN ETHNOLOGY ^^^^ 121This probably was due to the fact that the structure was built on adecided slope. For the upper floor at the ventilator side of the roomthe hei<^ht of the wall was 3 feet 2 inches (96.52 cm) ; at the oppositeside it was 3 feet 6 inches (1.066 m) . At the west side of the chamberthe wall was 4 feet 1 inch (1.244 m) above the floor and at the east2 feet 9 inches (83.82 cm). The latter was in part attributable tothe slope of the ground and to some extent to the weathering awayof the surface at that point.The holes for the support posts averaged 1 foot 1 inch (33.02 cm)in diameter. Their depth varied from 1 foot 6 inches (45.72 cm) to2 feet 3 inches (68.58 cm). Their distances from the wall were 1foot 8 inches (50.8 cm) for the north hole, 41/2 inches (11.43 cm) forthe east post, 10 inches (25.4 cm) for the south, and 1 foot 6 inches(45.72 cm) for the west.The fire pit for the upper floor had diameters of 1 foot 8 inches(50.8 cm) and 1 foot 10 inches (55.88 cm). It was 6^4 inches (15.87cm) deep. The rectangular pit measured 1 foot 11/2 inches (34.29 cm)by 1 foot 1 inch (33.02 cm) . It had a depth of 5% inches (14.61 cm)
.
There was no sipapu or deflector on this level. The only hole in thefloor, in addition to the places where the posts had been set and to thefire and ladder pits, was the small circular one (fig. 4, A) at thesouth side of the room. It had a diameter of 5 inches (12.7 cm) andwas 7 inches (17.78 cm) deep.The low walls which formed the binlike compartment at the ven-tilator side of the chamber varied in width. That at the north sideof the ladder pit averaged 1 foot (30.48 cm), while that at the southranged from 8 inches (20.32 cm) to 11 inches (27.94 cm). The heightalso varied. That at the north end averaged 7^/^ inches (19.05 cm)and the one at the south lOi/^ inches (26.67 cm). It is possible thatthere was not as great a discrepancy between the heights of the twoat the time when the house was occupied because there were clearindications that a portion of the top of the north wall had beenbroken off. The north partition at the end next to the support postwas 9 inches (22.86 cm) from the wall, and in front of the ventilatorwas 2 feet 2 inches (66.04 cm) from it. The south partition was 1foot (30.48 cm) from the wall of the chamber at the end next to theroof post and 2 feet 21/2 inches (67.31 cm) separated it from the ven-tilator opening. The stones which formed the ends of the partitionsnear the ventilator opening had heights of 7 inches (17.78 cm) forthe north one and 6I/2 inches (16.51 cm) for the south. That at thenorth side had a thickness of 1% inches (4.76 cm), while the southstone measured 1% inches (4.45 cm).The pilaster erected behind the support post at the south side of theroom was 10 inches (25.4 cm) thick. Along the wall it measured 81/2inches (21.59 cm) and where the post stood 4 inches (10.16 cm). Its
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA 45height at the time when the room was excavated was 2 feet 8 inches(81.28 cm). There was sufficient material in the debris at the base,however, to indicate that it had originally risen to approximatelythe height of the wall.The ventilator opening was 7 inches (17.78 cm) wide at the bottomand 6 inches (15.24 cm) at the top. The aperture was 8 inches (20.32cm) high. The sill was 4 inches (10.16 cm) above the floor level. Thepassage from the opening in the wall of the chamber to the base of theshaft was 5 feet 1 inch (1.549 m) long. The opening into the shaft was10 inches (25.4 cm) high and 9 inches (22.86 cm) wide. The shafthad a diameter of 1 foot 9 inches (53.34 cm) on the passage line and1 foot 10 inches (55.88 cm) at right angles to it. The shaft was 2 feet6 inches (76.2 cm) deep at the tunnel side and 2 feet (60.96 cm) at theother.The wall heights for the original occupation of No. 2 were 3 feet9 inches (1.143 m) at the ventilator, and 4 feet 2 inches (1.270 m)at the opposite side of the chamber, 3 feet 9 inches (1.143 m) at thenorth, and 4 feet 10 inches (1.473 m) at the south side of the room.The latter measurement was taken from the top of the wall at itspresent level, which is that of the second occupation of the structure.Making allowances for the additions to the wall when the second floorwas placed in the chamber, the original height above the floor at thatside was 4 feet (1.219 m).The floor features associated with the first level of occupation haddifferent measurements from those on the later one. The fire pit haddiameters of 2 feet 4 inches (71.12 cm) and 1 foot 8 inches (50.8 cm).It was shallower than the upper pit and had a depth of only 3 inches(7.62 cm). The ladder pit was 3^ inches (8.99 cm) from the firepit. It had a length of 1 foot 7^2 inches (49.53 cm) and a widthof 7 inches (17.78 cm). On the fire pit side it had a depth of 31^inches (8.89 cm) and on the side toward the ventilator was 5 inches(12.7 cm) deep. This difference was due to the fact that the floorlevel in the com.partment was higher than that throughout the restof the chamber.The rectangular depression just inside the compartment, which itwas suggested might represent the place where the deflector had stood,measured 1 foot 6 inches (45.72 cm) long and 4% inches (12.065 cm)wide. It had an average depth of 1^/^ inches (3.81 cm). The oval-shaped hole near one end of this groove (fig. 5, i) had diameters of4 inches (10.16 cm) and 3 inches (7.62 cm) and was 2% inches (6.67cm) deep.The hole in the floor between the ends of the compartment ridges(fig. 5, g) had diameters of 3iA inches (8.89 cm) and 4i/^ inches(11.43 cm). It had a depth of 314 inches (8.26 cm). The sipapu was
. g BUREAU OF AMERICAN ETHNOLOGY [Bull. 1211 foot 3 inches (38.1 cm) from the fire pit and 5 feet 51/2 inches (1.663m) from the wall. The hole had diameters of Q% inches (17.15 cm)and 41/0 inches (11.43 cm). It was much deeper than the averagewith a measurement of 1 foot 41/2 inches (41.91 cm) from the floorlevel to its bottom.The adobe ridges which formed the compartment at the ventilatorside of the chamber were more irregular in width than the wallsto the similar inclosure on the upper level. The ridge at the northside was 1 foot 1 inch (33.02 cm) wide at the base of the supportpost and tapered down to a width of 41/2 inches (11.43 cm) at theend in front of the ventilator opening. The average height of thisridge was 3 inches (7.62 cm). The south ridge was 8I/2 inches (21.59cm) wide at the base of the post and 6 inches (15.24 cm) across atthe end in front of the ventilator. There was a circular hole in thefloor at that end of the ridge which had a diameter of 6 inches (15.24cm) and a depth of 4% inches (10.48 cm). The average height ofthe south ridge was 3 inches (7.62 cm).The ventilator opening in its original form was 1 foot 3 inches(38.1 cm) wide and 1 foot 8 inches (50.8 cm) high. The sill was 3inches (7.62 cm) above the floor. The passage had the same lengthduring this occupation as it had for the later one. The width of thetunnel where it entered the shaft was 9I/2 inches (24.13 cm) and itsheight 10 inches (25.4 cm) . There was no difference in the diametersof the shaft between this stage and the later one. There was somevariation in depth at the side where the tunnel opened into the shaft.During the original occupation this was only 2 feet 2i/^ inches (67.31cm) instead of 2 feet 6 inches (76.2 cm) as recorded for the laterinhabitation. The depth at the side opposite the tunnel opening wasthe same. STRUCTURE 3Structure No. 3 was located between Nos. 1 and 2, as shown onthe ground plan (fig. 1). No indications of this ruin were ob-served on the surface and it was not until after Nos. 1 and 2 hadbeen cleared completely of their accumulated debris and a trenchwas run between them that the third pit was found. The excavatedportion of No. 3 was roughly D -shaped (fig. 6). The wall alongthe ventilator side was flattened to some extent and produced thisform. The structure is of particular interest because it gave dis-tinct evidence of having been covered with a cribbed roof. It wasthe only one in the entire group which appeared to have had thattype of covering erected over the pit. The roof had been destroyedby fire but the timbers were only partially consumed and as a con-sequence niost of them were lying on the floor in positions whichclearly indicated the manner in which they had been placed in theframework.
ROBBKTS] ARCHEOLOGICAL REMAINS IN ARIZONA 47The superstructure was supported by four posts set in the floorand one pilaster or masonry pillar. Two of the support postshad been braced with mud plaster placed between them and the wallin the same fashion that the post in No. 2 had been buttressed (pi.7, 6). One curious factor in this comiection is that in both Nos. 2
FiODRK 6.?Structure 3. a, support posts for superstructure ; h, pilasters ; c, storagebasin ; d, storage liole ; e, sipapu ; f, fire pit ; g, ladder box ; h, storage hole ; i, stoneembedded in floor ; j, secondary fire pit ; k, ventilator shaft ; m, sand accumulation aboveold surface.and 3 one of the same posts, that at the south corner of the room,had been braced. Whether this was a "happenstance" or had somemarked significance is not known. It seems rather curious thatposts in the same position in different houses would have neededbracing, although it is possible that the superstructure in each case
.g BUREAU OF AMERICAN ETHNOLOGY [Boll. 121may have been forced out of position as the result of winter snowsresting upon the north side of the roof for a longer period than onthe south. Wliile this explanation is a plausible one there possiblywas some other reason more specifically responsible, but what thatmay have been can not be stated.The cribbed portion of the roof conformed to the general typethat has been observed throughout the northern portion of theSouthwest, the plateau area. Logs were laid from support postto support post, including the masonry pilaster, around the wallof the chamber, forming a pentagonal framework. None of thebeams passed over the center of the room, as in the case of thoseforming the main frame in the type of superstructure described forthe other structures, but followed the periphery. In this exampleof cribbing, the timbers were placed in pairs, possibly because theywere rather small, although single beams were frequently employedin building roofs of that type. Each succeeding row of logs wasplaced so that the beams cut across the corners of the precedingseries. Each new set also formed a smaller pentagon and ap-proached nearer to the center of the chamber until the frameworkhad been built up into a kind of dome (fig. 7). The portion abovethe central part of the room generally had a flat surface borderingthe smoke hole and entrance hatchway. The proportion of crib-bing to flat construction varied from structure to structure. Insome a large area constituted a flat ceiling while in others the cover-ing was almost entirely dome-shaped. In the case of No. 3 thereprobably was more of the cribbing, although it cannot be definitelyso stated. Indications suggested such a condition but the evidencewas not clear enough to warrant a positive assertion to that effect.As was mentioned in the introduction to the subject of houses,it has been thought that a cribbed roof was characteristic of kivas orceremonial chambers. When No. 3 was in the x)rocess of excavationthe belief was that it represented another structure similar to Nos.1 and 2 which appeared to have functioned as dwellings. Afterthe roof timbers were found on the floor in positions showing defi-nitely that the roof construction had been of the cribbed type, thequestion was raised as to whether or not this structure had func-tioned primarily as a domicile or as a ceremonial chamber. Afterall of the debris had been removed there was little in the pit to sug-gest a dwelling. Furthermore, indications were that it was muchlater than either of the adjacent pits. This evidence was later sub-stantiated by the date obtained from the timbers. The latter showsthat No. 3 was erected 50 years after the remodeling activities in No.2, at approximately 867 A. D. In view of this fact, and also of thesimilarity between pottery fragments found in this structure andthose obtained from the nearby surface remains, it would seem that
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA 49No. 3 may have been the ceremonial chamber for the dwellers in thathouse (fig. 1, Surface Kemains A). If this were true the presenceof a cribbed roof can be explained on the grounds that it was kivaand true to type, although still in an early stage of development.On the other hand, it is possible that the pit-house builders adopteda crib style of roofing before the transition from semisubterraneanto above-ground houses. As the structure under consideration rep-resents a stage approximately coinciding with the period of changein this district it may qualify as an example of the evolution inroofing practices. The shift from house to ceremonial chamber isshown in other structures dating from this same horizon, but thisis the earliest cribbed roof at the site.There was no bench around No. 3. The sides of the pit extendedfrom the surface of the ground to the floor without a break. Inthis the structure corresponded to those previously described. In-terior features were simple. Near the center of the room was a sipapu(fig. 6, e), fire pit, and ladder box (fig. 6, /, g). There were twostorage pits near the west and north support posts (fig. 6, c, h). Atthe base of the south post and extending along the wall of thechamber was a shallow depression lined with stone slabs (fig 6, j)that appeared to have functioned as a secondary fire pit. It had notbeen used to any great extent, however. This may possibly be at-tributed to the fact that it was so close to the support post that itwould have been a decided fire hazard. Its position, also, was suchthat smoke from it would have tended to cling about the ceiling andwould not have passed out through the smoke hole as readily ascould be desired. One function which suggests itself is that the pitwas not employed for actual fires but rather as a container for livecoals from the main pit near the center of the room. In such acapacity it could have served as a crude form of brazier. Charcoalfires of that type are not known to have been prevalent in the South-west, but they were in use in regions to the south and may havebeen employed occasionally in this district.Embedded in the floor at the south side of the fire pit, about mid-way between it and the south support post, was a slab of stone thetop of which was flush with the floor (fig. 6, i). Except for beingplaced at a different side of the chamber, this stone correspondedto those described for both levels in No. 2 and may have served thesame purpose.The fire and ladder pits in this structure (fig. 6, /, g) were com-bined in that they were surrounded by a rim of adobe plaster whichinclosed both basins. Both were dug into the floor and were linedwith plaster. A low ridge of the same material separated them.Two stones were set in the ridge and reinforced it. The ladder
KQ BUREAU OF AMERICAN ETHNOLOGY [Boll. 121pit had a fairly large stone incorporated in it near the east cornerand another at the south. The fire pit was rather shallow but itsdepth was increased to some extent by the bordering plaster. Therewere no ashes in the ladder pit and its bottom and end showedclearly the marks of abrasion caused by the ladder poles. Therewas no deflector between the ladder pit and the ventilator, but pro-vision for shutting off the draft was made in the form of a smallstone cover to place over the opening in the wall.There was no bin or compartment at the southeast side of thechamber, not even an indication of one. The pilaster at the northside of the room (fig. 7, b) was constructed of large blocks of stone,but roughly shaped, laid in adobe plaster. The amount of plasterin proportion to the stone varied from course to course. In somecases it was quite thick; in others it barely separated the stone.The face and sides of the pilaster were not true. One stone in acourse might i)roject some distance beyond those immediately aboveand below, or another not extend to the edge of the pillar. Thisunevenness was lessened to some extent by the application of thickcoats of plaster, so that when the structure was in use the pilasterhad a fairly regular appearance. Despite its crude construction itseems to have functioned in a satisfactory manner so far as support-ing the roof was concerned.The ventilator was smaller than the average for such structures,although the shaft was comparatively large. The horizontal por-tion of the ventilator had been worked out by trenching. The aper-ture in the wall of the chamber was at a higher level than in mostof the structures and had a sill of stone embedded in adobe plaster.The oj)ening was reinforced with a rim of plaster, containing a slightoffset in which the cover stone rested. There was an upward slant tothe floor of the passage and the shaft at the outer end was almostcircular in form.No stratigraphic study was made of the fill in No. 3 because the])ilhir which was left in the center of the room collapsed before itcould be removed and put through a screen in the usual way. Pre-liminary notes made during the course of tlie excavation, however,record the fact that most of the upper half of the debris in the pitconsisted mainly of clean wind-blown sand. There was very littlerefuse or broken material in the remaining portion above the burnedtimbers of the superstructure. There was little indication of thepit having served as a dumping place. As a matter of fact the fillsuggested that in the interval shortly after the structure was de-stroyed by fire there was very little activity in its vicinity and thematerial which collected in the pit was an accumulation due to theaction of surface water rather than human agents. The timbersalvaged from the floor of the pit was carefullly studied and gave
BOBBSTB] ARCHEOLOGICAL REMAINS IN ARIZONA 51
boa
^2 BUREAU OF AMERICAN ETHNOLOGY [Boll. 121the dates 842, 852?5, and 867, which are from 38 to 52 years laterthan those obtained in No. 2.No. 3 measured 11 feet 9 inches (3.581 m) in diameter on the linethrough the sipapu, fire pit, and ladder pit. At right angles to thisline it was 13 feet 2 inches (4.013 m) from wall to wall. At theventilator opening the wall was 4 feet 4 inches (1.320 m) high. Thefill above the old ground level was 1 foot 8 inches (50.8 cm) deep,so that the floor when the structure was excavated was 6 feet(1.828 m) below the present surface. At the opposite side of thechamber the wall was 4 feet 61/0 inches (1.384 m) high at the timeof occupancy. Wlien excavated the sand accumulation had increasedthe measurement to 7 feet 3 inches (2.209 m). The posts used instructure No. 3 ranged in size from 5 to 9 inches (12.7 to 22.86 cm).The charred ends of all of them were present in the floor. The northpost stood 1 foot 4 inches (40.64 cm) from the wall. That at the eastcorner was 2 inches (5.08 cm) from it. The south pole was 3i/^inches (8.89 cm) from the wall and the space between was filled withadobe plaster. The third post was partially embedded in the wall.The pilaster at the east side of the chamber (fig. 6, h) measured6 by 10 inches (15.24 by 25.4 cm) . It was 3 feet 3 inches (99.06 cm)high when the pit was excavated. Around its base there was anadobe reinforcement over the floor. The edges of this feature started1 foot 71/2 inches (49.53 cm) from the base of the pillar at the northside, was 1 foot 10 inches (55.88 cm) from its face, and 1 foot(30.48 cm) from its south side. This reinforcement was flush withthe floor at its outer edge and rose to a level 2 inches (5.08 cm) highat the base of the pilaster.The fire pit had diameters of 1 foot 3% inches (39.37 cm) and 2feet 1/2 inch (62.23 cm). It had a depth of 4 inches (10.16 cm). Therim around the pit had an average width of 8 inches (20.32 cm) andwas 2 inches (5.08 cm) high. It sloped upward from its outer bor-ders to form a summit almost at the edge of the fire pit. The ridgeseparating the fire and ladder pits was 5 inches (12.7 cm) wide and314 inches (8.99 cm) high. The ladder pit measured 1 foot 7 inches(48.26 cm) by 1 foot 5 inches (43.18 cm). It was 7 inches (17.78cm) deep at the side toward the ventilator. The adobe rim at theend of the ladder pit was 614 inches (16.51 cm) wide and li/^ inches(3.81 cm) high.The sipapu (fig. 6, e) was 6 inches (15.24 cm) from the fire pit.It had diameters of 10 inches (25.4 cm) and 11 inches (27.94 cm),and a depth of 91/2 inches (24.13 cm). The hole at the east side ofthe fire pit, ladder pit combination (fig. 6, d) was 61/2 inches (16.51cm) from the adobe rim around the pits. The hole had diametersof 101/2 and 11 inches (26.67 and 27.94 cm). It had a depth of 6inches (15.24 cm). The oval depression between the north roof post
ROBERTS I ARCHEOLOGICAL REMAINS IN ARIZONA 53and the wall (fig. 6, c) had diameters of 1 foot 5 inches (43.18 cm)and 11 inches (27.94 cm). Its average depth was 41/2 inches (11.43cm). The hole near the third post (fig. 6, 5) was 21/2 inches (6.35cm) from the wall; it had diameters of 9 inches (22.86 cm) and 10inches (25.4 cm) on the floor level. At the depth of 21/2 inches(6.35 cm) below the floor a second inner hole with diameters of 5inches (12.7 cm) and 31/2 inches (8.89 cm) extended downward 4inches (10.16 cm). The oval-shaped depression near the south roofpost (fig. 6, /) had diameters of 1 foot 7 inches (48.26 cm) and 11inches (27.94 cm) ; it was 4 inches (10.16 cm) deep.The ventilator opening was 1 foot 21/2 inches (36.83 cm) high and1 foot 11/4 inches (34.29 cm) wide. The sill of the opening was 8inches (20.32 cm) above the floor. The ventilator passage had alength of 4 feet 9 inches (1.447 m). The shaft had diameters of2 feet 31/2 inches (69.85 cm) and 2 feet 51/2 inches (74.93 cm). Itsoriginal depth was 3 feet 21/2 inches (97.79 cm). At the time of theexcavation the sand which had drifted across the top of this partof the site increased the depth to 4 feet 6 inches (1.371 m). The pas-sage opening into the shaft was 1 foot 3 inches (38.1 cm) wide and1 foot 3 inches (38.1 cm) high.SURFACE HOUSE AND BRUSH SHELTERS NEAR STRUCTURES 1,2, AND RJust west of the pit remains described in preceding pages werethe ruins of a small surface house and three brush shelters. Onlythe foundations of the house and a few courses of stone remained(fig. 8). The structure appeared to have been one containing fourrooms. Three of these were built in a row, while the fourth waserected at the north and projecting toward the east so that a smallcourt was formed. A fragment of the wall south of this room ex-tended on toward structure 3, suggesting to an even greater degreeprovision for a court. Only portions of the two inclosures at thenorth end of the building were present; the rest of the foundationspresumably were washed away. The ruin was located on the edgeof a slope whose surface gave indications of having been eroded to aconsiderable extent and there was sufiicient wall material scatteredover the slope to warrant belief that the walls necessary for the com-pletion of the building once stood there. The probable extent ofthese walls is indicated in the drawing (fig. 8) by dotted lines. Thechamber that formed the east wing of this small dwelling containeda fire pit. The room at the south end, that is the remaining portionof it, had a paved floor. The other two rooms had little of interestabout them. The masonry in the walls was rather crude. Most ofthe stones employed in the construction were used in their naturalstate, little attempt having been made to shape them. The builders68764?39 5
54 BUREAU OF AMERICAN ETHNOLOGY [Bull. 121used a large amount of adobe mortar to compensate for the irregularity of the stones. There was very little fallen wall material orthe surface around the house. This lack may be explained on tli(grounds that the builders of the later structure probably robbed th(older ruin of its stone for use in the newer construction. At twcplaces in the central room in the west tier the foundation and walU
D tL=^ pijrw m U
10 FEETFIGURE 8^?Remains of surface building A. a and 6, fire pits beneath foundations ; c, firepitin floor of room; d, opening into pit oven; e, t, and g, placements for portico posts;?, ^,i, k, I, m, n, and 0, placements for shelter posts ; m, location of fragment of child'sskull ; y, location of blacli-on-white bowl.extended across former fire pits (fig. 8, a, h). At the time thestructure was built these pits were apparently on or just below theexisting level of occupation. They were filled with stones and mudbefore the wall was carried across them (pi. 8, a). Another feature,that of a porch or portico, was indicated by the row of post moldsor holes along the front of the structure. A similar feature wasfound near house 15 and inasmuch as the latter exhibited better theI
BUREAU OF AMERICAN ETHNOLOGY BULLETIN 121 PLATE 7
a. Compartment side of stnic-ture 2.
//. I'la.-ter iiila>ier lieliiii'l w It'll roof supporl in structures.
BUREAU OF AMERICAN ETHNOLOGY BULLETIN 121 PLATE
n. Klie pit luMR'iith wall (ilsmlace strtieture A.
Ii^ <_i!X'.iiiirj iiiiri pii ovpii. Trciwcl iinlioates 111
BUREAU OF AMERICAN ETHNOLOGY BULLETIN 121 PLATE 9
n. Stone ventilator aperture in stnulnre li
b. Fire pit in fill of structure li.
BUREAU OF AMERICAN ETHNOLOGY BULLETIN 121 PLATE 10'
(I. Wnlil >ti>i- ^ile 111 slriK'ttire 9.
b. Portion of structure 11.
Roberts] AtlCfiEOLOGlCAL REMAINS IN ARIZONA 55nature of the construction, detailed discussion will be reserved forthe section describing that dwelling. As was suggested in connectionwith structure 3, the latter may have been a ceremonial chamber.If it was a kiva for the small house, the combination would be agood example of an early form of the unit-type structure of theDevelopmental Pueblo period. Potsherds from the floor of the sur-face house are typical of the vessel styles associated with that era.The pit oven in the court formed by the angle in the surface housedid not belong to the same period as the house (fig. 8, d). The topof the pit was on the level of occupation corresponding to that ofstructure No. 1 and seemed unquestionably to belong to that horizon.Above the oven was a layer of clean sand on which there was a harderlayer of adobe clay. The foundations for the surface structurerested upon this level, which showed conclusively that they were of alater date than the oven (fig. 10, upper section). The oven was ofparticular interest for two reasons : one was associated with the actualform of the pit itself and the other its similarity to an example foundpreviously on the Zuiii Reservation in western New Mexico.^'' Thedistinctive feature of both of these pits was a flue extending fromone side to the ground level. The pit was jug-shaped with a narrowopening at the top (fig. 9).The interior gave evidence of hot fires having burned there untilthe plaster was baked to a bricklike consistency. The earth wallsback of the lining were colored a distinct red by the heat. Just whatovens of tliis type were used for is not known. It is possible thatthey functioned in the same way as the pi-gummi ovens of the Hopi,in wliich a corn-meal mush bearing the same name is cooked. Thelatter ovens have been known for a long time but it is only recentlythat prehistoric examples have been found. In Mindeleff's descrip-tion of the modern type he states that the flue was provided not, asmight be expected, to increase the draft but in order that the fireburning in the interior might be poked and arranged from time totime.^^ Mindeleff also suggested that baking pits of this type prob-ably were the stem upon which the domelike Spanish oven now foundin all the pueblos was grafted. They are alike in principle andgeneral form and it would be a simple step to shift from the one tothe other. In general practice the methods of heating and of usagewould be similar. The main difference, of course, is that one wasbeneath the ground level while the other was built above it. Thecustom was to light large fires in the pits and heat the walls to thedesired temperature. The fire, charcoal, and ashes were then re-moved, the material to be cooked was placed inside, and the openingsealed.
8" Roberts, 1932, pp. 44-46.31 Mindeleff, 1891, pp. 163-164.
56 BTJREAtJ OF AMERICAN ETHNOLOGY [Bull. 121There was a slight difference between the example found on theZufii Keservation and that uncovered at this location. In the Zunipit the flue was oval to circular in form. It sloped obliquely upwardand its outer end opened on the surface some distance from the mainaperture of the oven.^^ The flue of the oven in the court of Surface
Z FEET
FiGDDE 9 h, opening on ground level.House A was a rectangularly shaped slit rising vertically from themain pit. The opening on the ground level was only a few inchesaway from the mouth of the oven (pi. 8, 6; fig. 9, a).The brush shelters were simple arborlike structures erected on theground level. They consisted of four upright posts which probably
32 Roberts, 1932, fig. 6.
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA 57supported fiat roofs of brush. The framework may have been m-closecl on two or three sides by poles and branches leaning againstthe framework, but there were no indications that side pieces hadbeen embedded in the ground. The remains of these structures sug-gest that they were like the summer hogans and wickiups used bymodern Indians in the region. The people may have lived in thepit structures in the winter and during the suimner months spentmost of their time on the surface outside. Evidence from earlyPueblo villages in various sections has indicated that considerableuse was made of such shelters or arbors, particularly for cooking andallied tasks. Harrington reports such remains in Nevada.^^ Houghfound them near Luna, N. Mex.^^ Judd observed them in the courseof his work in Utah,^^ and an example was found by the writer inhis work in southern Colorado.^^ It is even possible that during thegrowing season similar temporary shelters were erected on the val-ley bottom below and the people lived there during the time whenthe crops were reaching maturity. After the harvest they could havereturned to the hilltop location for the winter season. All of thebrush shelters indicated on the plot near the surface remains werenot contemporaneous. Two of them lettered A and B on the plan(fig. 8) were on the same level as the surface house. O, however,was on the level of occupation which corresponded to that of the pitoven and structure 1 (fig. 10, lower section). The shelter locatedsome distance from the surface house group, southwest from struc-ture 1, and midway between it and No. 4, figure 1, also dated fromthe early horizon.The surface house had a total length of 25 feet (7.G20 m) anda width at the north end of 15 feet 9 inches (4.800 m). Across themiddle of the central room the total width was 7 feet 7 inches (2.311m). The east room at the north end of the building had an east-west measurement of 7 feet 11 inches (2.413 m). The remaining eastwall was 1 foot 5 inches (43.18 cm) long and that at the west 1 foot11 inches (58.42 cm). The west room at the north end of the housewas 5 feet 4 inches (1.625 m) wide and the remaining west wallwas 5 feet 9 inches (1.752 m) long. The central room was 5 feet8 inches (1.727 m) wide and 8 feet 51/2 inches (2.577 m) long. Thesouth room had a width of 5 feet 7 inches (1.701 m) and the westwall, which was the longest remaining, measured 9 feet 3 inches(2.819 m).The opening to the pit oven was oval in shape. It measured 2 feet7 inches (78.74 cm) on the long diameter and 2 feet 1 inch (63.5
S3 Harrington, 1927, p. 267.s* Hough, 1919, pp. 415-416, pi. 35.3^ Judd, 1926, p. 29.3? Roberts, 1930, p. 72.
58 BUREAU OF AMERICAN ETHNOLOGY [Bdll. li^lcm) on the short. The pit had a depth of 2 feet 6 inches (76.2cm). At its bottom the diameter was 3 feet 11 inches (1.193 m).The flue was 1 foot 1 inch (33.02 cm) from the mouth of the pit. Theflue measured 9 inches (22.86 cm) by 3 inches (7.62 cm). It was 2feet (60.96 cm) long. The bottom of the flue was 6 inches (15.24cm) above the floor of the oven.SURFACEHOUSE
SHELTER B__
10FEETFigure 10.?Sections through surface remains A, pit oven and structure 3, shelter andstructure 1. a, accumulation of saud above original surface of occupation.STRUCTURE 4Structure No. 4 stood alone and was placed in the earth along theridge some distance from the group of three pits and surface re-mains discussed in preceding pages (fig. 1). It had a roughly D-shaped pit; two sides were curved and two were practically straight(fig. 11). The pit was excavated with straight sides from groundlevel to floor and had no bench. There were four holes for postsnear the corners of the chamber, indicating that the superstructurehad been of the flat-ceiling, sloping-side-wall type common in themajority of the structures at the site. None of the timbers remained,however, and only the holes where the main posts stood were left toshow that they had been present.
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA 59The interior features of No. 4 were simple. There was a fire pitnear the center of the room (fig. 11, c), a small irregularly shapeddepression in the floor at the north side of the chamber (fig. 11, h),and between the fire pit and ventilator opening a deflector (fig. 11, e).The latter was flanked at either end by additional upright stonesembedded in the floor (fig. 11, d, f) . The latter suggested the begin-nings of a bin or compartment. There was nothing to show that theventilator side of the room had been completely separated from theremaining portions of the chamber, as in Nos. 1 and 2, hence, thetwo stones must be considered as augmenting the deflector ratherthan forming a compartment.The fire pit was a simple depression in the floor and was facedwith plaster. No stones were used in its lining. In the floor box atthe north side of the room (fig. 11, h) stones were used to reinforcethe sides. What this box may have been used for is not known. Itprobably functioned as a storage place, since there was no indicationthat fires had been laid in it. No traces could be found of a sipapuand there was no ladder pit. The surface of the floor between the firepit and the deflector was roughened and abraded to some extent, sug-gesting that the base end of a ladder had rested there. Possibly thedeflector stone itself kept the poles fi"om slipping on the surface ofthe floor. The deflector consisted of an upright slab of stone. It ispossible that the spaces between it and the other two stones (fig. 11,
c?, /) at one time had been filled with plaster, making the entiregroup one large deflector or fire screen, although no evidence of suchplaster remained.The ventilator simply consisted of a tunnel and shaft. The featuredistinguishing it from the ventilators in the structures previouslydescribed was that of the shortening of the tunnel. It was quite smallin comparison with the others. The shaft at the outer end wascomparable to those previously described. The main difference wasthat in No. 4 the bottom of the shaft had a pronounced slope fromfront to back, while in the others it was practically level (fig. 11,section) . The passage portion of the ventilator was tunneled throughthe earth between the bottom of the shaft and the chamber, and thefloor of the shaft sloped slightly downward toward the tunnel.The fill in No. 4 had nothing of significance to offer. The upperpart of the pillar was clean sand and the lower was of sand slightlystained with ash. There were no distinct strata in it. Charcoal waslargely absent and there were no potsherds. The pit unquestionablywas never used as a dumping place. Furthermore, there was nothingto indicate the relationship between this and the other structures.The few potsherds on the floor were from culinary vessels of theDevelopmental Pueblo banded-neck type.
QQ BUREAU OF AMERICAN ETHNOLOGY [Boll. 121No. 4 had a diameter of 10 feet 41/2 inches (3.162 m) on the linethrough the ventilator and fire pit. Across the other way it measured12 feet 91/2 inches (3.898 m) from wall to wall. At the ventilator sidethe original depth of the pit was 3 feet 51/2 inches (1.054 m). There
t::r ^
SFESr
Figure 11.?Structure 4. a, holes for main support posts ; h, storage tox iu floor ; c, firepit ; d, stauding stone ; e, deflector ; f, standing stone ; g, ventilator sliaft ; h, sand accumulation above original surface.were 2 feet (60.96 cm) of fill above the old level which gave the pita total depth of 5 feet 51/0 inches (1.663 m) when it was excavated.At the opposite side of the room the original wall rose 5 feet (1.524m) above the floor level. The accumulation of sand above the orig-
Roberts] ARCHEOLOGICAL REMAINS IN ARIZONA Qlinal surface was only 1 foot 9^2 inches (54.61 cm) deep. This, how-ever, gave the pit a total depth of 6 feet 9i^ inches (2.070 m) at thetime when the debris was removed from its interior.The holes for the posts varied somewhat in size but not to asmarked a degree as in the case of some of the other structures. Thehole at the northwest corner of the room had diameters of 8I/2 inches(21.59 cm) and 10 inches (25.4 cm). It was 1 foot 1 inch (33.02 cm)from the wall and was 1 foot 3 inches (38.1 cm) in depth. The holeat the northeast corner had diameters of 7^4 inches (19.05 cm) and 8inches (20.32 cm). It was 1 foot 5 inches (43.18 cm) deep and stood10 inches (25.4 cm) from the wall. The southwest hole was moreoval in shape with a long diameter of 9 inches (22.86 cm) and a shortone of 6 inches (15.24 cm). It was only 1 foot 2 inches (35.56 cm)deep and was placed 71/2 inches (19.05 cm) from the wall. The holeat the southwest corner was the smallest of the group and morenearly circular in contour; it had diameters of 7^2 inches (19.05 cm)and the depth was 1 foot 7 inches (48.26 cm). From the edge of thehole to the wall measured 1 foot 3i^ inches (39.37 cm).The fire pit (fig. 11, c) had a diameter of 1 foot 5 inches (43.18 cm)on the ventilator line, and 1 foot 9 inches (53.34 cm) at right anglesto it. Its depth was 6^ inches (16.51 cm). The slab-lined pit at thenorth side of the chamber (fig. 11, h) was practically midway betweenthe two support posts at that side of the room. It was 2 feet 3I/2inches (69.85 cm) from the northwest pole and 2 feet 7 inches (78.74cm) from the one at the northeast corner. It was 1 foot 7% inches(49.53 cm) from the wall and 2 feet 4i/^ inches (72.39 cm) from thefire pit. The box itself measured 1 foot 3 inches (38.1 cm) by 1 foot% inch (31.75 cm). The box was 6 inches (15.24 cm) deep.The main deflector stone (fig. 11, e) stood 1 foot 6 inches (45.72 cm)from the fire pit. The stone measured 1 foot 9 inches (53.34 cm) inlength, had a thickness of 21/2 inches (6.35 cm) , and stood 11 inches(27.94 cm) above the floor. A space of 4 inches (10.16 cm) separatedit from the stone at its north end. This stone (fig. 11, d) had a lengthof 61/2 inches (16.51 cm), a thickness of 2 inches (5.08 cm), and aheight of 1 foot 4 inches (40.64 cm). The third stone in the groupwas 3 inches (7.62 cm) from the south end of the deflector (fig." 11 /).The stone had a total length of 1 foot 5^/^ inches (44.45 cm), wasiy2 inches (3.81 cm) thick, and stood 1 foot 11 inches (58.42 cm)above the floor. From the edge of the deflector to the ventilatoropening measured 2 feet l^ inch (62.23 cm).The ventilator opening was 1 foot (30.48 cm) high and 1 foot 2inches (35.56 cm) wide. The sill was 3 inches (7.62 cm) above thefloor. The passage was 1 foot lOi/^ inches (57.15 cm) long. Whereit entered the shaft it was 1 foot 1 inch (33.02 cm) high and 1 foot
g2 BUREAU OF AMERICAN ETHNOLOGY [Bull. 1212 inches (35.56 cm) wide. The ventilator shaft was oval in shapewith a long diameter on the passage line of 2 feet 6 inches (76.2 cm)and a short diameter of 2 feet 21/2 inches (67.31 cm). The originaldepth of the shaft at the passageway was 3 feet 41/2 inches (1.028 m).The 2 feet 21/2 inches (67.31 cm) fill of sand above the old groundlevel increased the depth to 5 feet 6 inches (1.676 m). At the timewhen the excavation was made at the back of the shaft the originaldepth was 2 feet 9 inches (83.82 cm), with a sand fill of 2 feet 41/2inches (72.39 cm). The depth was increased to 5 feet 11/2 inches(1.562 m). STRUCTURES 5A AND 5BThe remains of No. 5 were among the most interesting of the findsmade during the season of 1931. Two structures had occupied prac-tically the same position at different periods and for that reason weredesignated as 5a and 5b. Originally there had been a fairly largestructure of the pit type located at that particular spot. After it wasabandoned the pit became filled with refuse and debris to the groundlevel and a second structure was built. The pit portion of the secondand later house was dug into part of the fill in the old structure. Theremaining portion of the newer pit was excavated in undisturbed soil.The floor level of the second house was higher than that of the original.The ventilator and practically all of the east half of the room inthe subsequent structure were in the fill of the first. The material inthe pit did not provide a solid face for the walls of the second struc-ture and it was necessary to reinforce them with stone slabs andwattlework.The existence of the later and smaller structure was discoveredwhile the workmen were cleaning the debris from the original house.The exploratory trench dug across the ridge at this point penetratedNo. 5a at its southeast corner. The workmen then proceeded to fol-low the wall to the floor and were cleaning out the accumulation whenthe ventilator and east wall of No. 5b were discovered. Unfortunatelythe wall of the second structure which was in the fill of the first couldnot be left standing since all of the material that backed it had beenremoved. Its presence and relation to both structures was noted andmeasured, however, and the combination of two houses is shown infigure 12.Because of the unexpected development of finding one pit housepartially overlying and within the confines of another it was notpossible to retain a pillar for stratigraphic study. The pillar leftat the start of the work was found to consist largely of the ventilatorof structure 5b and for that reason was of little value from the view-point of a study of the strata. Good stratigraphic evidence on theceramic sequence was furnished between the two floor levels. Pot-
ROBEKTS] ARCHEOLOGIOAL REMAINS IN ARIZONA 63sherds from No. 5a indicated an older, less developed ware than thosefrom the floor of 5b. The walls of the pit of 5a were very irregular.The structure had a tendency toward a rectangular shape but some
5 FEET
Figure 12.?Structures 5a and 5t). a, holes for main superstructure supports ; &, fire pits
;
c, stone embedded in floor ; 6,, basin in floor ; e, hole in floor ; f and g, storage holes ; li, ven-tilator shaft ; i, sand accumulation above old surface.portions of the walls were concave and others convex, so that each sideof the chamber was quite uneven. In its general features the struc-ture had been comparable to the others previously described. There
/.^ BUREAU OF AMERICAN ETHNOLOGY IBoll. 121was no bench and the superstructure had been supported by fourupright posts set in the floor at some distance from the corners.There was a fire pit near the center of the room (fig. 12, &). It waspeculiar in shape, two sides and a portion of the third being prac-tically straight. The fourth was a long curve. It really consistedof the fourth and part of the third sides combined. One end and aportion of one side had a facing of stone ; the rest of the pit was linedwith plaster. A stone was set in the floor, similar to those in No. 2, be-tween the fire pit and the northeast roof post (fig. 12, c) . Close tothis stone was an oval pit (fig. 12, d). A circular hole was placed inthe floor in front of the ventilator opening (fig. 12, e) and there weretwo additional small holes of the storage type at the south side of thechamber, (fig. 12, /, g). There was no sipapu, no ladder pit, and nodeflector. The floor between the fire pit and the ventilator openingwas broken, however, and it is possible that there had been a ventilatorstone that was pushed out during the reconstruction process associatedwith the erection of structure 5b. The stone in the floor between thefire pit and the northeast post probably served a similar purpose tothat suggested for the stones in No. 2. All indications were that thesmaller holes in the floor had served for storage purposes.Just what the purpose of the depression in the floor in front of theventilator opening (fig 12, e) may have been is debatable. A smallhole is frequently found in such a position in houses of this type.Some investigators have expressed the belief that they represent asipapu placed in a different position from that usually found. Insome of the structures at the Long H Ranch holes of this type wereobserved in which there were indications that timbers had been setthere." In discussing this feature in the report dealing with thosehouses it was suggested that a pole had been embedded in that posi-tion to hold a cover stone in front of the ventilator opening. In thepresent example, however, the depression was more like a basin thana post hole and there was no indication that a piece of wood had beenplaced in it. What other purpose it may have served was not in-dicated. That it functioned as a sipapu does not seem logical inas-much as that feature, if present at this site, was consistently in theportion of the floor between the fire pit and the wall opposite from theventilator.The ventilator in No. 5a was of the tunneled form with a shortpassage and rather large shaft. The exact limits of the passage andshaft could not be determined because there had been considerableweathering at that portion of the structure. They were approximatelyas shown in the drawing but may have varied a few inches from thesize indicated.
'^Roberts. 1931.
Roberts) ARCHEOLOGICAL REMAINS IN ARIZONA g5The interior of structure 5b was without extensive furnishings.The only different feature found was a slab-lined fire pit placed inthe fill above the old floor. The four holes near the corners of thestructure showed that the roof had been supported by four uprightposts and indicated that the covering of the pit was similar in formto those already described. The ventilator, because of its positionAvholly within the fill in 5a, was of the constructed type. The passagewas covered with poles and brush in order that the fill would notfall into it, and the shaft was reinforced with sticks and plaster.As will be seen from the diagram (fig. 12) No. 5b was smaller than5a. The exact relationship of the two structures, from the stand-point of the position of one lying within a part of the other, isclearly shown by the section at the lower half of figure 12. It willsuffice to say that the potsherds from the floor of 5b were Develop-mental Pueblo, while those from 5a were Modifled Basket Maker.This indicates that an interval of some length elapsed between theabandonment and filling of 5a and the digging of the pit and erec-tion of 5b, although in the general form of the two structures thereis little variation to be observed.On the fire pit, ventilator line, structure 5a measured 12 feet 1inch (3.683 m) from wall to wall and at right angles to that linethe room was 12 feet 6 inches (3.810 m) across. The wall height atthe ventilator side of the chamber was originally 3 feet 8 inches(1.117 m). With the accumulated sand the depth at the time ofexcavation was 5 feet 1 inch (1.549 m). The exact wall height atthe opposite side of the room could not be determined because it hadbeen dug away to make the pit for 5b. The point where pits 5a and5b coincided, however, near the southwest support post placement of5a, gave a wall height of 4 feet 9 inches (1.447 m). This probablyapproximates the original height for the west wall of 5a. The accu-mulation of sand over the remains of the structure made the depthfrom the ground level to the floor 6 feet 5 inches (1.955 m).When the debris was cleared from the pit the holes for the uprightsfor 5a showed that fairly heavy timbers had been employed to sup-port the superstructure. The hole at the northwest corner of theroom was practically circular in form with a diameter of 11 inches(27.94 cm). It was 1 foot (30.48 cm) from the wall and had a depthof 1 foot 6 inches (45.72 cm). The northeast hole was more ovalin outline and had diameters of 11 and 9 inches (27.94 and 22.86 cm).It was 1 foot 3 inches (38.1 cm) deep. The distance between the holeand the wall was 9 inches (22.86 cm). The southeast support-posthole was 11 inches (27.94 cm) in diameter and 1 foot 4 inches (40.64cm) deep. It was placed 11 inches (27.94 cm) from the wall. Thesouthwest-post hole was circular in form with a diameter of lOi^
gg BUREAU OF AMERICAN ETHNOLOGY [Boi^ 121inches (26.67 cm) and a depth of 1 foot 3 inches (38.1 cm). It waslocated 91/2 inches (24.13 cm) from the wall.The fire pit measured 1 foot 6 inches (45.72 cm) by 1 foot (30.48cm) and was 4 inches (10.16 cm) deep. The hole in the floor (fig.12, d) had diameters of 7 inches (17.78 cm) and 91/2 inches (24.13cm). It had a depth of 7 inches (17.78 cm). The basinlike depres-sion in front of the ventilator opening (fig. 12, e) had a diameterof 81/2 inches (21.59 cm) and a depth of 31/2 inches (8.89 cm). Theother two holes in the floor (fig. 12, /, g) had diameters of 4 inches(10.16 cm) and 51/2 inches (13.97 cm) and depths of 81/2 inches (21.59cm) and 9 inches (22.86 cm).The ventilator opening was approximately 1 foot (30.48 cm) high,and 1 foot 5 inches (43.18 cm) wide. Its approximate length was2 feet 8 inches (81.28 cm). The shaft was roughly circular in formwith a 3-foot 6-inch (1.066 m) diameter. The shaft had an origi-nal depth of 3 feet 3 inches (99.06 cm) and an accumulated depthof 4 feet 6 inches (1.371 m).The pit portion of structure 5b had a diameter on the fire pit,ventilator line, of approximately 10 feet (3.048 m). This measure-ment may have varied slightly since it was not possible to determineexactly where the interior surface of the wall had been. Acrossthe opposite direction the room was 11 feet 3 inches (3.429 m) fromside to side. The depth of the ventilator side can only be approxi-mated since a portion of the floor was removed before the measure-ments were taken. Eoughly the wall stood 3 feet 3 inches (99.06 cm)above the floor at the time of occupation. The later accumulationof sand on the surface increased this depth to 4 feet lOi/^ inches(1.485 m). At the opposite side of the chamber the original groundlevel was 4 feet 5 inches (1.346 m) above the floor. The driftedsand increased the depth to 6 feet 2 inches (1.879 m).The holes for the support posts were smaller than those in 5a.The northwest hole in 5b had a diameter of 6 inches (15.24 cm)and a depth of 1 foot (30.48 cm). The hole was 5 inches (12.7 cm)from the wall. The northeast post was set in a hole 6 by 71/2 inches(15.24 cm by 19.05 cm) in diameter. A portion of the depth ofthis hole had been in the fill above the floor of 5a, so that when themeasurement was taken only 7 inches (17.78 cm) of the total re-mained. This hole had been approximately 3 inches (7.62 cm)from the wall. The southeast post hole had diameters of 61/2 and11 inches (16.51 and 27.94 cm). It also had passed through the fillabove the floor of 5a and only 5 inches (12.7 cm) of the total depthremained. The hole had been approximately 9 inches (22.86 cm)from the wall. The hole at the southwest corner of the room meas-ured 7 inches (17.78 cm) and 814 inches (21.59 cm) in diameter. Itsdepth was 1 foot 6 inches (45.72 cm) and its distance from the waU
Roberts 1 ARCHEOLOGICAL REMAINS IN ARIZONA 671 foot 1 inch (33.02 cm). The fire pit measured 1 foot 3 inches(38.1 cm) by 1 foot 2l^ inches (36.83 cm). It had an average depthof 81/2 inches (21.59 cm).Inasmuch as any measurements which might be given for theventilator and ventilator shaft of 5b would be only a rough approxi-mation of the original figures they will not be included in the report.The diagram (fig. 12) shows very nearly the size of each and itsrelation to the rest of the structure and this will have to suffice sincethe exact data cannot be presented.
1
FiGDEE 13.?Surface structure B and associated pit remains.STRUCTDBES 6, 7, AND 8A short distance along the ridge southwest from structure No. 5were the remains of surface house B and a group of pits numbered6, 7, and 8. The remains of the semisubterranean structures formedone of the groups of three connected houses previously mentionedin the introduction (fig. 13). Two of them, 6 and 7, were actuallyparts of one large structure, as there was a fairly large opening be-tween them. The third, No. 8, except for a narrow slitlike openingin the wall between it and No. 7, was independent of the other two.This opening was large enough to pennit intercourse between thetwo chambers and it would have been possible to pass small objects
go BUREAU OF AMERICAN ETHNOLOGY [Bull. 121from one to the other through it. The aperture was too small, how-ever, to have permitted a person, or even a small child, to pass fromone room to another. Anyone desiring to do so would have foundit necessary to go outside by means of the smoke-hole hatchway fromone and enter the other through the corresponding entrance. Thestructures, individually, corresponded closely to the general type ofpit structure. Each constituted a unit and will be discussed as such.Structure No. 6 had a roughly rectangTilar pit with slightly curvedwalls and rounded corners. The wall at the ventilator side of thechamber was more irregular than the other three. This was probablydue to the fact that the original aperture had been blocked and anew opening made. Like the structure previously described. No. 6had no bench. The sides of the pit extended perpendicularly from thefloor to the ground level. There were four holes near the cornersof the chamber, indicating that the pit portion had a superstructurenot unlike those previously described.The interior features were simple. There was a center fire pit, aladder pit, sipapu, and storage hole in the floor (pi. 9, a). Betweenthe ladder pit and the eastern support post a large slab of stone hadbeen set in the floor to form a binlike compartment at that side ofthe chamber. In this respect the structure corresponded to No. 1where, it will be recalled, there was a similar inclosure, although onethat was more elaborately made.No. 6 contained one feature which was not present in the structurespreviously described, namely, a niche in the wall opposite the venti-lator side of the chamber (fig. 14, h). The location of the niche inthe wall a short distance above the floor suggests the Katcina Kihufound in kivas of the more fully developed Pueblo periods. TheKatcina niches were usually of two shapes, cylindrical and rectan-gular. The example in structure 6 was not of these forms. The sec-tion through the pit (fig. 14) shows the shape. The evidence wasthat while it was in the position usually occupied by the KatcinaKihu, in which ceremonial paraphernalia was usually kept, it shouldnot be considered in the same classification because indications werethat a brace for the superstructure had been placed there. It isprobable that the main superstructure framework began to sag towardthat wall and in order to check the tendency to collapse a brace wassupplied and the niche h was the place where it had rested. Theniche was not the only provision made for bracing the superstructure.In the wall at the east corner there was another hole where a bracehad been set (pi. 9, a). The latter had a decided slant and sug-gested that a timber with a forked end had been embedded there.The crotch being placed against the upright at about the height ofthe main stringers would have braced the framework in an efficientmanner.
Roberts] ARCHEOLOGICAL REMAINS IN ARIZONA gQThe fire pit in No. 6 was roughly circular in form (fig. 14, /).It had been dug through the floor mto the undisturbed soil be-neath and the sides were plastered with adobe. Around its edges,on the floor, was a plaster ridge similar to the one discussed forstructure No. 3. It differed from the latter, however, in that therim did not include the ladder pit. The ladder pit was a shallowdepression with carefully plastered sides and bottom. A small stoneslab was set in it close to the edge on the ventilator si< le in such away that the base of the ladder might rest against it and not marthe wall. The nature of this object is illustrated in the drawing(fig. 14, g). The stone may have served also as a step for personsusing the ladder.One of the interesting features in connection with No. 6 was thatof the ventilator. As mentioned before, for some reason or otherthe original passage was blocked and a new opening into the cham-ber provided. The relationship of the new aperture and passageto the old is shown in the drawing (fig. 14) . As a result of the con-struction of the passage that part of the ventilator was lengthened.Wliat advantage a longer passage might have had over a shorterone is not known, but the builders must have had good reason formaking the change. The aperture from the chamber passage wasunique as far as the structures at this site were concerned, becauseit had a stone frame for the openmg (pi. 9, a) . The stone employedm this capacity was a metate which had been used until the troughportion became very tliin. A rectangular hole with rounded cor-ners was then pecked through the grooved portion of the stone andthe ventilator opening thus provided. When the debris was removedfrom the interior of the pit a carefully worked cover stone wasfound in position over the opening. The use of stone as an aper-ture frame is not unknown in the Southwest, although examplesof it are not common. In one of the kivas excavated at a site inNutria Canyon, northwest from the Pueblo of Zufii, New Mexico, asomewhat similar treatment of the ventilator opening was observed.In this case, however, the stone used was a large flat slab and notan old metate. Furthermore, the vent frame was cut in the shapeof a horseshoe rather than in the form observed in house 6.^*Perforated stone slabs have been found at various sites, particu-larly in the Little Colorado region, but in most cases there has beenlittle to indicate what their function may have been. Fewkes insome of his reports has suggested that they may have been built intowalls of rooms to partially close the passageway, but he found noexample in position. Most of those that he recovered were lying
' Roberts, 1932, p. 73, pi. 13, a.68764?39 6
70 BUREAU OF AMERICAN ETHNOLOGY [BDLL. 121over burials. The explanation, or rather suggestion, made m thisconnection was that the hole was for the escape of the soul or breathbody.?? Hough during his investigations in the Milky Hollow dis-trict of Arizona, east of the Petrified Forest and southwest from the
1=3 cS-^"
FiGUEH 14.?Structure 6. a, holes for main roof supports ; 6, placement for roof brace
;
c, storage hole ; d, sipapu ; e, fire pit in fill ; f, fire pit for 6 ; g, ladder pit ; h, stone form-ing bin ; i, ventilator stone ; j, ventilator shaft ; k, remains of granary ; m, sand accumu-lation above old surface.present location, discovered perforated slabs that were used as topson fire pits.^? Just north of the region where Hough made his dis-covery a group of students from Colorado College found similar
??Fewkes, 1904, pp. 106, 160-102."Hough, 1903, p. 230.
Roberts] ARCHEOLOGICAL REMAINS IN ARIZONA 71specimens. During the course of his work at the old Zuni Village ofHawikuh, Hodge conducted some excavations in a small pre-Hawikuhsite nearby and uncovered a perforated stone which was in positionover the fire pit in a dwelling room. A similar find was also madein one of the early Hawikuh dwellings.*^ The preparation of suchstones for use over fire pits may have suggested the use of similarobjects as a frame for the ventilator opening in houses. The ideaseems plausible, in the light of present knowledge on the subject,since the former have been found in larger numbers than the latter.One curious feature about the ventilator for No. 6 was that theshaft at the outer end had an unusually large opening. In fact, theline measurement for the shaft was practically the same as the lengthof the original passage; the second one had been somewhat longer.Furthermore, this shaft was the only one in the whole group ofstructures excavated which had a rectangular form. Because of itssize it suggests that its original purpose may have been that of anactual entryway rather than merely that of a ventilator. However,the shaft and tunnel could not have functioned as an entrance inthe condition found when the chamber was excavated. The openingin the stone which framed the aperture was too small to permit thepassage of a person other than a very small child. On the otherhand, the original tunnel was large enough to have permitted its useas an entrance. Structure No. 6 may be an example of a dwellingoccupied at the time when the change from entrance to ventilator wasmade and the means of gaining access to the house was throughthe smoke hole and down a ladder. Evidence from No. 6 suggeststhat the structure was built and occupied at about the stage whenModified Basket Maker was giving way to Developmental Pueblo.In view of this it would be quite possible for the ventilator changesto have been made as a part of the transition accompanying thereadjustments occurring at that time.In the fill at structure 6 was a fire pit dating from a later stagein the occupancy of the site (fig. 14, e; pi. 9, h). As far as couldbe determined from the general position of this fire pit, and the ma-terial in the fill, it belonged to the stage represented by surface houseB. There was nothing unusual in any way about the pit itself.It was oval in outline, the sides were faced with stone slabs, andthere was stone paving on the bottom. The pit was filled withashes and charcoal and the stones gave evidence of considerableburning, so that it is probable that fires were lighted in it over afairly long interval.The pillar left in this structure for stratigraphic study showed thatfor an appreciable period after it had been abandoned and gone into
? Hodge, 1923, pi. xvi, p. 26.
-2 BUREAU OF AMERICAN ETHNOLOGY [Bull. 121decay there was little activity in its immediate vicinity. The firstWo feet (45.72 cm) of fill above the floor was clean sand. Fromthat point to the original surface the depression had been used asa dump and the material was typically refuse-mound deposit. Itcontained a large amount of ashes, bits of charcoal, broken bones,potsherds, and stone chips. From the top of the secondary fire pit tothe present ground level the fill was largely clean sand, probablyblown over the house after that portion of the site was abandoned.Structure No. 6 measured 10 feet 6 inches (3.2 m) from wall towall on the line through the sipapu, fire pit, and ladder pit. Acrossthe opposite way it was 10 feet 8 inches (3.251 m) from wall to wall.The depth showed more variation than the foregoing measurements.At the ventilator side of the chamber the original ground level was2 feet 8 inches (81.28 cm) above the floor level. Due to an accu-mulation of drift sand the depth at the time of excavation was 4 feet2 inches (1.27 m). At the opposite side from the ventilator the wallfrom the floor to the original ground level measured 4 feet 31/2 inches(1.308 m). The accumulation of drift sand added 1 foot 2 inches(35.56 cm) to this measurement, making a total of 5 feet 51/2 inches(1.663 m) for the depth at the time of excavation.The holes for the support posts showed some variation in size.That at the north corner of the room had a diameter of 1 foot 1 inch(33.02 cm). Its depth was 1 foot 7 inches (48.26 cm). It was lOiAinches (26.67 cm) from the wall. A smaller post was used at the eastcorner since the diameter of the hole where it was placed was only 9inches (22.86 cm). The hole had a depth of 1 foot 3 inches (38.1 cm)and was 41/0 inches (11.43 cm) from the wall. The post at thesouth corner of the room was set against the wall of the chamber.The hole in which it stood measured 1 foot I1/2 inches (34.29 cm)on one diameter and 1 foot 3 inches (38.1 cm) on another. Thehole had a depth of 1 foot 4 inches (40.64 cm). The post had notcompletely filled the hole as a stone slab was set in it to help brace thetimber. The hole at the west corner of the room had a diameter of 1foot 1 inch (33.02 cm), a depth of 2 feet II/2 inches (64.77 cm), andwas 6 inches (15.24 cm) from the wall.The fire pit (fig. 14, /) had a diameter of 1 foot 61/2 inches (46.99cm) on the ventilator line and 1 foot 6 inches (45.72 cm) at rightangles to it. The depth was 31^ inches (8.89 cm). This measure-ment includes the adobe rim which had a height of II/2 inches (3.81cm) above the floor. The rim ranged from 3 to 4 inches (7.62 to10.16 cm) in width.The ladder pit (fig. 14, g) was 10 inches (25.4 cm) wide and 1 foot8 inches (50.8 cm) long. It had a depth on the fire pit side of II/2inches (3.81 cm) which was provided in part by the plaster rimi
Roberts] ARCHEOLOGICAL REMAINS IN ARIZONA 73of the fire pit. At the opposite side, from the bottom of the pit tothe floor level was 5 inches (12.7 cm). The small stone set in thefloor of the ladder pit was 3 inches (7.62 cm) high, 8 inches (20.32cm) long, and II/4 inches (3.175 cm) thick.The sipapu (fig. 14, d) was 1 foot 4 inches (40.64 cm) from theedge of the fire-pit rim. The hole had diameters of 3% inches(8.89 cm) and 2I/2 inches (6.35 cm). Its depth was 6 inches (15.24cm) . The storage hole, or perchance a second sipapu (fig. 14, c) , was4I/2 inches (11.43 cm) from the sipapu. It had a diameter of 4inches (10.16 cm) and a depth of 5I/2 inches (13.97 cm). Tliis secondhole was 2 feet 10 inches (86.36 cm) from the wall. The niche inthe wall (fig. 14, h) measured 6 inches (15.24 cm) in width and 1foot 7 inches (48.26 cm) from top to bottom along the wall. At thetop it was 31/^ inches (8.89 cm) deep. Midway between top and bot-tom it was 4 inches (10.16 cm) deep and at the bottom measured 8inches (20.32 cm) from front to back. The bottom of the niche was1 foot 4 inches (40.64 cm) above the floor. From the top of the nicheto the original ground level was 1 foot 4i/^ inches (41.91 cm).The ventilator stone (fig. 14, i) was 2% inches (6.985 cm) thickat the bottom and l^ inch (0.635 cm) at the top. It was 13 inches(33.02 cm) wide at the bottom and top as it stood in the wall. Theend embedded in the floor, however, tapered to a width of 8^/4 inches(20.955 cm) at the bottom. The total height of the stone was 22inches (55.88 cm), but only 16 inches (40.64 cm) rose above thefloor when the stone was in position. The opening measured 8 by 10inches (20.32 by 25.4 cm). The passage was 3 feet 9l^ inches (1.156m) long. It had a width of 1 foot 6 inches (45.72 cm) where itentered the shaft. At the room end it was 1 foot 2 inches (35.56 cm)liigh, and at the shaft 1 foot 1 inch (33.02 cm). The shaft measured3 feet 5 inches (1.041 m) by 1 foot 5 inches (43.18 cm). Its originaldepth at the passage side was 1 foot 11 inches (58.42 cm), but withthe sand accumulation measured 3 feet IOI/2 inches (1.181 m) fromground level to the floor. At the back side of the shaft the originaldepth was 9 inches (22.86 cm) , but at the time of excavation measured3 feet 2 inches (96.52 cm) from top to bottom. The original passagehad a length of 3 feet 9 inches (1.143 m). At the aperture end itswidth was 1 foot 1 inch (33.02 cm), and where it entered the shaft1 foot 6 inches (45.72 cm).The fire pit built in the fill which accumulated in the roomafter it had been abandoned (fig. 14, e) was 2 feet 3 inches (68.58 cm)above the floor. The pit had diameters of 2 feet 2 inches (66.04 cm)and 1 foot 714 inches (49.53 cm). Its depth averaged 1 foot 1 inch(33.02 cm).Structure No. 7 was more irregular in shape than most of thoseexcavated during the season's investigations. This was in part due
74 BUREAU OF AMERICAN ETHNOLOGY [Bull. 121to the fact that a segment of the north wall was cut by a portion ofNo. 6 which overlapped No. 7 at that place. The pit was dug into theground and had no bench. Four posts were used to support thesuperstructure, as in most of other houses. One of these posts, how-ever, that at the north corner of the room, was not actually in No. 7but instead was placed in the floor of No. 6. The interior features forstructure 7 were quite simple. There was a sipapu and a fire pit
5FEET
I .,j i 1
Figure 15.?Structure 7. a, holes for main support posts ; h, placement for brace ; c, stor-age hole ; d, sipapu ; e, fire pit ; t> ventilator shaft ; g, opening into 8 ; h, accumulationabove original surface.near the center of the chamber and at the north side of the fire pit asmall storage hole (fig. 15).The fire pit differed from those previously described in that itwas both built up and dug into the floor (fig. 15, e). The floor levelat the ventilator side of the chamber was somewhat higher than thatthroughout the remainder of the room. Consequently about half ofthe fire pit was dug into the higher floor level, while^ the other halfwas formed by a plaster ridge which encircled the pit. Only two
Roberts] ARCHEOLOGICAL REMAINS IN ARIZONA 75stones were used in its construction. One was set in the adobe rimat the north side of the pit, its top flush with that of the rim. Theother was placed at the ventilator side in such a position that itprobably served as a deflector. Tliis was the only example of a de-flector stone actually incorporated in and forming a part of the firepit.The stone did not rise as high above the floor as in some of the otherstructures, but it was of sufficient height to have served such a pur-pose. There was no ladder pit, although the floor between the fire-pit rim and the ventilator opening was abraded to a degree which indi-cated that the base of a ladder had rested there.The sipapu in this structure was almost rectangular in shape, aform not usually observed (fig. 15, d). The sides were practicallystraight, but the corners were rounded sufficiently to suggest the ovalform. The inside and bottom were carefully plastered. Between thefire pit and the east support was a hole in the floor that was oval inoutline (fig. 15, <?). It contained no evidence to show what its purposewas. As in the case of similar features in other floors it is possiblethat it served as a depository for small objects.A niche at one side of the chamber (fig. 15, h) was very much likethat described for structure 6, the main difference between the twobeing that this example was not at the center of the wall but wasplaced quite close to the west support post. Its location and generalform are shown in the plan and section (fig. 15, 6). Like the previousone, it probably was the placement for a roof brace.The slitlike opening in the wall between structures 7 and 8 startedjust above the floor level and extended almost to the top of the wall.The opening penetrated the partition obliquely as is shown in figure15, g. The sides of the slit were carefully plastered, an indication thatthe feature was intentional and not an accidental break in the narrowsection of earth between the two chambers.The ventilator in No. 7 consisted of a simple tunnel and shaft. Thepassage had actually been tunneled through the earth and was not ofthe constructed type. The aperture was faced with plaster andthe walls of the passage were treated in the same fashion. A singleblock of stone resting on the floor of the chamber served as a sill forthe opening. The shaft at the outer end was oval in form.Structure No. 7 measured 8 feet 1 inch (2.463 m) from wall to wallon the fire pit, ventilator line. From the point on the north wallwhere it coincided with No. 6 to the opposite side of the chamber was9 feet 11/^ inches (2.780 m). The wall at the ventilator side measured2 feet 4 inches (71.12 cm) from the floor to the old ground level. Thepresent ground level is 3 feet 3 inches (99.06 cm) above the bottom ofthe pit. At the opposite side of the room the floor was 2 feet 7 inches
yg BUREAU OF AMERICAN ETHNOLOGY [Boll. 121(78.74 cm) below the old ground level. From the present groundlevel to the floor of the pit was 4 feet 1 inch (1.244 m).The north support post was set in the floor of No. 6 and touchedthe edge of the floor of No. 7. The latter was 1 foot 3 inches (38.1cm) higher than that of No. 6. The hole for the support post had adiameter of 71/2 inches (19.05 cm). It was not very deep, however,measuring only 6 inches (15.24 cm) from the floor level to the bottom.It would seem that the builders had depended in part on the wall ofthe step between the two chambers for additional support. The holefor the support post at the east side had a diameter of 9 inches (22.86cm) and a depth of 1 foot (30.48 cm). It was only 1 inch (2.54 cm)from the wall. The hole for the south support was broken out alongone edge so that it had a long diameter of 1 foot (30.48 cm) and a shortdiameter of 7 inches (17.78 cm). Indications were, however, that ithad originally been circular in form with an average diameter of 7niches (17.78 cm). This post must have slanted somewhat, becausethe wall of the pit slightly overhung the hole and the top of the wallextended into the room 31/2 inches (8.89 cm) beyond a vertical linefrom the point where the floor and bottom of the wall met. The holefor the post was 11/2 inches (3.81 cm) from the wall at the floor level.It had a depth of 1 foot 2 mches (35.56 cm). The west support postseems to have been the smallest of the group because the remaininghole had diameters of only 5 inches (12.7 cm). It had a depth of 1foot (30.48 cm). As in the case of the other support posts, this onewas very close to the wall; there was only 1 inch (2.54 cm) betweenit and the edge of the hole.The fire pit (fig. 15, e) measured 1 foot 2I/2 inches (36.83 cm) by111/^ inches (29.21 cm). It had a depth at the ventilator side of thechamber, where it was dug into the floor, of 6 inches (15.24 cm). Atthe opposite side, where it depended entirely upon the adobe rimfor its depth, the measurement was 21/2 inches (6.35 cm). The rimaround the pit ranged from 5I/2 inches (13.97 cm) to 10 inches (25.4cm) in width. At its inside edge, the highest part of the rim, it hadan average height of 21/4 inches (6.35 cm) . Just back of the stone slab,set on edge in the position of a deflector, the height of the rim was 31/2inches (8.89 cm). The stone was 1 foot 5 inches (43.18 cm) long, 11/4inches (3.81 cm) thick, and rose 6 inches (15.24 cm) above the floor.The top of the stone was 9 inches (22.86 cm) above the bottom ofthe fire pit.The sipapu (fig. 15, d) was 11 inches (27.94 cm) from the edge ofthe fire-pit rim. The hole measured 8 inches (20.32 cm) by 6 inches(15.24 cm) and had a depth of 51/2 inches (13.97 cm).The storage hole at the northeast corner of the room (fig. 15, c) was101/2 inches (26.67 cm) from the fire-pit rim and 11 inches (27.94 cm)from the east support post. This hole had diameters of 9 inches (22.86
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA 77cm) and 6i/^ inches (16.51 cm), and had a depth of 3i/^ inches(8.89 cm).The niche near the west support post (fig. 15, h) was 6 inches (15.24cm) wide and 1 foot 9^2 inches (54.61 cm) high. At the top of thewall it was 2i/^ inches (6.35 cm) deep, and at the bottom the back was6 inches (15.24 cm) from the face of the wall. It was 9 inches (22.86cm) above the floor. The top was on the line of the old ground level.The aperture between Nos. 7 and 8 (fig. 15, g) was 8 inches(20.32 cm) Avide and 2 feet 8 inches (81.28 cm) high. The wall betweenthe two chambers at that point was 7% inches (19.68 cm) thick. Theopening between 6 and 7 at the opposite side of the room was 4 feet5 inches (1.346 m) wide.The ventilator opening was 1 foot 2 inches (35.56 cm) wide at thebottom and narrowed down to 9 inches (22.86 cm) at the top. Theaperture was 1 foot l^ inch (31.75 cm) high. The bottom of thepassage was 3i/2 inches (8.89 cm) above the floor. The stone whichformed a sill for the aperture was 3 inches (7.62 cm) high, 2% inches(6.35 cm) wide, and 1 foot 2 inches (35.56 cm) long.The passage was 2 feet 11 inches (89.9 cm) long. Where it openedinto the shaft it was 2 feet (60.96 cm) wide and had a height of 1 foot1 inch (33.02 cm) . The shaft was oval in form with diameters of 2 feet(60.96 cm) and 1 foot 5 inches (43.18 cm). Its original depth was 1foot 11 inches (58.42 cm) at the passage side and 1 foot 10 inches(55.88 cm) at the back. At the time when the structure was excavatedthis depth had been augmented by a 9i/2-inch (24.13-cm) layer ofblow sand.The third structure in this group. No. 8, was typical of the pithouses at this location. The subterranean portion was dug intothe earth and no bench provided. Three of the sides were fairlysymmetrical arcs but the fourth was very irregular and undulating(fig. 16). This was the ventilator side of the chamber and it is pos-sible that the construction of this feature may in part have been re-sponsible for the unevenness of the wall. The pit for this struc-ture was somewhat shallower than in the case of some of thosepreviously described. The tendency of the people in the earliersouthwestern horizons to construct their pit houses on a slope iswell illustrated in the case of the remains of No. 8. The ventilatorside of the structure was much lower than the 'other and any surfacewater would have drained away from the house quite rapidly.This was also true of structures 6 and 7 and the sections in the sev-eral drawings illustrate this point better than could any extendeddescriptions.The covering over the pit of No. 8 was supported by four mainposts set in the floor close to the corners of the chamber. The roofprobably was of the flat ceiling, sloping sides type previously de-
yg BUREAU OF AMERICAN ETHNOLOGY fBtJLL. 121scribed. The interior features were simple indeed, as they consistedsolely of a sipapu and a fire pit. The only other item of consequencewas the opening in the wall between this structure and house 7. ^The fire pit was merely a depression in the floor with plaster sidesand bottom (fig. 16, d). The sipapu was a roughly circular hole inthe floor close to the fire pit (fig. 16, c). The sides and bottom ofthis hole were also plastered. The ventilator was of the simplestform. The passage was tunneled through the earth to the oval-shaped shaft at the outer end (fig. 16, e). The aperture in thechamber wall was larger than the remaining portions of the passage.The ceiling of the latter was practically horizontal but the floor hada decided upward slant to the bottom of the shaft. The ventilatoras a whole was not large enough to have functioned in the capacityof an entrance and must have served only to bring fresh air intothe chamber.Structure 8 measured 9 feet 8 inches (2.946 m) from wall to wallon the fire pit, ventilator line. The diameter at right angles tothe latter measurement was 8 feet 10 inches (2.692 m) . At the venti-lator side of the chamber the wall was 1 foot lll^ inches (59,69 cm)high. Above the original ground level was an accumulation of sandwhich increased the depth of the pit at the time of excavation to 3feet 41/^ inches (1.028 m). At the opposite side of the chamber theoriginal wall height was 3 feet 2 inches (96.52 cm). This was in-creased to 4 feet 4 inches (1.320 m) by the accumulation of sandthat drifted over the site after the abandonment of the structure.The hole for a support post near the north corner of the room haddiameters of 7i/^ inches (19.05 cm) and 91/2 inches (24.13 cm). Thehole was 1 inch (2.54 cm) from the wall and had a depth of 91/2inches (24.13 cm). The hole for the east post measured 9i/^ inches(24.13 cm) by IOI/2 inches (26.67 cm). Its depth was 1 foot 1 inch(33.02 cm). The hole was placed 13^ inches (4.44 cm) from the wall.The hole for the south post had diameters of 9 inches (22.86 cm) and8I/0 inches (21.59 cm) . Its depth was 11 inches (27.94 cm) . The holeactually touched the wall. The west-support hole was 8 and 71/2inches (20.32 and 19.05 cm) in diameter. Its depth was 1 foot (30.48cm). As in the case of the hole at the south corner, this one alsotouched the wall.The fire pit (fig. 16, d), with diameters of 1 foot 3 inches (38.1cm) and 1 foot 7 inches (48.26 cm), was quite shallow, having adepth of only 4 inches (10.16 cm).The sipapu (fig. 16 c) was 7 inches (17.78 cm) from the fire pitand 3 feet 11 1/0 inches (1.206 m) from the west wall of the chamber.The hole had diameters of 3 and 21/2 inches (7.62 and 6.35 cm). Itwas 5 inches (12.7 cm) deep.
Roberts ] ARCHEOLOGICAL REMAINS IN ARIZONA 79The slitlike opening in the wall between structures 7 and 8 waslarger in No. 8 than it was in No. 7. The width at the side of thewall was 9 inches (22.86 cm) and the height 2 feet 10 inches (86.36cm). The greater height was made possible because of the factthat the floor of room 8 was 3i/^ inches (8.89 cm) lower than thatin room 7.The ventilator opening was 1 foot 1 inch (33.02 cm) wide and 1foot 11/2 inches (34.29 cm) high. The sill of the opening was 2^^
"_::?
5FEET
Figure 16.?Structure 8. a, holes for main supports ; 6, opening into 7 ; c, sipapu ; rf, firepit ; e, ventilator shaft ; t> recent accumulation of sand above original surface.inches (6.35 cm) above the floor. The passage was 2 feet 8^inches (69.85 cm) long and where it entered the shaft was 1 foot 3inches (38.1 cm) and 1 foot l^ inch (31.75 cm). At the passageside its original depth was 1 foot 4I/2 inches (41.91 cm), which wasincreased to 2 feet 11 inches (88.9 cm) by the sand drift. At theback of the shaft the original depth was 1 foot 4 inches (40.64 cm)and that at the time of excavation 2 feet lOi/^ inches (87.63 cmh
80 BUREAU OF AMERICAN ETHNOLOGY [Bull. 121In the preceding paragraphs the three structures forming thisparticular group have been discussed as single units. They shouldbe considered, however, from the standpoint of the series as a wholeand the relationship existing between them. No. 8 stood suffi-ciently apart from the other two to constitute a single structure.Nos. 6 and 7 were so closely connected that, to all outward appear-ances, at the time when they were occupied they must have seemedto be a single building.The coverings over these two pits were predominantly of the flatceiling, sloping side wall type of roof. However, the section wherethe two houses opened into each other could not have had the slant-ing poles and in all probability the space where the opening occurredmust have had a flat roof. This could have been provided throughthe use of timbers extending from one rectangular framework to the
uj__l ! I L. 10FEETFigure 17.?Section through 6, 7, and 8. a, sand accumulation above original surface:&, opening between 7 and 8 ; c, fire pit built on fill in structure 6.
other, forming a continuous flat ceiling over that part of the pits.The sides could have been closed with no difficulty by using thecustomary sloping poles. The earth wall between Nos. 7 and 8 wassufficiently thick to have supported sloping timbers for each of theroofs. If the poles had been placed alternately one sloping to theframework in No. 7, the next to that of No. 8, etc., the sides for thecoverings over the two pits would have been of the usual type andquite satisfactory. They would have met at the bottom in the formof a V which, if properly plastered, would have functioned as agutter for drainage purposes.Despite the large opening between No. 6 and No. 7 each chamberseems to have functioned as a separate unit because the evidencewas that each had had its own smoke hole and ladder. When thepits were first being excavated it was thought that either No. 6or No. 7 was the older and had been abandoned before the otherhouse was erected. All of the evidence, however, pointed to a con-temporaneity between the two. The fill extended from one throughthe opening into the other and fragments from the same potteryvessels were found on the floors of both. Consequently it was con-
itoBERTsl ARCHBOLOGICAL remains in ARIZONA gleluded that they must have been occupied at the same time. Ifmore privacy than such an arrangement permitted had been desiredit would have been an easy matter to hang a blanket or erect a brushscreen in the opening, but there was nothing to show that any typeof partition had been placed there.The relationship between the floor levels of the three structures,6, 7, and 8, is shown by the section (fig. 17). The drawing makesclear the marked cliange in level between Nos. 6 and 7 and showsthat No. 8 was slightly lower than No. 7, although not in as pro-nounced a degree as in the case of the other structure. What thereason may have been for the break in levels between 6 and 7 is notknown. Possibly the step was in the nature of a property line,inasmuch as there was no partition wall between the two chambers.It would have served to designate where one room left off and theother began, even though one was not completely separated fromthe other. A similar break in floor levels occurred between two ofthe structures in the group of three united rooms, discussed in fol-lowing pages, and because of its presence there it is thought thatthere was some specific significance to the feature, even though nosatisfactory explanation for it is forthcoming.STRUCTURES 9, 10, AND 11The group of three structures Nos. 9, 10, and 11 was located ashort distance, 18 feet (5.486 m), east of structure No. 6 (fig. 13).This series of three was particularly interesting because while eachwas an individual unit complete in itself there were no partitionwalls between them, and all three formed one large structure. Asin the description of the preceding group, each of the structures willbe considered separately and then the group will be discussed asa whole.Structure No. 9 was roughly D -shaped in outline, although oneside was quite uneven (fig. 18). As noted in some of the otherstructures, the wall that was most irregular was the one where theventilator occurred. The pit was dug into the native earth withpractically vertical sides. There was no bench encircling it. Theexcavation was somewhat deeper than the subterranean portion ofsome of the other structures at the site, but in its general aspectswas quite similar to them. The superstructure had been supportedby four upright posts and in general seems to have been of the pre-vailing type. One side of the roof must have been an exception, be-cause of the fact that there was no wall between this structureand the next, through the omission of the sloping poles. Timberslying flat and extending from the rectangular framework on topof the main support posts in this house to that in the next probably
82 BUREAU OF AMERICAN ETHNOLOGY [Bull. 121covered the space at the south side, in the same manner as describedfor Nos. 6 and T. This construction will be discussed more fully,however, in the consideration of the three structures as a singlebuilding.The interior features in No. 9 were somewhat more elaborate than
; k !
FiGOEE 18.?Structure 9. a, holes for main support posts ; ft, sipapu ; c, fire pit ; d, fallenstone from compartment wall ; e, holes for small poles ; f, adobe plaster placement forcompartment wall slabs ; g, stones in compartment wall ; h, location of support post forsuperstructure over 10 ; i, storage hole ; j, ventilator opening and sill ; k, ventilator shaft
;
I, placement for roof brace ; m, sand accumulation above old surface.those in the three structures forming the group described in preced-ing pages. Near the center of the room was a slab-lined fire pit.Between the fire pit and the ventilator a group of upright stone slabswere arranged in such a manner as to suggest the former existenceof a compartment similar to the one in structure No. 1. All of the
Roberts] ARCHEOLOGICAL REMAINS IN ARIZONA ggstones were not present when the house was excavated but a few ofthem were still in position and indications were that the low wallIiad extended from one support post to the other at that side of thechamber (pi. 10, a). There were several holes and depressions scat-tered about in the floor. One of these was the sipapu, placed aboutmidway between the fire pit and the west wall of the chamber. An-other was probably for storage purposes and two additional onesgave evidence of having had posts set in them.The fire pit was not elaborately constructed (fig. 18, c). A holeof sufficient size was dug into the floor and the interior walls onthree sides were lined with stone. The fourth side toward the ven-tilator opening was merely faced with a heavy coating of plaster.The bottom of the pit was also plastered. There was no ladder pit,although a small opening was present in the compartment wall be-tween the fire pit and the ventilator aperture and there were two shal-low depressions where the ladder ends probably rested. The openingin the compartment was flanked on either side by upright stonesprobably set in the floor, in the positions found, to protect the endsof the low wall from the feet of people ascending and descendingthe ladder.Evidence that small poles were set in the floor in the main part ofthe chamber about midway of each partition wall (fig. 18, e) sug-gests that in structure No. 9 there was a similar feature to the up-right stones set in approximately the same positions in No. 2. Aswas stated in the discussion of the latter, however, their purpose isnot known. It is possible that they had some connection with thefunction of the binlike compartment at that side of the chamber, butwhat the relationship may have been is a problem still to be solved.In the case of No. 9 the small poles may have been incased in plasterand thus have presented about the same appearance as the stone ex-amples. The compartment walls were not only of stone (fig. 18,/, p') , but in addition a large amount of adobe plaster was used andthe intervening spaces between the slabs was completely filled withthat substance.There was a niche in the wall of structure No. 9 like those de-scribed for Nos. 6 and 7. The niche was in the southwest corner, justback of the support post (fig. 18, I), and suggested that here also itwas necessary to supply a brace to counteract a shifting of the roofframework. The niche in this case was higher in the wall than in theother chambers and more cylindrical in form. It appeared to havebeen dug in a shape that would conform to the contours of thetimber.The ventilator for No. 9 had a tunneled passage. The block ofearth above the horizontal portion of the ventilator was of sufficientsize and compactness to hold its form without extensive reinforce-
g^ BUREAU OF AMERICAN ETHNOLOGY I Bull. 121ment and the pit was at a depth that would have required a consid-erable amount of digging had the passage been constructed by thetrench and cover method. In view of this it would appear that thebuilders adopted the easiest method of procedure and simply tun-neled a passage. The aperture had a stone sill (fig. 18, j) and alintel of the same material. The sides of the opening were carefullyplastered (pi. 10, a) at the outer end of the ventilator passage. Theexact nature and size of the shaft was not determined because a largecedar tree was growing at approximately the place where it oc-curred and the information was not deemed of sufficient importanceto warrant the destruction of the tree. The approximate location ofthe shaft was determined by cleaning out the passage and removingas much as possible of the debris that had collected in it. Wliileexact measurements could not be obtained in this fashion the approxi-mation was enough to suffice.Only one pillar was left in removing the fill from the three struc-tures. This was in the center of No. 10 and since the fill was thesame in all three pits the material in the pillar will be discussed inconnection with all of the houses as a single building.Structure No. 9 had a diameter of 11 feet 1 inch (3.378 m) on thefire pit, ventilator line. Across the opposite way the distance fromthe north wall to a line drawn across between the two points thatapproximately designated the division between 9 and 10, was 11 feet9 inches (3.581 m). At the ventilator side the original wall rose3 feet 1114 inches (1.206 m) above the floor. Above this there wasan accumulation of 10 inches (25.4 cm) of sand which increased thedepth to 4 feet 9I/2 inches (1.460 m) . At the opposite side of the cham-ber the wall was 3 feet 914 inches (1.155 m) from the old groundlevel to the floor. The sand accumulation increased the depth ofthe pit to 5 feet Qy^ inches (1.689 m) at the time of exacavation.The poles for the support posts in this structure had a greatersize variation than in some of the other chambers. That at thenorthwest corner of the room measured 8 inches (20.32 cm) and 9inches (22.86 cm) in diameter. It had a depth of 1 foot (30.48 cm)and was 10 inches (25.4 cm) from the wall. The hole at the north-east corner had diameters of 61/2 and 7 inches (16.51 and 17.78 cm).It was 91/^ inches (24.13 cm) deep and was placed 10 inches (25.4 cm)from the wall. The diameters of the southeast hole were 61/2 and 7inches (16.51 and 17.78 cm). The hole was 4 inches (10.16 cm) fromthe wall and 1 foot 4 inches (40.64 cm) deep. The hole at the south-west corner was broken out on the floor level but near the bottom itwas possible to obtain the diameters, which were 10 inches (25.4 cm)and 11 inches (27.94 cm). The hole had a depth of 1 foot 1 inch(33.02 cm) and it was 1 foot 6 inches (45.72 cm) from the wall.
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA ?5The fire pit (fig. 18, c) measured 1 foot 71/2 inches (49.53 cm) onthe ventilator line and 1 foot 614 inches (46.99 cm) in the oppositedirection. The pit was 5 inches (12.7 cm) deep. The slabs whichlined the three sides projected above the floor level from ^ inch(6.35 mm) to 2 inches (5.08 cm). The probabilities are that whenthe structure was occupied the stones were covered with plaster andthe discrepancies in their heights were compensated for by the plasterand the rim around the pit was practically level.The sipapu (fig. 18, b) was 2 feet 1 inch (63.5 cm) from the firepit and 3 feet 8i/^ inches (1.130 m) from the wall. The hole wasoval in form with diameters of 4i/^ inches (11.43 cm) and 3^ inches(8.99 cm). It was 6 inches (15.24 cm) deep.The storage hole in the floor south of the fire pit (fig. 18, i) had adiameter of 3 inches (7.62 cm) and a depth of 21^ inches (6.35 cm).The holes in the floor where small poles had been set near the com-partment (fig. 18, e) had diameters of 3I/2 inches (8.89 cm) and adepth of 6 inches (15.24 cm).The stones forming facings for the ends of the low compartmentwalls were approximately the same size. The one at the end of thenorth wall was IQi^ inches (26.67 cm) long and 2 inches (5.08 cm)wide. That at the end of the south wall was 10 inches (25.4 cm) longand 2 inches (5.08 cm) thick. The one on the north wall was 514inches (13.97 cm) in height and the other was 4^4 inches (10.79 cm).The space between them measured 11 inches (27.94 cm) . The stonesused in forming the bin ranged from 6 inches (15.24 cm) to 1 foot1 inch (33.02 cm) in length, from 5^/2 inches (13.97 cm) to 1 foot7 inches (48.26 cm) in height, and 1 inch (2.54 cm) to 3 inches(7.62 cm) in thickness. The adobe ridge in which they were setaveraged 5 inches (12.7 cm) in height. In front of the ventilatoraperture the end of the north partition was 2 feet 2 inches (66.04 cm)from the wall, and that of the south, 2 feet 3 inches (68.58 cm)from the openmg.The ventilator opening measured 1 foot 1 inch (33.02 cm) in widthand was 1 foot 1 inch (33.02 cm) in height. The sill was 3I/2 inches(8.89 cm) above the floor. The passage had a length of 1 foot 10inches (55.88 cm), was 11 inches (27.94 cm) high, and 1 foot 2 inches(35.56 cm) wide where it entered the shaft. The shaft was ap-proximately 2 feet (60.96 cm) in diameter.Structure No. 10 had only two walls, one where the ventilatoropened into the chamber and the other at the opposite side. Wlierethe additional sides of the pit would normally have been the structureopened into Nos. 9 and 11. The division line between Nos. 10 and 9was indicated by the juncture of the curving east-and-west walls ofthe two pits, but it was not a distinct boundary between the rooms.68764?39 7
86 BUREAU OF AMERICAN ETHNOLOGY IBdll. 121The floor was continuous and showed clearly that it had been laidas a single operation, as it extended from one chamber into the other.In the plan (fig. 19) a boundary between structures 9 and 10 hasbeen indicated by a dotted line. At the opposite side of the chamberthere was a differentiation in rooms. The floor of No. 10 was
Figure 19.?Structure 10. a, holes for support posts ; h, sipapu ; c. Are pit ; d, ladder box
;
e, ventilator opening ; f, ventilator shaft ; g, accumulation of sand above original surface.distinct from that of No. 11 because the latter was at a higher leveland as a consequence formed a line of demarcation between the twochambei's.The interior features of structure No. 10 Avere very simple. Therewere the four holes for support posts, one of which, that at the north,really fell within the limits of house No. 9. Near the center of theroom was a fire pit and ladder box and two sipapu holes. There
ROBBETS] AKCHBOLOGICAL REMAINS IN ARIZONA 87were no other depressions in the floor and there was no deflector infront of the ventilator oi)ening. The superstructure supported onthe four main posts probably had a flat ceiling and sloping poleson two sides. At the ends where there were no walls the roof wasprobably constructed in the manner suggested in the discussion ofNo, 9, namely, timbers were extended from the rectangular frame-works in each adjacent structure across the intervening space tothe frame in No. 10 to form a flat ceiling above the sides where thesloping walls would normally have been. Small poles were un-doubtedly slanted from these beams to the ground so that the super-structure presented an unbroken exterior surface.The fire pit in the center of room No. 10 was roughly circular inform (fig. 19, c). It had been dug into the earth floor and its sidesand bottom were faced with adobe plaster. Encircling it on the floorlevel was a rim of plaster that increased the depth and served as acoping to protect the edges of the pit. A stone slab was embeddedin the face of the pit at the side toward the ladder box. This nodoubt functioned as a reinforcement to prevent damage which mightbe caused by the ladder or people using it. At first it was thoughtthat the stone originally was of sufficient height to serve as a deflector,but this did not seem to be the case. Had it extended some distanceabove the top of the adobe ridge it would have formed a deflectorsimilar to the one embedded in the plaster rim around the pit instructure No. 7. That it had done so can not be assumed from theevidence in the pit, despite the fact the top of the stone suggestedthat it might have broken off.The ladder box (fig. 19, d) was a simple oval-shaped depression ad-joining the fire pit. Its sides and bottom were plastered. The bottomand one edge, where the base ends of the ladder poles had rested, wereroughened and broken.In the floor midway between the fire pit and the west wall of thechamber were two circular holes (fig. 19, h) occupying the positionwhere the sipapu is normally found. It was not possible to tell whichof the two had been intended for the sipapu. Both were carefullyfinished and there was little difference in their diameters and depths.Because of this both have been considered as sipapus.The ventilator was similar to those previously described, althoughthe passage was somewhat longer than in some of the other structuresand the shaft at the outer end was nearly circular in form. The pas-sage was of the tunneled type. Where it opened into the chambera stone slab was embedded in the floor to form a sill. The floor of thepassage sloped slightly upward to the bottom of the shaft. Thisslant was almost imperceptible to the eye and became apparent onlywhen the section through the room was made with levelinginstruments.
gg BUREAU OF AMERICAN ETHNOLOGY [Boll. 121The pit of structure No. 10 measured 10 feet 6 inches (3.2 m) fromwall to wall on the line passing through the ventilator, fire pit, andsipapu. Between the raised floor level separating Nos. 10 and 11 anda line drawn between the two points where the arcs of the walls forNos. 9 and 10 intersected, the room measured 10 feet 5 inches (3.175 m)
.
At the ventilator side of the chamber the pit had a depth of 4 feet11/^ inches (1.257 m) below the old ground level and 5 feet 4 inches(1.625 m) below the present ground level. At the opposite side of theroom the floor was 4 feet 21/2 inches (1.282 m) below the ground levelat the time of occupation. From the present level to the floor was 5feet 71/2 inches (1.714 m).The hole for the north support post had diameters of 1 foot 2 inches(35.56 cm) and 1 foot 4 inches (40.64 cm). The hole was 1 foot 2inches (35.56 cm) from the point where the walls of Nos. 9 and 10intersected. The hole for the east support post had diameters of 8inches (20.32 cm) and 6 inches (15.24 cm). It had a depth of 1 foot51/^ inches (44.45 cm) and was placed li/^ inches (3.81 cm) from thewall. The hole at the south corner of the room had diameters of 9inches (22.86 cm) and 8 inches (20.32 cm) and was 4 inches (10.16 cm)from the wall. The hole near the west corner of the chamber touchedthe floor of structure No. 11. It had diameters of 1 foot 3 inches(38.1 cm) and 11 inches (27.94 cm). It was 7 inches (17.78 cm) fromthe corner of the room and had a depth of 1 foot 41/2 inches (41.91 cm)
.
The fire pit (fig. 19, c) in No. 10 had a diameter of 1 foot 5 inches(43.18 cm) on the ventilator line. At right angles to this measure-ment the pit was 1 foot 8i/^ inches (52.07 cm) across. At the ventilatorside the bottom of the pit was 4 inches (10.16 cm) below the floor,and at the opposite side was 3 inches (7.62 cm) below that level. Thisdepth, however, was increased by the adobe rim which encircled thepit. The average height of the rim was fi/^ inches (3.81 cm). Thewidth of the ridge varied from 5 inches (12.7 cm) to 8i/^ inches (21.59cm). The stone set in the face of the pit at the ventilator side was 1foot 11/2 inches (34.29 cm) long, 21/2 inches (6.35 cm) wide, and 8inches (20.32 cm) high.The space between the fire pit and the ladder pit was 51/2 inches(13.97 cm). The ladder pit (fig. 19, d) was 1 foot 10 inches (55.88cm) long, 11 inches (27.94 cm) wide, and 6 inches (15.24 cm) deep.From the edge of the ladder pit to the ventilator opening was 2 feet 2inches (66.04 cm).The sipapu, or rather two holes which may have functioned in thatcapacity (fig. 19, 5), were 1 foot 9 inches (53.34 cm) from the edgeof the fire pit rim and 2 feet 11 inches (88.9 cm) from the wall atthat side of the chamber. The one at the north had a depth of 41/2inches (11.43 cm) and the other was 61/2 inches (16.51 cm) deep. Theholes were 1 inch (2.54 cm) apart.
Roberts] ARCHEOLOGICAL REMAINS IN ARIZONA g^The ventilator opening was 1 foot 3I/2 inches (39.37 cm) wide and1 foot 71^ inches (49.53 cm) high. The stone slab which formed thesill for the opening was 1 foot 3 inches (38.1 cm) long, 5 inches(12.7 cm) wide, and li^ inches (3.81 cm) thick. The passage had alength of 5 feet 9 inches (1.752 m). It was 1 foot 1 inch (33.02 cm)wide where it opened into the shaft and was 1 foot 8 inches (50.8cm) high. The shaft had diameters of 2 feet 6 inches (76.2 cm)and 2 feet 10 inches (86.36 cm) at the passage side. The originaldepth was 4 feet 7l^ mches (1.409 m) and the depth at the time ofexcavation was 5 feet 5 inches (1.651 m), the increase being due tosand accumulation on the surface. At the back of the shaft the orig-inal depth was 5 feet 2 inches (1.574 m) and the present depth 5feet 81/2 mches (1.739 m).Structure No. 11 was somewhat more elaborate than No. 10 and inthis respect corresponded better to No. 9. The pit was quite irregularm shape (fig. 20) . The west wall was roughly crescentic in contour,but the east was decidedly uneven. The south side was comparativelystraight, although of course there were rounded corners. At thenorth the chamber opened into structure No. 10. As in the case ofthe other rooms there were the holes for the four posts upon whichthe superstructure had rested. The covering was probably like thatsuggested for No. 9. It was typical of the flat ceiling, sloping-sidedtype except vrhere it opened into room 10. As previously suggested,there possibly was a flat roof between the main frameworks of Nos.10 and 11 and slanting poles were placed at either side to fill in thespace between the ceiling and the walls of the excavation.The interior features of structure No. 11 (fig. 20) were fairlycomplete. Near the center of the chamber was a slab-lined fire pit,and at the ventilator side were the remains of a compartment similarto the one described for Nos. 1 and 9 (pi. 10, h). At the center ofthe room were two holes in the floor, either one of which might havefunctioned as a sipapu. They were not in the same position, how-ever, with respect to each other as those in No. 10. There was nodeflector and no ladder pit in this structure.The fire pit (fig. 20, e) had been dug into the surface of the floorto a depth which corresponded with that of the pits in other houses.The walls of this excavation were then lined with blocks of stonewhich were covered with adobe plaster. The plaster extended overthe faces of the stone and down across the bottom of the pit. Bothof the holes near the fire pit in the approximate position of thesipapu had been carefully plastered (fig. 20, c, d).Only the stone portion of the wall which formed the compartmentat the ventilator side of the room was in position when the house wascleared of the debris which had accumulated in it. Some of the slabswere found lying where they had fallen and grooves in the floor
90 BUREAU OF AMERICAN ETHNOLOGY [Bull. 121gave evidence that there had been additional ones. AU of the plasterthat had been placed between and over the stones had slumped awayas a result of moisture. There seems little question but what therewas a complete partition at that side of the chamber, except for aspace at the center between the fire pit and ventilator opening wherethere was a doorway. A metate was in the north end of the com-partment. In this connection it will be recalled that there were two
5FEET
Figure 20.?Structure 11. a, holes for main support posts ; ft, stone in compartment wall
;
c and d, holes in floor in position of sipapu ; e, fire pit ; f, stone in compartment wall
;
g, storage basin ; h, stone mold ; i, ventilator shaft ; ;, accumulation of sand aboveoriginal surface.milling stones in the bin in structure 1. That in No. 11 had beenplaced against the wall of the compartment but the collapse of thelatter had permitted it to fall to the floor.The floor space for No. 11 was distinct from that of Nos. 9 and10 inasmuch as it was at a higher level (fig. 21). In this respect itcorresponded to the relationship between structures 6 and 7 wherea difference in levels was also noted.
ROBBBTS] ARCHEOLOGICAL REMAINS IN ARIZONA QlThe ventilator was of the tunneled form. As in the case of someother examples there "was a slight although almost imperceptibleupward slant to the passage floor. The shaft at the outer end wasdamaged and it was impossible to obtain full information concerningits character. It apparently was almost circular in form and ratherlarge when compared with some of the others, but beyond this noth-ing could be ascertained.The pit for No. 11 measured 11 feet 1 inch (3.378 m) on thesipapu, fire pit, ventilator line. At right angles to this measure-ment there were 9 feet IQi^ inches (3.009 m) between the southwall and the edge of the floor between Nos. 10 and 11. At the ven-tilator side of the chamber the floor was 3 feet 5 inches (1.041 m)below the old ground level and 4 feet 7 inches (1.397 m) below thepresent surface. At the opposite side of the room the pit had adepth of 3 feet 7 inches (1.092 m) below the surface at the time ofoccupation. The present ground level is 4 feet 9 inches (1.477 m)above the floor.As was the case in some of the other structures, there was con-siderable variation in the size of the holes in which the main supportposts were set. The hole at the north corner of the chamber haddiameters of 11 inches (27.94 cm) and 10 inches (25.4 cm). Thedepth was 1 foot 4 inches (40.64 cm) and the edge of the hole was214 inches (6.35 cm) from the wall. The east hole had diametersof 6 inches (15.24 cm) and 7 inches (17.78 cm). It had a depth of1 foot 3 inches (38.1 cm) and was 91/2 inches (24.13 cm) from thewall. The south hole was decidedly oval in form, with diametersof 91/^ inches (24.13 cm) and 5i/^ inches (13.97 cm). The hole was1 foot (30. 48 cm) deep and was placed 1 foot 21/2 inches (36.83 cm)from the wall. The west hole had diameters of 9 inches (22.86 cm)and 10 inches (25.4 cm). The depth was 1 foot 1 inch (33.02 cm),and the distance between the edge of the hole and the wall of thechamber measured 1 inch (2.54 cm).The fire pit (fig. 20, e) measured 1 foot 4i/^ inches (41.91 cm) onthe ventilator line. In the opposite direction its greatest width was1 foot 7 inches (48.26 cm). The pit had a depth of 6 mches(15.24 cm).The two holes in the floor at the west side of the fire pit werecloser to the pit than ordinarily was found to be the case. Thefirst, or one immediately adjacent (fig. 20, d), was only 21/^ inches(6.35 cm) from the fire pit. The hole had diameters of 814 inches(21.59 cm) and 7 inches (17.78 cm). The depth was 41/2 inches(11.43 cm). The second hole (fig. 20, c), was 2 inches (5.08 cm)from the first, 1 foot 1 inch (33.02 cm) from the fire pit, and 3 feet6 inches (1.041 m) from the wall of the room. It had diameters of
92 BUREAU OF AMERICAN ETHNOLOGY [BULU 1219 inches (22.86 cm) and 91/2 inches (24.13 cm). Its depth was 31/2inches (8.89 cm).The only depression in the floor, in addition to the fire pit and t he-two holes near it, was located at the north end of the south partitionwall (fig. 20, g). The diameters of the hole were 1 foot 1 inch(33.02 cm) and 11 inches (27.94 cm). The depth was 4 inches(10.16 cm).The stones employed in the construction of the low wall formingthe compartment varied in size. The one standing between the e.istsupport and the wall of the chamber (fig. 20, h) was 8 inches(20.32 cm) long, 5 inches (12.7 cm) wide, and 1 foot 4 inches (40.64cm) high. The space between the end of the slab and the edge ofthe hole for the post measured 1 inch (2.54 cm). The only slabremaining in position between the hole for the east post and the pas-sage way between the fire pit and the ventilator opening (fig. 20, /)
lOrEETFigure 21.?Section through structures 9, 10, and 11. a, accumulation of drifts and aboveold surface.was 6 inches (15.24 cm) long, 4 inches (10.16 cm) wide, and 1 foot3 inches (38.1 cm) high. The stone that had stood between this oneand the east support post had a length of 1 foot 9 inches (53.34 cm),a width of 7 inches (17.78 cm), and a thickness of 3 inches (7.62 crn)!The opening in the center of the partition wall measured 1 foot 11inches (58.42 cm). The north stone in the south portion of thepartition was 9 inches (22.86 cm) long, 31/0 inches (8.89 cm) thick,and its top was 1 foot 1 inch (33.02 cm) above the floor. The nextstone stood 2 inches (5.08 cm) from the preceding one. It was tlielargest in the group, with a length of 1 foot 4 inches (40.64 cm),a thickness of 31/2 inches (8.89 cm), and a height of 1 foot 7 inches(48.26 cm). This stone had slumped out of position but the placewhere it originally stood was plainly evident (fig. 20, h). Thenext stone in the group was separated from the one just describedby a space of only 1 inch (2.54 cm). This slab had a length of 1foot (30.48 cm), a thickness of 21/2 inches (6.35 cm), and stood 1 foot6 inches (45.72 to) above the floor.
Roberts] ARCHEOLOGICAL REMAINS IN ARIZONA 93The ventilator opening was 1 foot 8% inches (52.71 cm) wide at thebottom. At the top it measured 1 foot 4 inches (40.64 cm). Theheight of the aperture was 1 foot 41/2 inches (41.91 cm). The sillof the opening was 6i/^ inches (16.51 cm) above the floor. The pas-sage w^as 2 feet 5 inches (73.66 cm) long. The shaft at the outerend was approximately 2 feet 11 inches (88.9 cm) in diameter. Theoriginal depth of the shaft was 2 feet 10 inches (86.36 cm), and atthe time of excavation the floor was 3 feet 11 inches (1.193 m) belowthe ground level.The pillar left in the center of structure No. 10 to facilitate astudy of the fill in the pits had a height of 5 feet 2 inches (1.574 m).There were 8 distinct strata in the pillar and the material containedin each layer v?^as quite dissimilar in character to that in the others(pi. 11, a). The first stratum at the top consisted of rather coarse,clean yellow sand that apparently had been carried into the pit bywind action. The layer was somewhat uneven, as the diagram (fig.22, a) shows. A shallow ditch ran through the top of the fill alongthe line that approximately cut the group of three structures throughthe longitudinal center. This may be attributed to the fact that thepits were located on the side of the ridge where there was a doubleslope, one extending from the west walls of the chambers toward theventilator sides of the structure and the other from No. 11 towardNo. 9. The original ground level at the north side of No. 9 wassomewhat lower than that at the south side of No. 11 (fig. 21).In addition, there originally had been a second ridge just beyondthe series of ventilator shafts. At the time when the houses wereoccupied there was a small gully between the edges of the ventilatorshafts and the slope of the second ridge. It no doubt carried awayall surface drainage from around the structures. After they wereabandoned, fell into decay, and the pits had become filled with ac-cumulated debris, the wind shifted sand across the site and piled itup against the ridge to the east. As a result its western slope wasextended until it covered the old ventilator shafts and shifted thedrainage channel, for the small area involved, several feet. It thenpassed over the subterranean portion of the houses, cutting the gutterindicated in layer a in the central face in the group of three shownin the diagram (fig. 22). That this took place subsequent to theabandonment of the site was indicated by the fact that there were nopotsherds, bone fragments, stone chips, or charcoal in the layer.Layer a varied in thickness from 6 inches (15.24 cm) at the centerof the ditch to 1 inch (2.54 cm) at the corner of the pillar. Thethinnest portion was at the east side where there was an upwardslope toward the east ridge beyond the ends of the ventilators.The second layer in the pillar (fig. 22, h) was composed of dis-colored sand mixed with a considerable amount of humus. This layer
94 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 121also appeared to be subsequent to the occupation of the site. Pre-sumably there had been considerable vegetation on the surfaceof h for a fairly long period of time prior to the deposition of thetop stratum of clean sand. This vegetation with the attendant decayof the plant matter was what gave the color to the sand and wasresponsible for the humus content. As will be seen from the drawing(fig. 22, &), this layer was somewhat irregular in outline. It wasthickest on the west side and as in the case of the upper layer slopedupward toward the east. The stratum ranged in thickness from61^ inches (16.51 cm) to 1 foot (30.48 cm).Stratum c was of interest because it was not continuous through-out the course of the three pits, in an east-and-west direction. Itoccurred as a lens extending from north to south along the eastiV
-;.?
-.c?.' -:V '.':'.'^ :"; .'?
^/;.'-:;--^/:<p;:;-^;;. :..;.;,;' .??.:? 'O. ?
, _ _ _ . fe^r>Vi;"-/vv'::i;A ji'^UVS'l-it.tj^'!!ij'j
SCALE FEETFigure 22.?Test pillar from fill in structure 10. N, W, S, and E, indicate north, west,south, and east corners.
side of the fill. The material apparently washed into the pits. Thelayer immediately below had been deposited with a pronounceddownward slope toward the eastern edge, and c evened up the sur-face of the fill. It extended across from approximately the highpoint on d, the stratum below, to the east edge of the excavation.Stratum c contained some fragments of charcoal, some ash-bearmgsand, and occasional nodules of adobe. In its general appearanceit was somewhat darker than the stratum immediately above. Fromthe point where it coincided with stratum d, layer c expanded to adepth of 10 inches (25.4 cm) at the south side of the pillar andnarrowed to 6 inches (15.24 cm) at the east side. At the west side ofthe pillar strata h and d touched.Stratum d was composed of light-colored sand containing occa-
EOBKETS] ARCHEOLOGICAL REMAINS IN ARIZONA 95sional fragments of charcoal but no ashes. There were no indica-tions of human handicrafts, such as worked bones, stones, or potteryfragments. As in the case of the two upper layers and the lens c,stratum d apparently was deposited at a time when there was little,if any, activity about the site. At the west side of the pillar thisstratum measured 7 inches (17.78 cm) thick. Near the center itincreased to 10 inches (25.4 cm) and at the east side narrowed to63^ inches (14.61 cm).Stratiun e was of particular interest because it consisted mainlyof refuse material such as is deposited around an occupied site, andin addition had a long oval depression through its center somewhatsimilar to the drainage channel described for the first level. As amatter of fact, it would be more appropriate to say the top layerof stratum / contained the hollow and that the refuse material com-prising stratum e had been deposited in the depression and over thegeneral surface surrounding it. Stratum e consisted largely of ash-bearing sand with a heavy content of charcoal. In addition, bonefragments, potsherds, and stone chips and spalls were abundant.This material was unquestionably waste matter from a habitation.Stratum e was thickest at the east side where it measured 1 foot(30.48 cm) in depth, and was thinnest at the south corner wherethere was only 3 inches (7.62 cm) of material. The channel or ditchin the center gave that portion of the stratum a thickness of 8 inches(20.32 cm).The next layer, stratum /, was composed chiefly of clean sand withsporadic fragments of charcoal and a few pieces of sandstone. Therewere no ashes, no bone, nor stone chips, such as resulted from themanufacture of stone implements, and potsherds. The layer as awhole gave the impression of a natural deposit that had accumulatedfrom the action of wind and water. The layer had drifted fromthe southern end into the elongated depression caused by the pitsand tapered off toward the north and east where the thinnest por-tion occurred. It was in the top of this layer that the hollow dis-cussed in connection with stratum e occurred. Stratum / ranged indepth from 3 inches (7.62 cm) to 1 foot 2 inches (35.56 cm).Stratum g was practically identical with / insofar as content wasconcerned. The distinction between the two was based on a thinlayer of compact clay and sand that had formed the top surface ofg prior to the deposition of /. The narrowest portion of g was atthe east side of the pillar. Unlike /, there were no indications ofa channel across the surface of g. No doubt this may be attributedto the fact that it was at a sufficient depth below the portion of thehouse pits to prevent a flow of water across its surface. Any waterwhich ran into the pits at this period would of necessity have toremain there. The greatest depth in the stratum was found toward
Qg BUREAU OF AMERICAN ETHNOLOGY [Bull. 121the south of the fill. Its surface sloped downward toward the northand upward toward the east. At the southwest corner of the pillarthe deposit measured 1 foot 4 inches (40.64 cm) . At the east side itwas only 3 inches (7.62 cm) deep, and at the north 10 inches(25.4 cm).The lowest layer in the fill, stratum h, consisted of clean sand thatgave evidence of having been deposited by wind and water action.The sand was a distinct yellow in color. The line of demarcationbetween the top of h and the bottom of g was very distinct. Stratumh contained no traces of human presence, except at the very bottomwhich was the floor of the houses and where a few potsherds andsome implements were found. The outstanding feature of the de-posit was that it reached its greatest depth along the east side of thefill and was little more than perceptible at the west side. In addi-tion, it sloped slightly upward toward the north. In other words,the first material that found its way into the pits where structures9, 10, and 11 had been was banked up against the east and northwalls of the series of three connected excavations. This would sug-gest that at the time of the deposition of this part of the fill theprevailing wind had been from the southwest, blowing the materialagainst those two sides. Stratum h ranged in thiclaiess from V-/2inches (3.81 cm) along the west wall to 1 foot 6 inches (45.72 cm) atthe east.From the data present in the fill of structures 9, 10, and 11, it isapparent that the pits were used as a dumping place for waste ma-terial from habitations at only one time during the interval inwhich the excavations were being filled to the ground level. Thistook place at about the middle period, as the refuse-bearing stratumwas practically midway in the pillar. This evidence seems to war-rant the conclusion that for some time after the combined struc-tures were abandoned the area immediately adjacent to them was un-occupied, since no signs of human endeavor found their way intothe pits. Then for a certain period people dwelt nearby and usedthe pits for a dumping place. Wlien this ceased the depressions con-tinued to fill in a natural manner with no assistance from humanhands. The potsherd evidence obtained from stratum e indicatesthat the refuse material may be closely correlated with the remains ofthe small surface structure to the west of the group. This also dem-onstrates that the three pit structures preceded the surface house andprobably represent an earlier stage in the development of the localcultural pattern.The combination of three structures into one large dwellingshows an interesting development in the matter of house construc-tion and thus far in the Southwest this group constitutes the onlyexample of such a procedure. As previously mentioned, the early
Roberts] ARCHEOLOGICAL REMAINS IN ARIZONA 97inhabitants of the district included in the Long H Ranch, some 42miles (67.592 k) southwest from the Allantown site, grouped theirpit houses in clusters of three but made no attempt to actually jointhem into one large structure.*^ Another example mentioned was thatof Martin's houses in southern Colorado.*^ Significant features inthe combination of structures 9, 10, and 11, as well as in 6 and 7,were that despite the fact that one long, narrow room was obtained,the units that went into its make-up were individually complete (pi.11, h). This may be considered as an indication that the peoplewere ready to combine their houses, but that they were not yet atthe stage where they were willing to give up house characteristicsthat had long been present in the single dwellings.Another factor of some importance is raised by this evidence ofthe joining together of three structures. In the early days of south-western archeology a theory became fairly well established to theeffect that the rectangular-roomed communal buildings of the Pueblopeople developed out of the practice of combining a number of cir-cular houses into one large structure and the subsequent discoverythat a better building could be erected if the walls of the chamberswere straight instead of curved.** Later work, particularly in themore northern parts of the area, contributed evidence which wentagainst this theory and demonstrated that the rectangular house de-veloped before the practice of combining many chambers into onelarge building.*^ If the present two groups comprising structures6, 7, 9, 10, and 11 constitute a prototype for the communal buildingerected above ground in later cultural horizons it is evident that inthis particular district the evolution of the house followed a differ-ent course from that observed for the more northern parts of thearea. Furthermore, it would indicate that the old theory was notafter all entirely wrong. However, the Allantown site contains onegroup of structures that seems to offer an example of a more logicalstep in house development. Because of this, the facts would sug-gest that the groups of combined pit structures, one containing twohouses and the other three, were peculiar local developments thatdid not have a direct bearing on the subject of communal houses ingeneral. They had a limited development, possibly the culminationof a purely local trend, that apparently had no influence whateveron subsequent Pueblo structures.At first consideration it would seem that the problem of erectinga superstructure over three pits combined in the manner of Nos. 9,10, and 11 would be a difficult undertaking. As a matter of fact.
*2 Roberts, 1931, fig. 1, pp. 25-40.*3 Jlartin, 1930, pp. 29-33, pi. vli, fig. 2."Gushing, 1806."Roberts, 1929, p. 147; 1930, p. 62.
gg BUREAU OF AMERICAN ETHNOLOGY TBull. 121the construction probably was quite simple and would have beeneasier to accomplish had the builders not retained as far as possibleall of the characteristics of individual and separate dwellings. Thepresence of four upright support posts in each chamber would ma-terially strengthen the roof, but the covering could have been con-structed with four less posts had the builders made two serve, insteadof four, at each opening between the chambers. The rectangularframeworks placed on the upright posts in each chamber probablycarried flat ceilings over the central portion and no doubt supportedsmaller poles whose upper ends rested against the main stringers andwhose base ends rested upon the ground back of the edges of the pit(fig. 23). In this respect the superstructure would in general havebeen similar to those erected over single pits. The only place wherea difference would have occurred was at the sides where the chambersopened into each other and here, as previously mentioned, there prob-ably was an extension of the flat portion of each roof by means oftimbers laid across the opening from chamber to chamber. Smallpoles bridged the space between these and the gTound. While thepostulated method of construction, as shown in figure 23, probablyerrs in minor details, it is thought that in general the reconstructionis fairly accurate and should give a good idea of the superstructureframework erected over the three pits.SUBrACE HOUSE B AND GRANARIESSurface house B was located 23 feet 6 inches (7.162 m) west of theNos. 9-11 group, and its south end was 4 feet (1.219 m) from the6-8 structures (fig. 13). As in the case of surface house A, littleremained of B. There were only a few courses of stone in positionabove the foundation and portions of the east and south walls wereentirely missing. There was some question at the time of excavationas to whether or not the walls actually represented the remains of ahouse or indicated an enclosed court which had low walls and waswithout a roof. The absence of building stones in the debris sug-gested that the walls had not risen to any gi-eat height. On theother hand, it is possible, as was mentioned in the discussion of sur-face house A, that the building was stone-robbed at the time whenthe large Pueblo structures were erected, the builders making use ofall available material in the vicinity. The stones used in laying thesewalls were only partially shaped for building purposes. The faceswere not carefully dressed and the blocks were irregular in formIf structure B was originally a house it consisted of a building con-tamm.g only one long narrow room (fig. 24). There were no indica-tions of walls or partitions. It is possible that the interior may havebeen separated into smaller chambers through the medium of wattle-
ROBEBTS] ARCHEOLOGICAL REMAINS IN ARIZONA 99
100 BUREAU OF AMERICAN ETHNOLOGY [Bull. 121and-daub walls, but if they were ever present no traces of themremained.At the south end of the masonry walls were the remains of a stor-age cist (fig. 24, a). This cist was roughly rectangular in form,judging by what remained of it, and although the sides were fairlystraight the corners were rounded. This place probably was a gran-ary in which corn was stored. Just what type of superstructure waserected over it is not known. Possibly a framework of poles andbrush was placed over it and the exterior surface was covered withplaster. The main timbers probably sloped from the ground level
-r-? M M ; f I II I f-^^-nbbE^
FiauHE 24.?Surface remains B in first group. Granary pits at a, b, c, and d; exteriorfire pit at e.to a point directly over the center of the cist where they joined.This would have formed a conical-shaped structure that in generalappearance would have been not unlike the present-day sweat housesof the Navajos. Granaries of similar construction have been foundin various portions of the Southwest, some of them practicallyintact.^^Because of this it seems logical to postulate that the method ofconstruction in the present case was similar. The floor of the gran-ary was paved with large slabs and around its periphery additionalstones were set up on edge to make a low wall. A portion of the
^'Morris, 1925, p. 270.
ROBERTS 1 ARCHEOLOGICAL REMAINS IN ARIZONA JQ]^floor and bounding wall at the east side was missing but the majoroutline was present.Within the inclosure formed by the rectangular wall were the re-mains of a similar cist (fig. 24, 6). If the surface structure had beena house it is hard to explain why a granary had been built within it.On the other hand, it is possible that the feature in the north end ofthe structure does not represent such a granary but merely a slab-lined depression in the floor which served either for storage pur-poses or as a mealing bin.Outside the inclosure at the north end was what remained of twoadditional cists (fig. 24, <?, d) . In addition, there was an exterior firepit 5 feet (1.524 m) from the northeast corner of the masonry walls(fig. 24, e).In general it may be said that the surface remains at tliis portionof the site corresponded in age to the group designated surface houseA. They gave evidence of being later than the nearby pit structures6, 7, 8, 9, 10, and 11. This was demonstrated by the fact that thefoundations of the stone walls and the granaries were at a higherlevel than the surface of occupation at the time when the pit struc-tures were inhabited. This is shown by the section through thebuilding (fig. 24).The masonry inclosure was 29 feet 10 inches (9.093 m) long and8 feet 8 inches (2.641 m) wide, outside measurement. Inside thewalls the length was 28 feet 8 inches (8.737 m) and the width 7 feet3 inches (2.209 m). The average height of the remaining stoneswas 1 foot 6 inches (45.72 cm). The storage bin at the south endof the structure measured 6 feet (1.828 m) across. The upright slabsaround the sides had an average height of 9 inches (22.86 cm). Thenorth wall of the granary was 5 inches (12.7 cm) from the southwall of the masonry structure. The bin in the north end of themasonry inclosure measured 4 feet 2 inches (1.270 m) by 6 feet(1.828 m). Its floor was 6 inches (15.24 cm) lower than that of theinclosure. The slabs which bordered the pit had an average heightof 1 foot (30.48 cm).Located along the top of the ridge southwest from surface houseB was a group of three granary pits (fig. 13). These pits were situ-ated 21 feet 3 inches (6,477 m) from the surface remains and were14 feet 6 inches (4.419 m) from pit structure No. 8. Two of thepits were roughly oval in form while the third was definitely rec-tangular. Two of the pits had slab facings on their walls whilethe tliird merely had a coating of plaster. All three had been de-stroyed by fire and there were large quantities of charred corn inthem. Indications were that the upper portion of these structureshad been of pole, brush, and plaster construction, such as that de-68764?39 8
2Q2 BUREAU OF AMERICAN ETHNOLOGY [Bull. 121scribed for the granary a, in connection with surface house B.Great chunks of plaster burned to a bricklike consistency were foundin the pits. These blocks of fired clay bore the imprints of thepoles and brush upon which they had rested. A few potsherdsfound in the fill of the pits indicated that they had been contem-poraneous with the semisubterranean structures rather than withthe surface remains.The two oval structures had diameters of 5 feet 11 inches and 5feet 3 inches (1.803 and 1.600 m), and 3 feet 8 inches and 4 feet(1.117 and 1.219 m). The rectangular structure measured 5 feetfeet 10 inches by 5 feet 9 inches (1.778 by 1.752 m). The depthsvaried. The northernmost pit in the group had an average depthof 8 inches (20.32 cm), the rectangular one had a depth of 1 foot(30.48 cm), and the southernmost of the group was 1 foot 6 inches(45.72 cm) in depth. Group No. 2STRUCTURE 12Structure 12 was the first of the remains uncovered in the secondgroup. It is located along the edge of the ridge north of the GreatPueblo ruins and at a considerable distance from the group discussedin preceding pages (fig. 25). A large portion of the east wall hadbeen carried away by erosion, but the damage was not sufficient todestroy all the evidence essential to a reconstruction of the structure.The pit was much deeper than in many of the other examples andfor that reason the major outline of the subterranean portion of thestructure was still intact. Structure 12 was the largest of the en-tire group and differed from the others in that it had a bench en-circling the upper borders of the pit. The pit was definitely D-shaped and more regular in outline than the majority of the excava-tions (fig. 26). The back of the bench was not as uniform as themain wall of the chamber but nevertheless conformed rather closelyto its perimeter.The superstructure erected over this pit was different from thetypes previously described. Because of the size of the pit additionaltimbers were needed to cover it. There had been four main supportposts placed in the floor near the corners of the chamber. In addi-tion there was a series of posts around the back of the bench. Thesealso seemed to have functioned in a support capacity. Three of themwere actually present and there were indications of a fourth wherea portion of the bench had been carried away. Along the easternand southern sides too much of the bench was missing to enabledefinite determination that such posts had been present. On thebasis of the distances between those found it may be suggested that
ROBEiBTS ] ARCHEOLOGICAL REMAINS IN ARIZONA 103there were possibly two more?one near the east corner and one atthe south. This would have made a total of six for the structure.The timber portions of the roof had been destroyed by fire and as aconsequence there were large sections of charred beams scattered aboutthe floor. The positions of the various fragments were carefully
/<^/ ^'SHELTERVv^-'granaries
FIRE ,?? PIT f/
.GRANARIES \ \ STRUCTURENO l?^^-?-^^
11 ? "3
y-v5t>'' SHELTERCRANARIES) PolUr;
>
.T?l,, )Bulil )JRCruSC UOUNO
I PILE OFIloosesrocrs
/V
.PoRGRANARIES
/^^?V \ STRUCTURES/ \ 1 NO, I3?\ A~7/ NO. 131
34 rcETTO SREAT PUEBLORUINS
e
FiotiEB 25.?Group 2 pit and surface remains. Numbered dots indicate location of burials.plotted in the hope that they would indicate the type of roof employed.The results were not entirely satisfactory, although they do givesome clue (fig. 27). One definite conclusion based on their evidenceis that the roof was not cribbed. As a postulation the most plausible
j^Q^ BUREAU OF AMERICAN ETHNOLOGY [Bull. 121form of construction seems to be as follows: The four main postsset in the floor of the chamber carried a rectangular framework suchas previously described for the smaller pit structures. In addition,the posts set in the back wall of the bench probably had stringersextending from the top of one to the top of another around theperiphery. This is indicated by two factors. Two of the outer up-rights were forked at their upper ends. This, together with a lineof ashes and burned plaster extending along the top of the benchbetween two others, is considered evidence for the use of stringers.The outer framework was not as high as the central one; in fact itstood only a short distance above the surrounding ground level.Smaller poles were sloped upward from it to the central frameworkand downward to the ground level around the pit. These in turn sup-ported light timbers placed crosswise, paralleling the outer framework.On top of them bark, brush, and a coating of plaster completed thecovering. This superstructure would compare quite favorably withthat of the large earth lodges of the Plains Indians."*^ The centralportion probably had a flat ceiling, as in the case of the other struc-tures, although it might have been partially cribbed with a flat por-tion near the center for the entrance and smoke-hole hatchway. Thefinished roof according to such construction possibly had an appear-ance closely approximating that shown in the postulated reconstruc-tion (fig. 28).The interior features for structure 12 were elaborate (pi. 12, a).Near the center of the chamber was a circular fire pit that had abordering rim of adobe and a stone slab reinforcement at the ladderside (fig, 26, j). Near the fire pit was a second smaller depression inthe position normally occupied by the ladder box (fig. 26, k), but inthis case it unquestionably functioned as a depository for ashes. Thissmaller depression was also bordered by a rim of plaster and at theventilator side had a large stone incorporated in the wall. This slabwas the deflector. The ladder used in gaining access to the cham-ber was of the runged type and the base ends of the side poles werestill embedded in the floor between the deflector and the ventilatoropening when the room was cleared of its accumulated debris (fig. 26,m). The charred ends of the posts were set at an angle whichshowed that the smoke-hole hatchway must have been located directlyabove the fire pit.There were numerous holes and depressions in the floor. In thespace between the fire pit and the wall at the proper location for asipapu there were two circular holes (fig. 26, i). Originally therehad been two similar holes in about the same positions but these hadbeen filled in and plastered over and two more provide'd. At one? Bushnell, 1922, pi. 38, a; pi. 40, pp. 132, 133. and 135.
BUREAU OF AMERICAN ETHNOLOGY BULLETIN 121 PLATE 11
View across structures 9, 10, and 11.
BUREAU OF AMERICAN ETHNOLOGY BULLETIN 121 PLATE 12
a. Floor in structure 12, showins: ladder holes, deflector, ash and fire pits.
It. l'iastei'ed-u\ er 'ifjiressioii in lioor oi srruetiire IL'.
BUREAU OF AMERICAN ETHNOLOGY BULLETIN 121 PLATE 13
BUREAU OF AMERICAN ETHNOLOGY BULLETIN 121 PLATE 14
S.t0>>^'-^
?lA
a. Floor of structure 13b.
b. Portion of floor area, structure 13b.
ROBHHTS 1 ARCHEOLOGICAL REMAINS IN ARIZONA 105
::::&
FiGDEE 26.?Structure 12. a, holes for main support posts ; b, secondary supports at backof bench ; c, paving on floor ; d, plastered-over hole in floor ; e, storage hole ; f, plastered-over basin ; g, storage hole in floor ; h, storage basin ; i, sipapu holes ; ;, fire pit ; k, ashpit ; I, deflector ; m, ends of ladder poles ; n, cover stone for ventilator ; o, shelf for coverstone and hole for small post to hold stone ; g, plaster ridges on bench ; r, plasteredover hole ; s, plastered over basin suggestive of subfloor kiva vault ; t, plastered-overholes ; m, ventilator shaft ; v, sand accumulation.
^Qg BUREAU OF AMERICAN ETHNOLOGY [Bo"- 121time or another several holes had been treated in like manner. Therewas one series of three with a possible fourth wliich had formed arectangle encompassing the central portion of the room (fig. 26, d, r,t). The fourth may have been in approximately the position occu-pied by the small hole (fig. 26, g), but a comparatively large areahad been dug out of the floor at that place, refilled and plasteredover (fig. 26, /). As a consequence all indications of a larger holecould have been obliterated. What these holes may have been foris not known, but the fact that they had once been present in soregular a grouping indicates something more than a casual placingof pockets for storage purposes. The thought occurred that pos-sibly structure 12 may at one time have been considerably smallerand posts have been placed there to support a superstructure overa lesser pit. When the excavation was enlarged to the size of No.12 it was found necessary to change the position of the uprights andto employ larger timbers. This is offered only as a suggestion,however, since there was no evidence to warrant the conclusion thatalterations of so extensive a nature had been made. That such aconclusion would not be entirely unwarranted, however, is demon-strated in the case of structures 13a and 13b discussed in followingpages.At the east side of the chamber two stones were embedded in thefloor, their tops flush with its upper surface. Along the fire-pit sideof these stones was a shallow, oval-shaped depression in the floor.The purpose of this combination (fig 26, c) was not clear. Thefeature in general was similar to one found in structure 13b de-scribed in a following section of the report. It may be suggestedthat a metate or grinding stone rested on the slab, while the depres-sion served as a container for corn to be ground or possibly as acatch basin for the meal as it dropped from the stone. The de-pression was bordered on the fire-pit side by a third stone laid in thefloor (fig. 26, /). Between the east support posts and the wall wasan oval-shaped depression (fig 26, h) that probably functioned asa storage place for small objects.At the southwest side of the fire pit there originally was a curi-ously shaped depression which had been filled in and plastered over(fig. 26, 5/ pi. 12, &). How long this depression had been in use be-fore it was abandoned and covered with flooring could not be de-termined. The pit itself consisted of two parts. The large circu-lar portion at the southwest side was quite shallow, while the longoval segment next to the fire pit was comparatively deep. Thenature of this combination is shown in section 2, s, figure 26. Inthe shallow circular part were two small holes that apparentlyhad contained upright sticks. The latter were not of sufficient size
Roberts ] ARCHEOLOGICAL REMAINS IN ARIZONA 107to have had any connection with the superstructure and must havehad some specific purpose apart from any association with theroof. What that may have been can be answered only by specula-tion. The important factor in connection with the presence of thisfloor feature lies in the fact that its general position and locationcorresponds in marked degree to the subfloor vaults found in kivasin ruins throughout this general district and also to the subfloorvaults generally present in great kivas or superceremonial cham-bers.*^ These vaults are generally rectangular in shape and ofgreater depth than the depression m structure 12. Their purpose
FiGUBB 27.?Positions of burned timbers on floor of etructure 12.in kivas has never been determined and as a consequence they canthrow little light on the present occurrence, but the analogy shouldbe mentioned. The significant feature is that such a depression hadbeen present in a position corresponding to that of the kiva vaults.Future investigations may in time furnish data that will definitelyexplain why they were placed in ceremonial chambers. It wouldbe interesting to know what reason the occupants of structure 12 hadfor discarding the depression and filling it in. Unfortunately, thatis one of many things that must remain unknown.There were low ridges of adobe plaster at two places on the topof the bench. These are particularly worthy of comment because of
*? Roberts, 1932, pp. 58-60, 69-70, 88-90.
-, Qg BUREAU OF AMERICAN ETHNOLOGY [Bull. 121the fact that they suggest a prototype for a feature commonly foundin the kivas in the ruins in the Chaco Canyon,*^ at Aztec,'? and atother sites belonging to the Chaco cultural pattern." A series oflow masonry boxes inclosing sections of logs are frequently notedin such ceremonial structures. These are located at intervals alongthe top of the bench. Generally there are four to six in the smallerkivas, the number increasing with the size of the structure. No tim-bers were incorporated in the plaster ridges in this structure, but inview of the fact that they extended out toward the face of the benchfrom the posts set in the wall at the back it may be suggested thatthere was a close correlation between them and the other forms.This tends rather definitely to indicate a relationship between themand the Chaco type. It is possible that the latter type was definitelyderived from such an architectural feature, although in the Wliite-water examples no definite function was indicated comparable tothat of the Chaco forms serving as a base for the bottom tier of logsin a cribbed roof.The ventilator in this structure had a low step of adobe plaster infront of the opening in the wall (fig. 26, o). The top of this ridgecontained a groove in which to set a cover stone. The stone was evenin position over the aperture when that portion of the room wasuncovered. In addition to the groove there was a small hole wherea piece of wood was placed in an upright position to aid further inkeeping the cover in position. The aperture was carefully finishedwith adobe plaster. The passage was long and comparatively small.It had been constructed by the trench-and-cover method. The floorsloped slightly upward toward the shaft at the outer end and thewidth increased in the same direction. The shaft was oval in form.Only a small portion of it remained, however, because it was locatedat the side of the structure where erosion had been most active and alarge portion of the earth was washed away.Structure 12 in general suggested much more the kiva or cere-monial chamber than did any of the other pit remains, with the ex-ception of structure 3. Furthermore, potsherds found in it were ofa more developed type of ceramics than those secured from the others.The date of the structure obtained from the timbers, 918 ?3,^' showsa later horizon. As a matter of fact the timbers covered a span of74 years, the dates being 844 to 918. Inasmuch as a little over one-third of the material gave the 844 date it is probable that those tim-bers were reused and may have come from a smaller structurelocated on the same spot, as was suggested by other features men-tioned in preceding pages. Considering all of the evidence, the? Judd, 1925, p. 88, figs. 99, 100. Pepper, 1920, pp. 104, 106''"Morris, 1924, p. 243.^iJeancon, 1922, p. 18.?= Miller, 1933, p. 31, listed as House A-1.
ARCHEOLOGICAL REMAINS IN ARIZONA 109
\>
a; C
a S
jlQ BUREAU OF AMERICAN ETHNOLOGY [Bum. 121conclusion was reached that structure 12 was a ceremonial cham-ber, probably belonging to the late Developmental Pueblo horizonat the site. Remains of late Developmental structures are incor-porated in the large ruins southwest from the pit of 12. It is pos-sible, of course, that people lived in pit dwellings at this place untillate Developmental Pueblo times, as dates from structure 15 over-lap a number of those from the beams in No. 12, but such was not acommon practice. In the Flagstaff, Ariz., area farther west, pithouses prevailed all through the Developmental period.^^The fill in the subterranean portion of this structure was somewhatdifferent in character from that found in most of the pits. Thefirst layer in the pillar (fig. 29, a) consisted of wind-blown sand.It contained some fragments of tabular sandstone, and potsherds werecomparatively numerous. The stratum was uneven, probably due tothe house being on the slope of the hillside. At the east corner ofthe pillar it measured 5 inches (15.27 cm), at the north corner 8i/^inches (21.59 cm), at the south 9 inches (22.86 cm), and at the westwas 1 foot 41/^ inches (41.91 cm) thick.The second level (fig. 29, h) was not a complete stratum, but partof a lens of material that did not extend entirely across the pit, onthe southwest side. At the north corner of the pillar the top and bot-tom lines met ; on the south face the lens extended only a few inchesbeyond the median line. The layer contained considerable charcoal,some adobe nodules, blow sand, and toward the bottom several streaksof water-deposited sand. Potsherds were not numerous and therewas no stone or bone. The maximum thickness of the layer was atthe east corner of the pillar where it measured 1 foot 2 inches (35.56cm) ; at the south it was 614 inches (16.51 cm).Layer c was practically the same in content as &, except for thefact that it contained less charcoal. The main line of demarcationbetween the two was a distinct streak of water-deposited sand andclay. This line was on an old surface of occupation at about the levelof the top of the bench. On this same surface, in another part of theroom, was a fire pit that had been placed in the fill at a considerableheight above the old floor (pi. 13, a). The material in stratum c wascharcoal, adobe nodules, sand, and some clay. There were a fewsandstone spalls and some bone fragments. Potsherds were present,but were not as numerous as in the layer above. Measurements forthis stratum were: north corner 1 foot 4 inches (40.64 cm), eastcorner IO1/2 inches (26.67 cm), south corner 91/2 inches (24.13 cm),and west corner 81/2 mches (21.59 cm). Strata a and c came togetherat the west corner because of the lens nature of stratum h.
wHargrave, 1930.
ROBBBTS] ARCHEOLOGICAL REMAINS IN ARIZONA 111The next two strata, d and e, were clearly defined by streaksof charcoal. The material in the layers was the same as that forstratum c. The main differences were that there was a slight increasein the amount of stone fragments present in d^ with a sharp drop ine. The number of potsherds present in d was the same as that forc, but a marked decline was noted in e. At the top of c?, in the faceof the pillar between the east and north corners (fig. 29), was a smallpocket filled with water-deposited sand. This apparently was theremnant of a shallow puddle where water had collected in a de-pression when the top of d was the surface of the ground at thatpoint. Stratum d measured : 6 inches (15.24 cm) at the north corner,
:--J-
t^ . : -N ^
,/:c^:^mssmm-m
^^2 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 121corner, 53^ inches (14.60 cm) at the east corner, 4 inches (10.16 cm)at the south corner, and 41/2 inches (11.43 cm) at the west corner.Along the course of the small channel the thickness was 9 inches(22.86 cm). Tliis latter feature probably represented the remainsof a natural drainage ditch, as there was water-deposited sand in itsbottom.Stratum g was separated from / by a layer of charcoal. Thestratum contained numerous lumps of clay, adobe nodules, someburned sand, and wood ashes. The amount of bone was the sameas that for the level above, but in addition there were fragments ofstone and a few potsherds. Stratum g measured 8 inches (20.32 cm)at the north corner, 7 inches (17.78 cm) at the east corner, 4 inches(10.16 cm) at the south corner, and 5^ inches (13.97 cm) at the west.The small channel noted in the discussion of / completely cutthrough g and at that point strata / and h touched. The channelwas 5 inches (12.7 cm) wide.Stratum h was separated from ^ by a definite line of charcoaland ash. The layer had a larger content of sand than was the casein g and in addition some ashes, bits of charcoal, lumps of adobe,and balls of clay. There was a total absence of bone, stone, andpotsherds. The layer measured : S^^ inches (21.59 cm) at the northcorner, 3% inches (9,53 cm) at the east corner, was missing at thesouth corner, and 7i/^ inches (19.05 cm) at the west corner. Aboutmidway of the west-south face an old depression in the top of thelayer below increased the thickness of h to 11 inches (27.94 cm).This pocket was 8 inches (20.32 cm) across and its bottom wasfilled with water-deposited sand.Stratum i rested on the floor of the structure and consisted largelyof burned plaster, charred beams, chunks of adobe, sand, andashes. There were numerous fragments of sandstone, chips of chertand chalcedony, bits of bone, and a fairly good showing of potsherds.This layer was also irregular in thickness due to an uneven top sur-face with humps and pockets attributable in the main to the timbersand charred bits of roofing in the debris and to features on the floor.The layer measured : 1 foot 6 inches (45.72 cm) at the north corner,8 inches (20.32 cm) at the east corner, 1 foot 2 inches (35.56 cm) atthe south corner, and 1 foot 4 inches (40.64 cm) at the west corner.The evidence in the content of the pillar points to two main fac-tors. The first, that the fill accumulating in the pit after the struc-ture was destroyed by fire and up to the level of the bench top,stratum c, was the result of a gradual and natural process. Therewas no intentional dumping of refuse and waste matter from occu-pied quarters in the pit during this interval. At no stage, however,was there complete absence of traces attributable to human activityand for this reason it may be concluded that the site in general was
ROBEHTs] ARCHEOLOGICAL REMAINS IN ARIZONA 113
still occupied while the hole was being filled. The second featureis that the interval that followed the period marked by the surfaceof occupation on the level with the top of the bench, the top ofstratum c, was characterized by deposits containing a large amountof refuse and debris, in this case definite dump material. This seemsto correlate with the major period of expansion that included exten-sive construction work around the largo stone buildings locatedsouthwest from structure 12. The potsherds from the upper levelsof the pillar check with those from the Great Pueblo ruins andsuggest a probable contemporaneity between the two.Structure 12 measured 21 feet 6 inches (6.553 m) from wall towall below the bench on the fire pit, ventilator line. Across in theopposite direction the measurement was 22 feet 8 inches (6.908 m).The bench varied in width from 1 foot 9 inches (53.34 cm) to 2feet 10 inches (86.36 cm). At the ventilator side of the chamberthe existing wall stood 2 feet 3 inches (68.58 cm). This, of course,does not represent the original height at that side of the room, sincea portion of the structure had been carried away by erosion. At theopposite side from the ventilator the top of the bench was 4 feet 8inches (1.422 m) above the floor. The back of the bench was 1 foot3 inches (38.1 cm) high. There was an accumulation of sand abovethe original ground level which added 2 feet 2 inches (66.04 cm)to the latter measurement.The base ends of the main support posts were still in the floor whenthe pit was uncovered, hence it is possible to give the actual measure-ments for the posts themselves. The north post had a diameter of10 inches (25.4 cm) ; it was 1 foot (30.48 cm) from the wall. Thetimber was removed from the hole in order that the wood might besaved for study purposes. After this was done it was found that thehole had a depth of 2 feet 10 inches (86.36 cm). The east supportpost had a diameter of lOi^ inches (26.67 cm) and was 8^/2 inches(21.59 cm) from the wall. The hole in which it was set was 1 foot 10inches (55.88 cm) in depth. The south support post had a diameterof 101/2 inches (26.67 cm), stood 9 inches (22.86 cm) from the wall,and was set at a depth of 1 foot 8 inches (50.8 cm) . The post nearthe west corner was 11 inches (27.94 cm) in diameter, stood 1 foot(30.48 cm) from the wall, and was set at a depth of 2 feet 5 inches(73.66 cm).The posts which were present in the bench were more variable insize than the central supports. That at the north side of the chamberhad been placed in a hole 1 foot 2 inches (35.56 cm) in diameter.The post itself, however, measured only 10 inches (25.4 cm). Themiddle one of the three remaining posts stood in a hole III/2 inches(29.21 cm) in diameter from front to back and 1 foot (30.48 cm)from side to side. The post placed in it had a diameter of 8i/^ inches
j^j4 BUREAU OF AMERICAN ETHNOLOGY [Buu^ 121(21.59 cm) . The post near the south corner of the chamber was thesmallest of the group; it was only 5 inches (12.7 cm) in diameter andwas set in a hole which measured 7% inches (19.68 cm) across. Mostof the larger timbers used in the superstructure had average diam-eters of 6 inches (15.24 cm). The smaller poles averaged 3 inches(7.62 cm) in diameter.The fire pit (fig. 26, j) had a diameter of 2 feet 1 inch (63.5 cm)on the ventilator line and 2 feet (60.96 cm) at right angles to it. Thebottom of the pit was only 4 inches (10.16 cm) below the floor level,but the depth was increased to 8 inches (20.32 cm) by the encirclingrim of adobe. The rim ranged from 7 inches (17.78 cm) to 11 inches(27.94 cm) in width. Its average height was 4 inches (10.16 cm).The stone slab incorporated in one side was 1 foot 9^/2 inches (54.61cm) long, 4 inches (10.16 cm) thick, and 9 inches (22.86 cm) high.The ladder box or ash pit (fig. 26, k) had a depth of 3 inches(7.62 cm). The adobe rim which encircled it ranged from 1^/2inches (3.81 cm) to 3 inches (7.62 cm) in width, and had an averageheight of 1% inches (3.81 cm). The deflector slab which was placedat the ventilator side of this depression was 1 foot 5 inches (43.18 cm)long at the bottom and 1 foot 6 inches (45.72 cm) at the top. Ithad a thickness of 1% inches (4.13 cm), and stood 1 foot 3 inches(38.1 cm) above the floor.The ladder poles (fig. 26, m) were set in the floor 6 inches (15.24cm) from the deflector. The holes were 1 foot 1 inch (33.02 cm)apart, indicating that the ladder was approximately 1 foot (30.48 cm)wide. The posts were 4i/^ inches (11.43 cm) and 4% inches (12.06 cm)in diameter. At the time when the debris was cleared from the floorthe charred butts stood 8i/^ inches and 8% inches (21.59 and 22.23 cm)above the floor.The sipapu holes (fig. 26, i) were 3 feet 5 inches (1.041 m) fromthe fire pit and 6 feet 6 inches (1.981 m) from the wall. The holenearest the fire pit had a diameter of 8 inches (20.32 cm) and a depthof 5 inches (12.7 cm). The second was 3 inches (7.62 cm) from thefirst. It had diameters of 6 inches (15.24 cm) and 9 inches (22.86 cm)and a depth of 1 foot (30.48 cm).The various holes scattered about in the floor had a considerablerange in measurements. The hole d, figure 26, had a diameter of 6inches (15.24 cm) and a depth of 6 inches (15.24 cm) ; e was 4 inches(10.16 cm) in diameter and 8% inches (22.22 cm) deep; g was 31/2inches (8.89 cm) in diameter and 1 foot 2 inches (35.56 cm) deep; hhad diameters of 5 inches (12.7 cm) and 10 inches (25.4 cm), and adepth of 3 inches (7.62 cm) ; r had a 7-inch (17.78 cm) diameter anda depth of 9 inches (22.86 cm)
-, t measured 9 inches (22.86 cm) and101/2 inches (26.67 cm) on two diameters, and had a depth of 11inches (27.94 cm).
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA 115The large depression in the floor at the west side of the fire pit(fig. 26, s) which had been plastered over gave the following meas-urements : The oval portion had diameters of 4 feet 5 inches (1.346 m)and 5 feet 21/0 inches (1.587 m). Its average depth was 2^4 inches(5.71 cm). The two small holes in this portion of the depression haddiameters of 31/2 inches (8.89 cm) and depths of 51/2 inches (13.97 cm)and 9 inches (22.86 cm). The rectangular pit at the side of thecircular depression had sloping sides so that the bottom was smallerthan the top. Along the top it measured 4 feet 7 inches (1.397 m)long and from 1 foot (30.48 cm) to 1 foot 4 inches (40.64 cm) inwidth. At the bottom the length was 4 feet (1.219 m) and thewidth 10 inches (25.4 cm). The average depth was 11 inches (27.94cm). The two small projections at the top of the east edge of thisrectangular depression were 5 inches (12.7 cm) and 6 inches (15.24cm) long. They were 4 inches (10.16 cm) wide. The depth aver-aged 2 inches (5.08 cm). The edge of the rectangular pit was 1 foot2 inches (35.56 cm) from the fire-pit rim.The oval depression at the east side of the chamber which hadflat stones at either side (fig. 26, c) had a long diameter of 2 feet 6inches (76.2 cm) and a short one of 1 foot 4 inches (40.64 cm). Thedepth was 8 inches (20.32 cm).The plaster ridges on the bench in front of the back support posts(fig. 26, q) ranged from 1 foot 8 inches (50.8 cm) to 2 feet (60.96 cm)in length and from 9 inches (22.86 cm) to 1 foot (30.48 cm) inwidth. The height was 2 inches (5.08 cm).The adobe step in front of the ventilator opening (fig. 26, 0) was1 foot 101/^ inches (57.15 cm) long and 10 inches (25.4 cm) wide. Itwas 3 1/2 inches (8.89 cm) high. The small hole in which a post forthe ventilator cover was set had a diameter of 2i/^ inches (6.35 cm)and a depth of 5 inches (12.7 cm). The groove in which the coverstone was placed had a depth of 1 inch (2.54 cm) and a length of1 foot 6 inches (45.72 cm) . It was li/^ inches (3.81 cm) wide.The ventilator opening was square and measured 111/4 inches(28.57 cm). The sill was at the same height as the bottom of thegroove for the cover stone. It was 2i/^ inches (6.35 cm) above thefloor level. The stone used to close the aperture measured 1 foot 3inches (38.1 cm) long by 1 foot % inch (32.38 cm) wide and I14inches (3.17 cm) thick. The ventilator passage was 7 feet 6 inches(2.286 m) long. Where it entered the shaft it was 1 foot (30.48cm) high and 1 foot 6 inches (45.72 cm) wide. The shaft haddiameters of 2 feet 4 inches (71.12 cm) and 1 foot 11 inches (58.42cm). The maximum depth was only 1 foot 6 inches (45.72 cm), butat the time of occupation this probably was about 6 feet (1.828 m).
^^Q BUREAU OF AMERICAN ETHNOLOGY [Boll. 12 1STRUCTUEES 13A AND 13BStructures 13a and 13b were somewhat similar in nature to 5aand 5b in that 13a had been erected at the spot where they were lo-cated, then abandoned and filled in. Later the larger pit was dug atthe same place and completely covered the older pit. In the processof investigation 13b Avas the first encountered. After it had beencompletely cleared a slight depression which encompassed most ofthe central part of the room was noted in the floor. A small sectionof the plaster was removed along the border of this depression and itwas noted that there yxas a wall extending downward from that level.With this discovery the remaining material was dug out and a smallerpit uncovered. Inasmuch as the latter represented the older structureit will be discussed first.The pit for 13a was smaller than the average for the site. It wasroughly circular in form and in general corresponded to the majorityof pits around the site (fig. 30; pi. 13, 5). The house presumably hada superstructure similar to the others. Four posts had been set nearthe corners of the room to support the roof. Each one of these posts,judging from the holes which remained, stood almost against the wall.The interior features of the structure consisted of a fire pit, ladderbox, sipapu, and five storage holes in the floor. In addition therewas a wall pocket or cupboard near the west support post.The fire pit in 13a was a simple basin that had been dug into thefloor and lined with plaster (fig. 30, g). It was encircled by a rim ofadobe plaster similar to those noted in some of the other houses. Theadjacent ladder box (fig. 30, h) consisted of a shallow depression in-closed on three sides by an extension of the fire-pit rim. The fourthside, that toward the ventilator opening, was not inclosed. As a mat-ter of fact a rim was not necessary along that part of the pit becausethe base ends of the ladder had been embedded in the floor. The sipapu(fig. 30, /) was located at the northwest side of the fire pit and con-sisted of a small circular hole. It differed from the many examplesdescribed in that it was jug-shaped. The walls and bottom were care-fully plastered. Nearby were two other holes that could have servedas additional sipapus (fig. 30, d, e) or for storage purposes. At thenorth side of the fire pit was such a hole (fig. 30, c). Near the eastsupport post was a shallow, basinlike depression (fig, 30, h) that nodoubt functioned in a similar capacity. The same was true for thatmarked i, figure 30.Perhaps the most interesting feature in the pit was the storagepocket in the wall (fig. 30, j) . This is one of the few examples at thesite of such a provision for placing objects out of the way. Thepocket was dug into the wall, its bottom slightly lower than the floorlevel. It was roughly oval in form and its sides, floor, and ceiling
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA 117
5 FEET
FiGDEE 30.?Structure 13a. a, holes for superstructure support posts ; h, c, d, and e, stor-age holes ; f, sipapu ; g, fire pit ; h, ladder box ; i, storage basin ; /, storage pocket in wall
;
fe, ventilator opening in chamber ; I, ventilator shaft ; m, accumulated sand. Dotted linesindicate position of 13b.
68764?39-
j^j^g BUREAU OF AMERICAN ETHNOLOGY [Boll. 121were carefully covered with plaster. A slab of stone was provided toclose the opening. This was in position at the time when the pit wasexcavated. There was an almost imperceptible depression in the floorin front of the opening that may have been provided to facilitate theplacing of the cover stone. Wall pockets of this type are not uncom-mon in some sections of the Southwest,^* but for some reason were notmade use of to any extent at this site. In some cases these recesseshave bottoms slightly below the floor ; in others they are slightly abovethe general level of the room. At the time when the cupboard wasopened it contained nothing beyond a few potsherds and several coresfrom which stone chips, probably for use in the making of arrow-heads, had been struck.The ventilator was small and the passage was of the constructedtype, that is a trench had been dug and then covered over withstones, plaster, and earth to form the horizontal portion of theventilator. The opening at the room end was framed with stone.There was a stone lintel and stone sill and large slabs placed ateither side extended back along the walls of the passage for somedistance. The shaft was oval in form. In the photograph (pi. 13, h)the upper end is shown encircled by stones. The latter were notfound in that position but were placed there to emphasize the presenceof the opening.The fill in house 13a indicated that the pit had not been used as ahabitation for some time previous to the construction of 13b. Onthe floor was a thick layer of turkey droppings. Several broken eggswere found near the north post and the skeleton of one bird waslying near the center of the chamber. This indicates that the in-closure was used as a turkey pen. After it had been employed forthat purpose the occupants of the site apparently decided to con-struct a second house on the same spot and accordingly filled thepit with refuse material and clean sand. Although it cannot bestated definitely that such was the case, it would seem that the fillwas obtained through the process of enlarging the pit to make thesecond structure, 13b. When the remaining portion of 13a was com-pletely filled an adobe plaster floor was laid across it. It was the set-tling of the material filling 13a and an attendant sinking of thefloor that indicated the presence of the earlier structure when 13b wasexcavated.The pit for 13a measured 6 feet 1 inch (1.854 m) on the sipapu, firepit, ventilator line, and 9 feet 91/2 inches (2.983 m) at right anglesto it. At the ventilator side of the chamber the floor of 13a was 2M Morris, 1919, p. 196. Guernsey and Kidder, 1921, p. 25. Roberts, 1929, p. 40 : 1930,pp. 49-50.
SoBBmxsj ARCHEOLOGIOAL REMAINS IN ARIZONA 119feet (60.96 cm) below that of 13b. At the opposite side the bottomof pit 13a was only 1 foot 8 inches (50.8 cm) below that of 13b.The holes for the support posts varied somewhat in size. That inthe north corner was irregular in shape and measured 9^/^ inches(24.13 cm) by 1 foot 1 inch (33.02 cm). The hole had a depth of1 foot 1 inch (33.02 cm). At one point on its periphery it touchedthe wall of the chamber. The hole at the east corner had diametersof 8 inches (20.32 cm) and 1 foot (30.48 cm). It had a depth of7^2 inches (19.05 cm). The edge of the hole did not touch the wall,although it was only l^ inch (1.27 cm) from it. The hole at thesouth corner of the room had diameters of 8^^ inches (21.59 cm) and9 inches (22.86 cm). The hole had a depth of 8 inches (20.32 cm)and touched the wall. The west hole had diameters of 1 foot (30.48cm) and 1 foot I/2 inch (31.75 cm). It had a depth of 1 foot 1 inch(33.02). This hole not only touched the wall but was partially in-corporated in it.The fire pit (fig. 30, g) had a diameter of 1 foot 4^ inches (41.91cm) on the ventilator line and 1 foot 4 inches (40.64 cm) at rightangles to it. The pit had a depth of 7 inches (17.78 cm) below thefloor level. This was increased 2 inches (5.08 cm) by the encirclingrim. The rim ranged from 5l^ inches (13.97 cm) to 7l^ inches(19.05 cm) in width. Its average height around the edges of thepit was 2 inches (5.08 cm) ; the outside borders merged with thefloor.The ladder pit, or rather the rectangular inclosure in which thebase ends of the ladder were inclosed (fig. 30, h), measured 1 foot1/2 inch (31.75 cm) by 1 foot 11/2 inches (34.29 cm). The holes inwhich the ladder poles were placed had diameters of 3 inches (7.62cm) and 3^^ inches (8.89 cm). The holes had depths of 3^ and 4inches (8.89 and 10.16 cm). A smaU hole adjacent to the one inwhich the east ladder pole had rested, and that probably was used forstorage purposes, had a diameter of 3^^ inches (8.89 cm) and a depthof 21/2 inches (6.35 cm).The sipapu (fig. 30, /) was 7 inches (17.78 cm) from the fire pitand 3 feet 4 inches (1.016 m) from the wall at the opposite side.The hole had a diameter of 4 inches (10.16 cm) on the floor level,but this was increased to 5i^ inches (13.97 cm) below the floor be-cause of the jug-like form in which it was made. The hole had adepth of 8 inches (20.32 cm). The storage holes located near thesipapu (fig. 30, d, e) had diameters of 6 inches (15.24 cm) and 7inches (17.78 cm). Their depths were 5 and 6 inches (12.7 and15.24 cm).The large depression at the northeast side of the fire pit (fig. 30, c)was 6l^ inches (16.51 cm) from the edge of the rim. It had diam-eters of 1 foot 1 inch (33.02 cm) and 1 foot (30.48 cm). It was 6
^2Q StlREAU Oi' AMERICAN ETHNOLOGY [Soll. 121inches (15.24 cm) deep. The storage hole near the northwest cornerof the pit (fig. 30, i) was 11 inches (27.94 cm) from the edge of therim. It had diameters of 8 inches (20.32 cm) and 81/2 inches (21.59cm). Its depth was 5 inches (12.7 cm). The depression near thehole for the east post (fig. 30, b) had diameters of 1 foot (30.48 cm)and 10 inches (25.4 cm). Its depth was 3 inches (7.62 cm).The storage recess in the wall near the west support post (fig. 30,j) measured 2 feet 2 inches (66.04 cm) by 1 foot 5 inches (43.18 cm).Its height was 1 foot 3 inches (38.1 cm).The ventilator opening had a width of 9 inches (22.86 cm) and aheight of 101/2 inches (26.67 cm). The sill was 51/2 inches (13.97cm) above the floor level. The sill projected into the room 31/2inches (8.89 cm) from the edge of the opening. The passage had alength of 2 feet 91/2 inches (85.09 cm). At the shaft end it had awidth of 9 inches (22.86 cm) and a height of IO1/2 inches (26.67 cm).The shaft measured 2 feet (60.96 cm) in diameter on the passageline and 1 foot 7 inches (48.26 cm) at right angles to it. The bottomof the shaft was 1 foot 8 inches (50.8 cm) below the floor level of 13b.Only a portion of the pit of structure 13b remained. It had beenbuilt on the edge of a sharp slope and weathering agents had washedaway a large part of the surface in the interval between the timewhen it was abandoned and the remains were uiicovered in the sum-mer of 1931. Not only had surface water carried away a consider-able portion of the slope but a large part of the walls along theeastern side of 13b had also disappeared. Enough remained, how-ever, to give a general idea of the nature of the structure (pi. 14, a,h). On the whole it compared quite Closely with the others de-scribed in preceding pages. The main difference between 13b anda majority of the others was that the pit was presumably muchshallower. The old ground level was only 2 feet (60.96 cm) abovethe floor.The shallowness of the excavation may in part account for theapparent ease with which the eastern walls were carried away.Despite the lack of depth to the pit there probably was not anymarked difference in the type of structure erected over it. Evidencewas that four upright posts had carried the main framework (fig.31, a) for the roof and that this covering had a flat central portionand sloping sides. The main difference between the superstructurein this instance and the others of like type probalaly was in the lengthof the slanting side poles. When the pit was shallow these timberswould of necessity have been longer and more of the structure wouldhave appeared above the ground level. Except for this there neednot have been any marked difference. The longer poles would tendto create more pressure on the central framework than under ordinarycircumstances and an additional brace seems to have been required.
Roberts] ARCHEOLOGICAL REMAINS IN ARIZONA 121
W//////////////////AFigure 31.?Structure 13b. o, holes for main support posts ; 6, beam placement ; c and d,storage holes ; e, groove in floor ; /, storage basin ; g, wall pocket or Katcina niche ; li,basin covered by floor plaster ; i, stones embedded in floor ; j, basin in floor ; h, fire pit
;
I, ladder box ; ra, deflector stone ; n and o, storage basins ; P, drift sand. Dotted linesshow position of 13a.
2^22 BUREAU OF AMERICAN ETHNOLOGY [Boi^. 121Evidence was that this was provided by a timber extending fromthe base of the wall at the north side of the chamber to the upperframework. The beam placement was present at 5, figure 31. Thiscorresponds to some extent with the situations described for struc-tures 6, 7, and 9, where the use of braces was noted.The pit of 13b more nearly approached the circular form thandid many of those previously described. The interior features ofthe house comprised fire and ladder pits and a number of storageholes in the floor (fig. 31). There was a small niche or pocket inthe west wall (fig. 31, g), suggestive of the Katcina Eahu in later-day ceremonial chambers. This was in contrast to the placementsfor roof braces noted in previous pits in that it miquestionably wasfor storage purposes. The size, shape, and finish indicated such afunction and despite the fact that it was on the floor level it ma;^have been a form of the Katcina niche,^^ a place where ceremonial ohjects were kept. Such wall pockets also occupied an analogous position to the niche for ceremonial articles in the wall opposite fromthe doorway in some of the Plain's earth lodges.^*'Near the center of the chamber was a small section of paving onthe floor that consisted of four stones (fig. 31, i), embedded in theplaster so that their upper surfaces were flush with the top of thefloor. Between the flagging and the fire pit was a shallow basin(fig. 31, j) , probably associated in purpose with the paving. A me-tate may have rested on the flagging and the concavity in the floorhave served as a storage place for grain or a catch basin for the pre-pared meal dropping from the end of the grinding stone. Thisrouglily oval-shaped depression was of particular interest from aconstructional point of view. It was not a simple basin formed byplastering the interior of a shallow pit dug below the floor level. Ithad been made with a substantial foundation. The latter consistedof an octagonal cist formed by lining a hole in the fill in 13a withsmall stone slabs (fig. 31, ;, section). The interior of the cist con-tained stones and mud plaster, tamped to form a solid, compact mass.All indications were that the latter was contemporaneous with theconstruction of the cist, not the result of later remodeling. Theelaborateness of the feature shows that it had a definite purpose re-quiring a certain degree of durability.The fire pit (fig. 31, k) was rouglily rectangular in form and linedwith stone slabs. The ladder pit (fig. 31, I) was also rectangular inshape and had stone borders on three sides. These stones were setin the floor so that they extended above it for several inches. Thehouse had no sipapu and there were no remaining traces of a ven-tilator. The position which it would have occupied with respect to
^^Mindeleff, 1891, p. 121.""Bushnell, 1922, p. 158.
ROBBBTS] ARCHEOLOGICAL REMAINS IN ARIZONA 123tlie rest of the structure was such, however, that it might easily havebeen carried away when the hillside was eroded. The fact that theend slab in the ladder pit was of sufficient height to have functionedas a deflector (fig. 31, m), suggests that there probably was a ven-tilator at that side of the chamber.The storage holes (fig. 31, <?, d, /, n, o) scattered about in thefloor were larger than average and had carefully plastered sides andbottoms. One depression located near the west wall of the chamberhad been filled in and plastered over (fig. 31, h). The crescent-shaped groove in the floor near the north support post (fig. 31, e) wasan unusual feature. Wliat purpose it may have had is not known.Nothing was found in it to suggest an explanation for its peculiarcontour.The fill in the pit consisted largely of drift sand containing somecharcoal. Across the top was a thin layer of stone spalls sugges-tive of mason's debris from the nearby large stone buildings. At anumber of places around the site the fragments of stone that wereleft after the building blocks were shaped were thrown out and thedeposit here seems to have been of that nature. This would indicatethat 13b antedated the large surface dwellings by an interval of somelength. At one place in the fill a peculiar cache was found. Thisconsisted of a large number of concretions, fossils, bones, some stonecylinders, and sandstone balls, suggestive of a medicine man's para-phernalia. The collection was just below the surface layer of brokenstones and apparently belonged to the stone-house stage ofoccupation.Complete measurements for 13b cannot be given because of the miss-ing east wall. From north to south, however, the structure measured20 feet 3 inches (6.172 m), which is a longer diameter than that foundin a majority of the structures. The wall along the west side was only2 feet (60.96 cm) high, measuring from the floor to the original groundlevel. The accumulation of drift sand and debris above the old groundlevel, however, increased the depth at the time of excavation to 5 feet9 inches (1.752 m).The north hole for a main support post had diameters of 9i/2inches (24.13 cm) and V/2 inches (19.05 cm). The depth was 1foot (30.48 cm). The hole for the east post had been carried awaywhen that portion of the house was destroyed. The hole for the southpost measured 9 inches (22.86 cm) and 10 inches (25.4 cm) in di-ameter. It had a depth of 1 foot 4 inches (40.64 cm). The south poststood 1 foot 1 inch (33.02 cm) from the wall. The hole for the westpost had diameters of 8 inches (20.32 cm) and 9I/2 inches (24.13 cm)and a depth of 1 foot 2 inches (35.56 cm). It stood 10 inches (25.4cm) from the wall. The placement for the roof brace (fig. 31, h) had
j^24 BUREAU OF AMERICAN ETHNOLOGY [Boll. 121diameters of 7 and 9 inches (17.78 and 22.86 cm). It was 10 inches(25.4 cm) deep.The niche in the wall (fig. 31, g) measured 7 inches (17.78 cm) inwidth and 1 foot (30.48 cm) in height. It was 1 foot (30.48 cm)deep.The fire pit (fig. 31, k) measured 1 foot 11 inches (58.42 cm) by2 feet 3 inches (68.58 cm) and it had a depth of 6 inches (15.24 cm).The ladder pit (fig. 31, I) was 7 inches (17.78 cm) from the fire pit.The depression measured 1 foot 11 inches (58.42 cm) by 1 foot 111/2inches (59.69 cm). Its depth was 4 inches (10.16 cm). The slabforming the end of the box toward the side where the ventilatorwould normally have been located stood 1 foot (30.48 cm) above thefloor level (fig. 31, m).The oval depression adjoining the paving (fig. 31 j) had a longdiameter of 2 feet 4 inches (71.12 cm) and a short one of 1 foot 6inches (45.72 cm). At the center the pit had a depth of 81/2 inclies(21.59 cm). The crescent-shaped groove (fig. 31, e) in the floor hadan average width of 21^ inches (6.35 cm). The groove was 3 inches(7.62 cm) deep.The small circular hole adjacent to the groove (fig. 31, d) haddiameters of 6 inches (15.24 cm) and 7 inches (17.78 cm). Its depthwas 6 inches (15.24 cm). The larger, roughly oval-shaped hole nearthe, south end of the crescent groove (fig. 31, c) had diameters of 1 foot4 inches (40.64 cm) and 11 inches (27.94 cm). Its depth was 8 inches(20.32 cm).The plastered-over basin (fig. 31, h) measured 2 feet (60.96 cm)by 1 foot 6 inches (45.72 cm). Its original depth was 6 inches(15.24 cm).The large pit just north of the west support post (fig. 31, n) haddiameters of 2 feet 6 inches (76.2 cm) and 2 feet 2 inches (66.04 cm).Its depth was 8 inches (20.32 cm). The pit southeast from the westsupport post (fig. 31, o) had a long diameter of 2 feet 11/2 inches(64.77 cm) and a short diameter of 1 foot IO1/2 inches (57.15 cm).The depth was 1 foot (30.48 cm)
.
The large pit at the northeast corner of tlie room (fig. 31, /) was 2feet 3 inches (68.58 cm) and 2 feet 2 inches (66.04 cm) in diameter.It had a depth of 2 feet (60.96 cm)
.
STRUCTURE 1 4 AND DANCE COURTStructure No. 14 was located on top of the ridge some distancewest of 12 and southwest from 13a and 13b (fig. 25). The sig-nificant factor about the remains of No. 14 was that they werepartially covered by a rather unusual feature in the form of adance court. The latter had been placed there after structure 14
Roberts ] ARGHEOLOGICAL REMAINS IN ARIZONA 125had gone into ruin and its pit had become filled with debris. Thatthis was not done immediately^, however, was indicated by the pres-ence in the fill of No. 14 of a small stone cist that was subsequent
Figure 32.?Structure 14. a, holes for main support posts ; b, wall pocket ; c, fire pit
;
d, ladder box ; e, ventilator opening ; f and g, exterior opening to ventilator shafth, platform of dance court dais ; i, floor of dance court.to the abandonment of the structure but also antedated the court.The latter probably belonged to a phase associated with the earlystages in the development of the large stone structure.The pit for structure No. 14 was roughly rectangular in shape,although the corners were rounded and the side where the ventilator
^2Q BUREAU OF AMERICAN ETHNOLOGY [Boll. 121opened into the chamber was slightly irregular in the same manneras discussed for most of the other structures (fig. 32) . The pit wasrather shallow and in that respect corresponded to the lack of depthnoted for structure 13b. The pit had been covered with a super-structure presumably similar to the flat-ceiling, sloping-sided typealready described. There were four holes near the corners of thechamber where the main supports had been placed that indicatea roof of the usual form.The interior features were simple indeed (pi. 15, a). There wasa fire pit (fig. 32, c), near the center of the room and not far fromit a ladder box (fig. 32, d). A small pocket in the wall near theeast support post probably served for storage purposes (fig. 32, b).This feature was not as elaborate or extensive as the one in 13aand differed from it in that it was partially below the floor level.There were no storage holes in the floor of the chamber and therewas no sipapu. Both the fire pit and the depression where the baseends of the ladder rested were simple basins cut in the floor andcovered with plaster. No stones were used in their construction.Perhaps the most interesting part of the construction of No. 14was that of the ventilator. The passage leading from the chamberto the shaft was comparatively small and had been tunneled throughthe earth. The shaft at the outer end exhibited a characteristicwhich was unique in that it extended to a considerable depth belowthe bottom of the passage (fig. 32, section). As a matter of factthere was more shaft below than above the level (fig. 32, g). Wliatthe significance or purpose of such an arrangement was is not known.The only explanation that seems logical is that the builders intendedto dig a deeper pit and prepared a ventilator shaft with that ideain mind. After the shaft had been excavated to the desired deptha change in plans was made and the pit for the structure was notcarried to a similar depth. The shaft opening on the ground levelsuggested a figure 8 in contour (pi. 15, b). One segment of the "8"(fig. 32, /) extended only a short distance below the surface whilethe other furnished the main portion of the shaft (fig. 32, ^). Hereagain, as in the case of the unusual depth below the passage, thereis no explanation for the peculiarity of construction.Potsherds found on the floor of the structure indicated that itbelonged to the same general horizon as the other pit remains,namely, early Developmental Pueblo. The overlying dance courtwas late Developmental Pueblo in its horizon.The dance court was so called because no other designation seemedmore appropriate. Whether the place actually was used for such apurpose or not can not be stated. The court consisted of a well-laidadobe floor placed slightly below the surface of the ground (pi. 16, a) .The floor was circular in contour and bordered by a series of small
BOBEBTS] ARCHEOLOGICAL REMAINS IN ARIZONA 127upright stone slabs (fig. 33). Near the center were three depressionsplaced in a row extending roughly on a north-south line. The north-ernmost of these pits (fig. 33, g)^ that was nearest the center of thepaved circle, had been used as a place for fires. The middle one (fig.33, h) was filled with wood ashes but gave no indication that fires
in floor ; f, basin in floor ; g, fire pit ; h, ash pit ; i, basin in floor. Dotted lines show loca-tion of structure 14.had been built in it. It presumably functioned as a depository forthe ashes from the main fire pit. There was nothing whatever toindicate the purpose of the third depression (fig. 33, i) . It was veryshallow and although a few stones had been placed along one sideit contained nothing but clean blow sand. A fourth basin was presentat some distance east of the fire pit (fig. 33, /) . The sides and bottom
J2g BUREAU OF AMERICAN ETHNOLOGY [Boll. 121of this hole in the pavement were carefully plastered. Except foran accumulation of drift sand the pit was empty when the floor wascleaned. At two places stone slabs were embedded in the floor, theirtops flush with the plaster paving (fig. 33, d, e). There was nothingto indicate why they should have been so placed and no evidence asto their function.At the north side of the circle, at a slightly higher elevation, werethe remains of a masonry wall that seemed to have formed part ofan enclosure (fig. 33, a; pi. 16, h). The evidence was not as satis-factory as might be desired but all indications were that the masonryconstituted the remnants of a three-sided structure that inclosed aform of dais along that portion of the circle. The side toward thecourt was open. This raised platform would have been a convenientplace from which to watch dances or ceremonies being performed inthe circle.Neither the dais nor the circle itself gave indications of a super-structure. It is possible that the platform had been covered with aflat roof and all evidence of it had disappeared. It does not seemlikely, however, that the circle could have been roofed over withoutleaving some indications of a superstructure. In view of these factsit seems logical to conclude that the court was not covered and thatall ceremonies or dances held there must have been in the open. Thereis no direct evidence to warrant such a deduction but it is hard toignore the suggestion of the basic plan of a great kiva or super-ceremonial chamber in these remains.^^ The size of the circle, to-gether with the raised platform and inclosure at the north end, bearsa striking similarity to the broad outlines of such structures. Manyof the interior features of the latter are, of course, absent. As apostulation it may be suggested that during Developmental Pueblotimes at this location such ceremonies and observances as later tookplace in the great kiva, one of which is present at the north end ofthe nearby Great Pueblo ruins, were held in the so-called dance court.At only one other location in the Southwest has a comparable fea-ture come to light. It was discovered during the same season, 1931,by Earl H. Morris in the course of his investigations in the ChuskaMountains along the Arizona-New Mexico border. Data concerningthe latter are not available, as yet, and no comparisons can be drawn.From what Mr. Morris has said concerning the remains he discovered,it would seem that there was a close parallel between the courts atthe two sites.Structure 14 measured 10 feet lOVs inches (3.314 m) on the fire pit,ventilator line. Across the opposite direction it was 11 feet 91/2 inches(3.593 m) from wall to wall. At the ventilator side of the chamber the
?' Roberts. 1932, pp. 50, 51, 98 ; discusses the problem of great klvas in general.
BUREAU OF AMERICAN ETHNOLOGY BULLETIN 121 PLATE 15
0. Structure 14.
b. Ventilator shaft for structure 1-1.
BUREAU OF AMERICAN ETHNOLOGY BULLETIN 121 PLATE 16
a. Dance court as seen from end of Great Pueblo ruin.
'.. I'-rM.in o; -inMiiv"- I I iiii<l-ri> :;,^ ^kmce court and dai;.
BUREAU OF AMERICAN ETHNOLOGY BULLETIN 121 PLATE 1 7
(7. Burned roof timbers in pit of structure 15.
b. Series ol slanting side pules fur roof uf structure 15.
lERICAN ETHNOI on- BULLETIN 121 PLATE If
a. Bowl below roof timber? of structure I.'j.
6. Jar containing carbonized ears of c-orn lying on top of roof .iehris m pit of structure 15.
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA 129floor was 1 foot 3 inches (38.1 cm) below the level of the court andat the opposite side 3 feet 3 inches (99.06 cm) below the ground level.The hole for the north post measured 1 foot i/o inch (31.75 cm)on its long diameter and 10% inches (26.67 cm) on the short. Thehole had a depth of 1 foot (30.48 cm) and was 10 inches (25.4 cm)from the wall. The hole for the east support post had diameters of9 inches (22.86 cm) and 11 inches (27.94 cm). The depth was 1 foot3 inches (38.1 cm) and the edge of the hole was 5% inches (13.97 cm)from the wall. The hole at the south corner of the room had diametersof 9 inches (22.86 cm) and 8I/2 inches (21.59 cm). The hole was 6inches (15.24 cm) from the wall and had a depth of 1 foot 1 inch(33.02 cm). The hole at the west corner was the largest of the groupwith a long diameter of 1 foot II/2 inches (34.29 cm) and II14 inches(29.21 cm). The depth was 1 foot 4 inches (40.64 cm) and the holewas 6 inches (15.24 cm) from the wall.The fire pit (fig. 32, c) had diameters of 1 foot 4i/^ inches (41.91cm) and 1 foot 5I/2 inches (44.45 cm). Its greatest depth was 6 inches(15.24 cm). The ladder pit (fig. 32, d) was 7 inches (17.78 cm) fromthe fire pit and measured 1 foot 6 inches (45.72 cm) by 1 foot (30.48cm). It had a depth of 3 inches (7.62 cm).The pocket in the wall near the east corner (fig. 32, h) had a diam-eter of 8 inches (20.32 cm) . The opening in the face of the wall was6 inches (15.24 cm) wide. The bottom was 2 inches (5.08 cm) belowthe floor level and the top 6 inches (15.24 cm) above it. The backwall sloped from the bottom to the top so that at the top of theopening the pocket had no depth.The ventilator opening (fig. 32, e) was 1 foot 3 inches (38.1 cm)wide and 1 foot 1 inch (33.02 cm) high. The passage was 3 feet 9inches (1.143 m) long. Wliere it opened into the shaft the passagemeasured 1 foot (30.48 cm) in width and was 1 foot 2I/2 inches (36.83cm) high. The shaft measured 2 feet 4 inches (71.12 cm) across thepassage line. On the ground line it measured 3 feet 9 inches (1.143 m)from wall to wall. At a depth of 1 foot 6 inches (45.72 cm), however,the diameter at right angles to the passage line narrowed to 2 feet9 inches (83.82 cm). The main part of the shaft had a depth of4 feet 1 inch (1.244 m) below the level of the court at the passage sideand 3 feet 11 inches (1.193 m) at the back wall.The dance court had a diameter on the north-south line of 36 feet8 inches (11.176 m) and on the east-west line measured 37 feet 6 inches(11.430 m). The floor of the court overlapped the house No. 14 pit,2 feet 9 inches (83.82 cm) at the east corner of the older structureand 7 feet (2.133 m) along the southwest wall. The fire pit at thecenter of the floor (fig. 33, {/) was 2 feet 4 inches (71.12 cm) indiameter and had a depth of 41/2 inches (11.43 cm). Between it andthe ash pit was a space of 7% inches (19.05 cm) . The ash pit (fig. 33, h)
22Q BUREAU OF AMERICAN ETHNOLOGY LBdll. 121had diameters of 1 foot 91/2 inches (54.61 cm) and 2 feet 3 inches(68.58 cm). Its depth was 4% inches (11.75 cm). The third depres-sion (fig. 33, i) was 1 foot 4 inches (40.64 cm) from the ash pit. Ithad diameters of 1 foot 51/2 inches (44.45 cm) and 1 foot 113^ inches(60.32 cm). The depth was 6 inches (15.24 cm). The fourth pit (fig.33, /) was 3 feet (91.44 cm) from the central fire pit. It had diametersof 1 foot 101/2 inches (57.15 cm) and 2 feet 2 inches (66.04 cm). Itsdepth was 11 inches (27.94 cm).The remaining wall of the inclosiire (fig. 33, a) at the north sideof the circular floor was 14 feet (4.267 m) long and averaged 10inches (25.4 cm) in width. It was only 1 foot (30.48 cm) in height.The east end of the wall was 4 feet (1.219 m) from the circle and thewest end 7 feet 2 inches (2.184 m) from the pavement. The floor ofthe inclosure had been 1 foot 5 inches (43.18 cm) above that of thecourt. SHELTER AND GRANARIESThe remains of a shelter 22 feet 6 inches (6.858 m) northwestfrom structure 12 and approximately midway between the dancecourt and structures 13a and 13b (fig. 25) were the best preservedof any found. The floor was roughly oval in outline (fig. 34) andalthough depressed below the surrounding ground level was notdeep enough to be called a pit. The low earth walls sloped up fromthe floor level to the ground, giving the depression a saucerlike crosssection. Indications were that the floor basin possibly resulted fromcontinued sweeping of the area with an attendant removal of par-ticles of earth on each occasion, rather than being intentionally dug.The floor was not plastered; the surface was merely hard-packedthrough use. Holes for posts to support a superstructure werepresent at four places near the periphery (fig. 34, a) and indicateda flat-roofed covering. The positions of these holes approximatelycoincided with the four major points of the compass. There wereno traces of slanting side timbers, hence it is not known whetherthe place was merely an arbor or had walls of flimsy construction.Because the floor features here were more elaborate than in the othershelters previously described, the first thought was that the structurewas a shallow-pit dwelling, but when no traces of pole, brush, andplaster walls could be found this idea was abandoned. The structureprobably was an outdoor summer kitchen for one of the nearbydomiciles.Floor features consisted of a fire pit, storage holes, the mold leftby a stone, and a niche in the wall that presumably was a seat fora roof brace. The fire pit consisted of a simple basin dug into thefloor. It was practically circular in form and the sides were morevertical than usual, although the walls did slope in to some extent
Roberts ] ARCHEOLOGICAL REMAINS IN ARIZONA 131
at the bottom. There was no rim around the edges of the pit on thefloor level. When uncovered the basin contained wood ashes andsmall bits of charcoal. The sides and bottom were burned a deep redcolor. Close to the fire pit, on the northeast side, was a peculiarly
vd) o?
^5 FEET
FiGDBH 34.?Brush shelter, a, holes for support posts ; t>, stone in floor ; c, storage pit infloor ; d, fire pit ; e, storage hole In floor ; /, placement for roof hrace ; g, slab m wall.shaped depression and hole (fig. 34, c) that possibly served for stor-age purposes or in connection with a mealing stone. There wasnothing to indicate its exact function. The outer portion was shal-low, the bottom curving down gradually to the edge of the inner
-I Q2 BUREAU OF AMERICAN ETHNOLOGY [Bull. 121hole in funnel-like fashion. The trough extending from one side(fig. 34, c') was blocked off from the main portion of the basin bya small stone set in the earth at that point. There is one suggestionin connection with the trough, namely, that it was a placement for astone, corresponding to that at the opposite side (fig. 34, h). Thelatter gave clear evidence that the base end of a slab, subsequentlyremoved, was embedded there. If a similar stone stood at the formerspot the two would have formed a rough bin beliind the funnel-shaped basin. The small circular hole in the floor at the southeastside (fig. 34, e) was unquestionably for the storage of small objects.The niche, suggestive of a placement for a roof brace, was locatednear the west support post. The bottom of the niche (fig. 34, /) wason the floor level. The sides were vertical, but the back sloped fromtop to bottom at an angle coincident with that which a slantingpole placed there would assume if its upper end joined the frame-work for the roof. A single slab of stone was set in the low wallnear the west post (fig. 34, p-). This was the only example of stonefacing in the structure and at no place around the remainder of theperiphery were there indications that similar slabs had been set butsubsequently removed. There was nothing to indicate why the onewas placed there.On a line through the north and south support posts the diameterof the floor measured 9 feet 7 inches (2.921 m) and from wall towall through the east-west posts was 9 feet 6 inches (2.895 m). Theaverage depth of the floor below the old ground level, that of theperiod of occupation, was 8 inches (20.32 cm). The present surfaceis 3 feet 9 inches (1.143 m) above the floor level on the south sideand 3 feet 2i^ inches (97.79 cm) on the north.The holes for the support posts did not vary greatly in size. Ingeneral they were smaller than those in most of the house structures.That at the north side had a diameter of 7i/2 inches (19.05 cm) anda depth of 1 foot 3 inches (38.1 cm). It was 2i/^ inches (6.35 cm)from the edge of the depression. The east hole was 61/0 inches(16.51 cm) in diameter and 1 foot 51/2 inches (44.45 cm) deep andstood 21/2 inches (6.35 cm) from the wall. The hole at the southside had an 8-inch (20.32 cm) diameter, a depth of 1 foot 61/2 inches(46.99 cm), and was I1/2 inches (3.81 cm) from the edge of tiie pit.The hole for the west post was oval in form, the long diameter beingparallel with the wall, and measured 6 inches by 71/2 inches (15.24by 19.05 cm). The depth was 1 foot 5 inches (43.18 cm) and thedistance from the wall 1 inch (2.54 cm).The fire-pit diameters were 1 foot 6 inches (45.72 cm) and 1 foot61/2 inches (46.99 cm) on the floor level. The sloping of the sidesnear the bottom reduced the diameter to 1 foot 2 inches (35.56 cm).The depth of the pit was 4% inches (11.75 cm).
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA 133The funnel-shaped basin (fig. 34, c) had diameters of 1 foot 41/0inches (41.91 cm) and 1 foot 7 inches (48.26 cm). The greatestdepth was around the edges of the central hole and measured 2%inches (60.32 cm). The central hole had a diameter of 7i/^ inches(19.05 cm) and a depth of 5 inches (12.7 cm) below the bottom ofthe encircling basin. The groove (fig. 34 c') was 7^/2 inches (19.05cm) long by 3 inches (7.62 cm) wide and had a depth of li/4 inches(3.17 cm). The stone at the basin end was 3 inches (7.62 cm) long,J inch (2.54 cm) thick at the west end and IV2 inches (3.81 cm) thickat the east end, and stood 3% inches (9,21 cm) above the bottomof the basin.The long oval-shaped groove (fig. 34, h) that had been the place-ment for a stone slab had a length of 1 foot 1 inch (33.02 cm), awidth of 3 inches (7,62 cm), and ranged in depth from 2I/2 inches(6.35 cm) at the south end to 3 inches (7.62 cm) at the north. Nearthe center its depth was 3^ inches (8.89 cm).The storage hole (fig. 34, e) was 51/2 inches (13.97 cm) in diameterand had a depth of 5 inches (12.7 cm).The roof-brace placement at the west side of the floor (fig. 34, /)was 41/^ inches (11.43 cm) across the opening in the low wall andmeasured 4l^ inches (11.43 cm) from front to back.The mound of accumulated sand, stones, and other material abovethe remains of the shelter suggested a number of interesting itemsin the growth of the site. After the abandonment of the sheltei-the area was covered with a layer of sand deposited, for the mostpart, by wind action. The remains of the superstructure of theshelter may have formed enough of an obstruction to constitute anucleus around which a low dune developed. The area was thenused from time to time as a "workshop" where stones employed inconstruction work were shaped and dressed. Mason's debris wasscattered all through the deposit and that in the lower levels seem-ingly correlated with the similar material, tabular bits of sandstone,noted in the top layer fill in structure No. 12, although the latterrepresented only the outer fringes of the accumulation. The mainstone-dressing activity apparently centered to a large extent in thesection overlying the shelter. After the mound had grown to aheight of approximately 2 feet 3 inches (68.58 cm) above the oldground level, building operations increased and the top 1 foot 3inches (38,1 cm) consisted almost entirely of such debris. This isshown in the section drawings (fig. 35, 2, 3). Potsherds scatteredthrough the layer of stones indicated that the deposit correspondedin time to the interval when considerable construction work wasunder way on the Great Pueblo dwellings.68764?39 10
134 BUREAU OP AMERICAN ETHNOLOGY [Bull. 121One concentration of building stones, in the mound (fig. 35, e, e),presents a puzzling problem. The level where they occurred wasonly 6 inches (15.24 cm) above that of the old ground surface andthey centered about the location of the shelter. When they wereencountered in the excavation work it was thought that they hadbeen used in walls for the structure, but the evidence was otherwise.The layer of wind-blown sand between them and the ground aroundDANCE COURT
SCALE FEET
SCALE FEET
SCALE FEETFigure 35.?Sections for structure 12, dance court and sbelter.the edges of the shelter floor demonstrated that there was a definitechronological break between the two features. To attribute thesituation to coincidence does not seem an entirely satisfactory ex-planation, yet after careful consideration of all aspects of the casethere can be no other conclusion.Farther along the side north from the shelter were the remains oftwo granaries. They were located 24 feet (7.315 m) northwest fromstructure 13b. Fire had destroyed both of them and the pit por-tions were filled with carbonized corn and charcoal from the super-
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA I35structures. In general they were quite similar to the granary pitsdiscussed in preceding pages. Shallow excavations had been cov-ered with a pole, brush, and plaster superstructure. One of the pitswas lined with stone slabs, while the walls of the other were cov-ered with mud plaster. The latter had slumped away from thewalls, hence the exact dimensions of the pit could not be learned.There was nothing in either pit to link them with any particularstructure. Judging from the normal location of granaries, with re-spect to other houses at the site, they would appear to belong with13a or 13b.The stone-lined pit was 5 feet 6 inches (1,676 m) long and 4 feet6 inches (1.371 m) wide. At the south end it had a depth of 10inches (25.4 cm) and at the north end 1 foot (30.48 cm). The sec-ond pit was slightly more pentagonal in outline and measured ap-proximately 5 feet by 5 feet (1,524 by 1.524 m). Its depth was 2feet (60,96 cm) at the south end and 2 feet 3l^ inches (69,85 cm)at the north.The pile of loose rocks located 27 feet (8.230 m) northwest fromthese granary pits (fig. 25) apparently consisted of material gath-ered for use in construction work. The stones were arranged inan orderly fashion, although simply piled there, and were not theremains of a fallen house. Perhaps they had been carried therefrom the sandstone ledges along the edge of the mesa and were wait-ing to be shaped and dressed for inclusion in a structure that wasnever built. There was nothing to show what stage in the historyof the site was represented by this feature, although their locationand the nature of the surrounding earth suggest that it was subse-quent to the existence of the not far distant structure No. 15.STRUCTURE 1 5 AND ASSOCIATED GRANARIESThe remains of structure 15 and the nearby granaries and shelterspresent an interesting example of a prototype for a unit or single-clan form of dwelling (fig. 25). The group as a whole is especiallysignificant from the standpoint of the evolution of a characteristiclate Developmental Pueblo type of house. Structure 15 was a typicalpit dwelling of the form more commonly found in the regionsnorth and east from the Allantown district, particularly the ChacoCanyon area. The pit was rouglily circular in form and had anencircling bench (fig. 36). Both bench and inner pit were dug intothe earth, although the bench was quite shallow. The back wall wasjust high enough to serve as a footing for the sloping side timbersof the superstructure. The top of the bench was covered with adobeplaster, but the back wall was not so treated and its face was merelythe native earth. The wall of the main pit, however, had a facing
136 BUREAU OF AMERICAN ETHNOLOGY [Boll. 321
FiGDKB 36.?Structure 15. a, holes for support posts ; b, stone embedded In bench top
;
c, metate on floor ; d, small storage hole ; e, sipapu ; f, fire pit ; g, basin in floor ; /', firepit ; i, ash pit ; ;, deflector ; k, step and ladder brace ; I, basin in floor ; m, storage hole infloor ; n, storage basin ; o, compartment wall ; p, hole for post to brace cover stoneq, ventilator opening ; r, base of ventilator shaft ; s, ventilator opening on ground
:
*, present surface ; u, sand accumulation ; v, original surface ; ic, wooden support formasonry ; x, floor in granary ; y, surface of accumulated material in shaft.
Roberts] ARCHEOLOGICAL REMAINS IN ARIZONA 137of stone slabs. These were covered with plaster. The latter hadfallen away in some places, although it was still in position aroundmost of the periphery.The superstructure had been of the characteristic flat-center, slop-ing-sided type. It differed from most examples, though, in one fea-ture. There were five main support posts instead of four. Threeof these were set into the face of the bench, while the other twostood out from the wall but were incorporated in the low partitionforming a compartment at the ventilator side of the room (fig. 36, a) .The construction of the superstructure was easily determined herebecause the house had been destroyed by fire. Most of the timberswere only charred, the collapse of the roof smothering the flamesbefore the wood was wholly consumed, and were lying in positionsthat clearly indicated their original locations (pi. 17, a). The fiveuprights carried a series of stringers that formed a pentagonalframework. Several large joists extended across the central portionfrom one side of this framework to the other and in turn supportedsmaller poles. The opening in the roof above the fire pit was rec-tangular in shape, the long measurement being on the fire pit, venti-lator line. Stone slabs framed the opening and formed a protectiveborder around its edges. The sloping ceiling consisted of a seriesof main beams slanting from the back of the bench to the heavystringers on the uprights. Smaller poles were placed, running hori-zontally around the framework, on top of these (pi. 17, &). Theplaster and earth outer covering was applied directly to the timbersto complete the roof.Wlien the structure was destroyed by fire two pottery vessels weresitting on the roof, one not far from the edge of the smoke-hole-hatchway, the other farther down the roof slope. They were found inthe fill just above the burned timbers. One was a bowl (pi. 18, a)and the other a flat-topped jar filled with ears of corn (pi. 18, h).The corn was charred by the heat of the conflagration. Wliy thevessels were sitting on the roof can only be answered by postulation.The surfaces of the low-rounding superstructure mounds probablyserved as gathering places for the families living there and it wouldbe natural that objects, for one reason or another, were occasionallyleft outside. Of course it is possible that an attempt was being madeto salvage belongings from the fire and that these were carried asfar as the roof and then abandoned. On the other hand, one of thewomen of the household may have been returning from the granarywith her jar of corn and the bowl and seeing smoke pouring fromthe dwelling, rushed to determine the cause, setting the vessels downin the process.The interior features of No. 15 were elaborate and more complexthan in many of those previously discussed. The fire pit was dug
T^gg BUREAU OF AMERICAN ETHNOLOGY fBuLU 121into the earth and was roughly oval in contour (fig. 36, h). On oneside two stones were set into the wall to reinforce it, but most ofthe periphery was covered with a heavy coating of mud plaster,burned a deep red in color. A low ridge of plaster encircled the piton the floor level. When the chamber was excavated the pit was halffilled with ashes and some small fragments of charcoal. On eitherside of the pit, in the floor, were two basins (fig. 36, g, I), of unknownpurpose. They had rounded bottoms and were carefully plastered.They may have functioned as pot rests, vessels being placed thereso that the heat from the fire would keep their contents from coolingtoo rapidly.Next to the fire pit, in the position normally occupied by the ladderbox, was a rectangular-shaped basin with rounded ends (fig. 36, i) . Aridge of plaster separated it from the fire pit. At the side towardthe ventilator the wall of the basin was reinforced by a stone slabthat also served as a deflector (fig. 36, j). The basin apparentlyfimctioned as a depository.for ashes, as it was a little more than halffilled with such material when first uncovered. The ashes could notbe attributed to the burning of the superstructure as they were fineand powdery and free from charcoal fragments. They were the kindproduced by long and continuous burning of wood in a pit and un-questionably were intentionally placed in the basin. A similar featurewas noted in some of the other houses or structures.The ends of the ladder rested against a large stone set just insidethe compartment (fig. 36, k). The poles were not embedded in thefloor, but they had worn small, cuplike depressions in the surface ofthe plaster. The bracing stone was low and its top probably servedas a step for anyone ascending or descending the ladder. Placed as itwas, the ladder blocked the opening into the compartment and onentering or leaving the chamber it would have been necessary for oneto step over the low partition. The position of the deflector stone(fig. 36, j) between the ladder and the fire pit was not common. Amajority of the structures showed the reverse, the deflector beingbetween the ladder base and the ventilator opening.In the main part of the chamber there were two storage holes inthe floor. They were on opposite sides of the fire pit and its adjacentbasins (fig. 36, d, m) . They were similar in form to those describedfor other houses. The sipapu was present in the customary position(fig. 36, e) and exhibited no unusual features. At the west side of thechamber, close to one of the main supports, was a large basin in thefloor that slightly undercut the wall above it (fig. 36, n). The sidesand bottom were carefully plastered and the top was encircled by a
? low ridge of plaster that raised the rim somewhat above the floorlevel. There was nothing in the basin at the time of excavation to
Roberts] ARCHEOLOGICAL REMAINS IN ARIZONA 139show the purpose. It would have been a convenient place to keep avariety of tilings, however.The metate or milling stone on the floor at the east side of thechamber (fig. 36, c; pi. 19, a) was in position when found. The heatfrom the burnmg roof timbers cracked the stone, but the pieces hadnot fallen apart. One hand stone (mano) was lying in the trough.Several additional manos were on the floor at one side of the metate.They varied from coarse, through medium, to fine in quality and nodoubt were used for grinding meal of different grades. The metatestone was raised above the floor by a base consisting of three stones.This additional height would make it possible to set a shallow basketor tray under the open end of the trough and catch the meal as itdropped from the grinding stone. Metates frequently occupied sucha position in houses of this period. It was not until the following orGreat Pueblo era that the practice developed of placing several suchstones in a bin.The compartment at the southeast side of the chamber was formedby a series of stone slabs placed on end in the floor. Several werebroken when the roof crashed in and others were cracked from theheat. They originally had been of approximately the same height,any discrepancy being compensated for by the heavy coating ofplaster that covered the stones. There was a break or doorway to thelow wall at the place where the ladder stood. There were no interiorfeatures in the compartment.The ventilator was well made and had several features that differedfrom those previously described. In common with several, however,it had been reduced in size. The original opening into the chamberwas considerably larger than that of the remodeled form. The firstaperture had a framing of stone; the second and smaller one wasentirely of plaster. In the final form there was a step or ledge thatformed a rest for a stone cover for the opening. This ledge was onthe level of the original floor of the passage. Just in front of thevent, in the compartment floor, was a hole in which a small post hadbeen set to hold the cover stone in place (fig. 36, ^). The front wallof the bench above the opening projected into the room to a degreenot noted elsewhere (fig. 36, section). Whether tliis was intentionalor had resulted from a sagging and displacement in the bench, withattendant repair work, is not Imown. The front edge was reinforcedwith a short log of wood (pi. 19, &), but it could not be determinedwhether it dated from the remodeling of the ventilator or had beenplaced there at some subsequent time. The wood was charred fromburning and some of it fell away when the debris was being removedfrom the pit. The full length had carried it w^ell across the openingand about midway of the upright stone slab that formed part of theoriginal aperture (pi. 19, 5), the stone at the right in the photograph.
. .Q BUREAU OF AMERICAN ETHNOLOGY ^^^^^ 1^1The passage was short and also had been reduced in size. This wasaccomplished through use of large quantities of mud plaster. Thewidth was narrowed and the floor raised to a higher level. It was notpossible to determine the status of the original passage roof becausethe old one was removed and a new one installed. The original floorsloped upward to the bottom of the shaft at the outer end, while thelater one was almost level. The passage was of the constructed type.The shaft at the outer end had several features differing some-what from others previously described. The lower and main partwas oval in outline and much larger than many of those discussedin preceding pages. The walls from the old ground level to thebottom of the shaft were native earth covered with plaster. Severalcourses of stone were laid around the aperture on the surface, pre-sumably to heighten the shaft. A similar condition, although insmaller degree, was mentioned in connection with structure No. 2.The presence of the masonry may indicate an attempt to preventdriftmg sand from falling into the opening and blocking the shaft,or an effort to improve the ventilating functions. The extensivedome-like structure shown in the upper section of the drawing(fig. 36) does not belong to the dwelling proper. It is of lat^rdate. Subsequent to the destruction of the house by fire theventilator shaft drifted full of sand to the height of the old groundlevel, as indicated by the dotted line in the shaft (fig. 36, y). Thecircle of rocks remained and was used as the foundation for astorage structure. The latter was built of stones laid in progres-sively contracting courses to produce the beeliive-shaped structureuncovered in the digging. Before erecting the upper part, dump-heap material was placed in the bottom to level and raise the floor.The latter consisted of adobe plaster laid on top of the fill andsmoothed to form an even surface (fig. 36, x). There was nothingto show the stage to which this structure belonged. Wliether it waserected shortly after the destruction of the house or following aninterval of some length is not known. That the upper courses wereof later date was clearl}^ shown by a difference in the color of themud mortar used in the masonry. That in the original sectionwas a dark yellow, discolored by smoke stains, while in the upperpart it was red and not smudged. The entire dome was coveredwith plaster when completed. ]Most of it had fallen away, how-ever, and only patches were clinging here and there when the re-mains were uncovered. The encircling courses of stone that werelaid around the top of the shaft in the first place were an additionmade after the completion of the ventilator in the original form.The stones in the arc on the house side of the shaft rested partially
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA 141on the ground and partially on two timbers (fig. 36, w) thatextended across that part of the excavation. The wood was in anadvanced stage of decay when found and was only held in placeby surrounding dirt. There were no traces of charring, indicatingthat the flames in the burning structure had not touched them. Asignificant factor in this connection is that the plaster on the wallsof the original part of the shaft showed direct effects of fire, whilethe stones above did not. The latter were smoke stained, but theyhad not come into actual contact with flames. This suggests thatthe structure was damaged to some extent on an occasion prior tothe addition of the stones to the shaft and later final destructionof the dwelling. The last conflagration probably produced thesmoke stains on the stones. There were some indications in thedwelling proper of an earlier period fire to substantiate those in theventilator shaft. Several places where the plaster had broken awayfrom the face of the bench and from the bench top, the underlyingmaterial gave evidence of burning. These lead to the conclusionthat there were two periods of occupation, each terminated by fire.
^Lfter the first the structure was rebuilt and after the second it wasabandoned.The shaft in its original status "v\'as large enough to serve as anantechamber and with the unreduced passage could have been theentrance to the house. When the passage was made smaller it wouldnot have been possible for a person to crawl through the apertureinto the room. The conditions warrant the suggestion that in thishouse there is a record of the shift from antechamber, passageentrance to access by means of a ladder through the smoke-hole-hatchway in the superstructure. This shift, if such actuallyoccurred, correlated with the first burning of the house and subse-quent remodeling.The fill in the main part of No. 15 consisted largely of charredtimbers, burned plaster, and drift sand. At no time was the pitused as a dumping place for refuse. Wlien the hole had filled tothe old original ground level a definite surface of occupation de-veloped and a fire pit was placed there (fig. 36, /). This pit waslined with stone slabs and had a large stone for its bottom. Theslabs around the sides projected slightly above the ground level,with the result that the fire basin was deeper than the hole in whichit was placed. This pit probably corresponded in time to the gran-ary or bin erected above the old ventilator shaft. The accumula-tion of sand above the fire pit and the beehive-shaped structure ap-parently dates from a period subsequent to the abandonment of both.The quantity of burned timber in the pit of No. 15 supplied amplematerial for dendrochronological studies and determining the age
^^2 BUREAU OF AMERICAN ETHNOLOGY IBdlu 121of the structure. The latest date obtained was 888, and the earliest857, a spread of 31 years.^^ That the life span of the structureshould be considered as of that duration is questionable. The logsof earlier date may have been taken from another structure. Thereis, of course, a possibility that they were part of the original super-structure and were employed in the rebuildmg that took place whenthe ventilator was reduced in size and the encircling stones laidaround the ground-level opening at the top of the shaft. To assumethat such was the case is not warranted by the evidence, but it doessuggest an interesting possibility. On the basis of the latest dateit can be said definitely that the house was occupied, that is its finalperiod of mhabitation, circa 888. This antedates by 30 years struc-ture No. 12 and its 918. There was an overlapping in dates for cer-tain logs in the two, however, as was mentioned in connection withthe discussion of No. 12. How long both were occupied from theperiod of their latest dates until destroyed by fire is not known.That No. 15 was burned before No. 12 is indicated by the differencein potsherds found on the floors of each. Those from No. 12 defi-nitely represent a later type.Structure No. 15 had a diameter of 13 feet 1 inch (3.987 m) belowthe bench on the ventilator, fire pit, sipapu line, and 19 feet 1 inch(5.816 m) above the bench. At right angles to this line the innerdiameter, below the bench, was 15 feet (4.572 m) and the outer, in-cluding the bench top, measured 20 feet 2 inches (6.146 m). On theventilator side of the chamber the top of the bench was 2 feet 8^/2inches (82.55 cm) above the floor. At the opposite side it was 3feet 1 inch (93.98 cm). The back wall of the bench was 6 inches(15.24 cm) high. The floor at the ventilator side was 3 feet 9 inches(1.143 m) below the original ground level and 8 feet 7i/4 inches (2.628m) below the present surface. At the opposite side of the room thefloor was the same distance below the old ground level, 3 feet 9 inches(1.143 m), but was only 7 feet 9i/^ inches (2.374 m) below the recentlevel.The holes for the support posts varied somewhat in size, but thiswas in part due to the edges being broken at the floor level. Thetimbers used for the main uprights apparently were about the samesize. They seem to have approximated closely a 6i^-inch (16.ol-cm)diameter. The post at the north side of the chamber was set intothe face of the bench for a little over half of the diameter. Theplaster of the bench had touched the timber all along its surfaceand showed the log had a 6-inch (15.24-cm) diameter. The hole inthe floor at the base of the bench was slightly oval in form withG-inch (15.24-cm) and 7-inch (17.78-cm) diameters. The depth was
^ MUler, 1934, p. 16, designated as House 4/32.
ROBERTS] AKCHEOLOGICAL REMAINS IN ARIZONA 143only 6 inches (15.24 cm). The enclosing bench wall furnished con-siderable support for the timber, hence greater depth was probablynot essential. The hole for the support post in the east compart-ment wall measured 6 inches (15.24 cm) by 10 inches (25.4 cm).The depth was 1 foot 9 inches (53.34 cm), a much deeper placementthan that in the bench. The post in the west compartment wall wasset in a hole with diameters of 8^/2 inches (21.59 cm) and 9i^ inches(24.13 cm) and a depth of 1 foot 10 inches (55.88 cm). The postat the southwest side, in the bench face, had a diameter of 6V2inches (16.51 cm). The charred butt was still in place and was notremoved, so that the depth of the hole is not known. The timberin the western arc of the bench face also measured 6I/2 inches (16.51cm) in diameter. The hole in the floor was 6 inches (15.24 cm) indepth.The foe pit (fig. 36, h) measured 2 feet 6I/2 inches (77.47 cm) onthe sipapu, ventilator line and 2 feet 1 inch (63.50 cm) across in theother direction. At the sipapu side it had a depth of 10 inches (25.4cm) and at the ventilator side 1 foot 1 inch (33.02 cm). The plasterridge separating it from the basin containing ashes was 1 inch(2.54 cm) thick at the narrowest point. The ash basin had a lengthof 1 foot 7 inches (48.26 cm) and a width of 111/2 inches (29.21 cm).At the fire-pit side it had a depth of 1 inch (2.54 cm) and in frontof the deflector stone it was li/^ inches (3.81 cm) deep. The deflectorstone (fig. 36, j) measured 8 inches (20.32 cm) by 2 inches (5.08 cm)and stood 1 foot (30.48 cm) above the floor.The sipapu (fig. 36, e) was 2 feet 3i/^ inches (69.85 cm) from thefire pit. It was slightly oval in form with diameters of 7 and 8 inches(17.78 and 20.32 cm). The depth was somewhat greater than inmany of the examples and measured 9% inches (24.76 cm).The hole in the floor (fig. 36, d) had a diameter of 5 mches (12.7cm) and a depth of 2 mches (5.08 cm). The hole was practicallymidway between the fire pit and the wall, being 3 feet 2 inches (96.52cm) from the latter and 3 feet (91.44 cm) from the former. It wasonly 9 inches (22.86 cm) from the end of the metate stone. The holein a corresponding location at the opposite side of the room (fig.36, m) had a diameter of 3 inches (7.62 cm) and a depth of 4 inches(10.16 cm). Tliis hole was 2 feet 3 inches (68.58 cm) from the firepit and 3 feet 3 inches (99.06 cm) from the wall.The basins on either side of the fire pit (fig. 36, g, I) differed insize. The length for g was 1 foot 10 inches (55.88 cm), the width 5I/2inches (13.97 cm) at the small end and 6i/^ mches (16.51 cm) at thelarge. The average depth was 5 inches (12.7 cm). This basin was61/^ inches (16.51 cm) from the fire pit. Basin I had a length of 1 foot7 inches (48.26 cm), a width of 9i/^ inches (24.13 cm) at the broadend and 8i/^ inches (21.59 cm) at the narrow. The depth averaged 4
jL^ BUREAU OF AMERICAN ETHNOLOGY [Bull. 121inches (10.16 cm), and the distance between the basin and the firepit measured 81/2 inches (21.59 cm).The storage place at the base of the wall (fig 36, n) had diametersof 1 foot 3 inches (38.1 cm) and 1 foot 4 inches (40.64 cm). Itmidercut the face of the bench I1/2 inches (3.81 cm) . The basin hada depth of 4 inches (10.16 cm). The plaster rim around the topaveraged 3 inches (7.62 cm) in width and at the edge of the basinwas 1 inch (2.54 cm) above the floor level.The metate (fig. 36, c) had a maximum length of 1 foot 81/2 inches(52.07 cm) and a maximum breadth of 1 foot 41/2 inches (41.91 cm).The trough in the stone measured 1 foot 1 inch (33.02 cm) long on amedian line and was 10 inches (25.4 cm) wide. The bulk of thestone was 5 inches (12.7 cm) thick and it was raised 3 inches (7.62cm) above the floor level.The stones forming the walls of the compartment varied in height,width, and thickness (fig. 36, <?). Some of the stones had been brokenwhen the roof fell, so that the measurements given are not in all casestheir original height. At the southwest corner of the compartmentthe wall height was 2 feet 5 inches (73.66 cm) and at the northeast endwas 1 foot 11 inches (58.42 cm). These measurements include theadobe plaster covering over the stones. Beginning at the northeastend of the compartment wall (fig. 36), the stones measured:First, 1 foot 6 inches (45.72 cm) long, 2% inches (6.98 cm) thick,and 1 foot 9 inches. (53.34 cm) high.Second, 9 inches (22.86 cm) long, 1 inch (2.54 cm) thick, and 1foot 8 inches (50.8 cm) high.Third, 9 inches (22.86 cm) long, 21/2 inches (6.35 cm) thick, and8 inches (20.32 cm) high.Fourth, 81/2 inches (21.59 cm) long, 31/0 inches (8.89 cm) thick,and 1 foot 2 inches (35.56 cm) high.Fifth, 5 inches (12.7 cm) long, 3yo inches (8.89 cm) thick, and 1foot (30.48 cm) high; the opening in the wall in front of the ven-tilator was 1 foot 9 inches (53.34 cm) wide.The sixth stone was 3yo inches (8.89 cm) long, 2 inches (5.08 cm)thick, and 8 inches (20.32 cm) high.The seventh measured 71/0 inches (19.05 cm) long, 31/2 inches(8.89 cm) thick, and 1 foot 4 inches (40.64 cm) high; the small brac-ing stone next to No. 7 was 5 inches (12.7 cm) long, 3 inches (7.62cm) thick, and 3 inclies (7.62 cm) high.The eighth stone was 1 foot (30.48 cm) long, 1 inch (2.54 cm)thick, and 1 foot 10 inches (55.88 cm) high.The ninth was 91/2 inches (24.13 cm) long, 21/0 inches (6.35 cm)thick, and 2 feet 1 inch (63.5 cm) high.The tenth stone was 1 foot (30.48 cm) long, 214 inches (5.71 cm)thick, and 2 feet 3 inches (68.58 cm) high.
ftoBEBTs] ARCHEOLOGICAL remains in ARIZONA ;[45The ladder brace, just inside tlie compartment (fig. 36, h) was 1foot 9I/2 inches (54.61 cm) long, 4 inches (10.16 cm) thick, and rose6 inches (15.24 cm) above the floor.The hole where a small pole had been set as a reinforcement forthe end of the compartment wall at the southwest side of the door-way (fig. 36) was 3I/2 inches (8.89 cm) in diameter and had adepth of 3 inches (7.62 cm). There was nothing to indicate whatthe height of the pole had been. The hole in front of the ventilatoropening, where a post to hold a cover stone in place over the venthad been set, measured 2 inches (5.08 cm) and 3 inches (7.62 cm)in diameter and had a depth of 2 inches (5.08 cm). The hole(fig. 36, p) was against the riser to the vent sill.The ventilator aperture, in its final form, was 10 inches (25.4 cm)wide and 9 inches (22.86 cm) high. The sill was 5 inches (12.7 cm)above the compartment floor. The offset for the cover slab was 3i/^inches (8.89 cm) deep at the sill. The original opening into the pas-sage was 1 foot 6 inches (45.72 cm) wide. The height could not bedetermined. The passage was 3 feet 2 inches (96.52 cm) long.Wliere it entered the shaft it was 1 foot 2 inches (35.56 cm) wideand 1 foot 1 inch (33.02 cm) high. The original width was 1 foot6 inches (45.72 cm). The shaft had a diameter of 3 feet 11 inches(1.193 m) on the passage line and 3 feet 2 inches (96.52 cm) atright angles to it. The bottom of the pit was 2 feet 11 inches (88.9cm) below the original ground level and 8 feet 5 inches (2.565 m)below the present surface. The circle of stones first placed aroundthe opening on the surface was 1 foot 2 inches (35.56 cm) high. Thecompleted structure (fig. 36, s) rose 3 feet 6 inches (1.066 m) abovethe ground level. Inside the total height, from the adobe floor tothe opening, was 2 feet 11 inches (88.9 cm). The opening at the topwas approximately 2 feet (60.96 cm).The fire pit built on top of the fill in the pit of No. 15, and datingfrom the same period as the stone structure above the vent shaft,was 2 feet 3 inches (68.58 cm) by 2 feet 5 inches (73.66 cm)and had a depth of 1 foot 2 inches (35.56 cm). Tlie edges ofthe stone slabs lining the pit projected above the surface of thelevel of occupation 3 inches (7.62 cm). The bottom of the pitwas 2 feet 9 inches (83.82 cm) above the floor of the dwelling. Thetop of the pit was 2 feet 2 inches (66.04 cm) below the presentsurface.The group of granaries associated with structure No. 15 consti-tuted one of the more significant features at the site. They were lo-cated 19 to 25 feet (5.791 m to 7.620 m) north and west from thedwelling and were built in an orderly row (fig. 37). The pit por-tions were rectangular and there was evidence to show that theyhad been covered with pole, brush, and plaster superstructures of the
146 BUREAU OF AMERICAN ETHNOLOGY [Bull. V^^truncated pyramidal type. The walls in this type had only a sligiitslant and the roofs were flat. No internal supports were used. Thecorner posts were placed just at the edge of the pit and carried a
: A \.^'^"\/ B /
V''
/ E /
'\ H
6 \. ^-^
\\
^'SHELTER
^^4>^''
8^ \
S^^-.^^a'' SHELTER
lOFEET
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA 147for use elsewhere. The outlines of the pit were not sharp and clearas the earth walls had slumped in places and traces of the stones thatmi^aht have been there were lost. This structure (fig. 37, G) wasfilled with corn when it burned and several inches of charred grainand cobs covered the floor. Granary E also burned and some of thestones were missing from its pit. There were signs on the walls,though, to show that stones were used there. In several of the ex-amples a number of courses of horizontally laid stones were usedto augment the height of the slabs and approximately half of onewall in another (fig. 37, B) consisted of this type of masonry. Thefloors of a number of the pits (J., Z>, and G) were completely pavedwith slabs, while in others scattered flagstones were embedded in theplaster. Two of the structures showed that small fires had beenkindled on the floor near the center of the pit, although there wasno actual fire basin present. This is an indication that on occasions,when they were not fulfilling their functions as storage places, thegranaries were used as temporary habitations.Two surface shelters were associated with the granaries. One ofthese was erected in front of the two at the north end of the row(fig. 37, J., B; pi. 20, a). An area on the surface was enclosed by aseries of stone slabs set upright in the ground (fig. 38) . In the centerof this space was a stone-lined fire pit. Six upright posts, two placedmidway along the front and back sides and four near the corners,had supported a rude arbor, probably a brush roof similar to thosepreviously described. Except for the sides of the granaries, whichwould have served in that capacity at the back of the enclosure, therewas nothing to indicate the former existence of any kind of walls.The rude structure thus formed would shelter the people from thesun and give some protection during light showers, but it would nothave been very serviceable in times of severe storms. The remainsof a similar shelter were present at the south end of the group. Thelatter was not as well preserved, although there were indications thatthe floor had been enclosed by slabs and that six posts were used tosupport the covering. The fire pit associated with this second struc-ture was not in the enclosed space but just outside, between it andthe dwelling. The important factor about this group of granariesis the manner in which they foreshadow the row of contiguous roomsand associated ceremonial chamber in the unit-type house of subse-quent years. When the general plan and grouping of the No. 15complex is compared with others discussed in following pages thisbecomes more apparent.The granary pits varied somewhat in size and depth. A was 5feet 10 inches (1.778 m) long and 5 feet (1.524 m) wide. At its northend it was 2 feet 6 inches (76.2 cm) deep and at the south end2 feet 6 inches (73.66 cm). The distance between it and B was 1
148 BUREAU OF AMERICAN ETHNOLOGYB A IBCLU 1210=^-
SHELTER
SFEET
FiGDEB 38.?Granaries A and B and shelter in No. 15 group, a, hole for support post
:
b, fire pit ; c, hole for support post ; d, present surface ; e, sand fill ; i, original surface
;
g, holes for corner supports for arbor.
HOBBRTsl ARCHEOLOGICAL REMAINS IN ARIZONA 149foot 7 inches (48.26 cm). B was 6 feet 4 inches (1.930 m) long, 4feet 9 inches (1.447 m) wide, and 1 foot 3 inches (38.1 cm) deep atboth ends. O was approximately 6 feet (1.8i28 m) square, but exactmeasurements could not be determined. The depth at the ends of thepit was 2 feet 8 inches (81.28 cm). D was 7 feet 3 inches (2.209 m)long, 6 feet 4 inches (1.930 m) wide. The depth at the north endwas 1 foot (30.48 cm) and at the south end 9 inches (22.86 cm).That was the shallowest pit in the group. E was 1 foot 6 inches(45.72 cm) from Z>, had a length of 7 feet 6 inches (2.286 m), awidth of 4 feet 6 inches (1.371 m), and a depth of 2 feet (60.96 cm)at both ends. F was only 1 foot 3 inches (38.1 cm) from E. Itslength was 6 feet (1.828 m), its width 4 feet 4 inches (1.320 m),the depth at the north end 2 feet 2 inches (66.04 cm) and at thesouth 2 feet 8 inches (81.28 cm). G was 19 feet (5.791 m) from F.The length of G was 5 feet 6 inches (1.676 m) and the width 4 feet6 inches (1.371 m). The depth was 1 foot (30,48 cm) at both ends.H was 8 feet (2.438 m) from G. The stone had been removed fromthe sides of the pit but its outline was distinct. It had a length of8 feet 2 inches (2.489 m) and a width of 5 feet 6 inches (1.676 m).Like G^ the depth at each end of H was 1 foot (30.48 cm)
.
The shelter in front of granaries A and B covered an area 10 feet(3.048 m) by 12 feet (3.657 m). The enclosing slabs averaged 8inches (20.32 cm) in height. The posts that supported the arboraveraged 6 inches (15.24 cm) in diameter. The fire pit near thecenter of the floor space had diameters of 1 foot 10 inches (55.88 cm)and 2 feet (60.96 cm) and was 10 inches (25.4 cm) deep. The shel-ter between granary G and the dwelling had a floor area 9 feet(2.743 m) by 12 feet (3.657 m). The posts used to support thearbor averaged 3 inches (7.62 cm) in diameter. This was about halfthe size of those noted for the other structure and the smallestnoted for the entire site. The arbor that they supported could nothave been as heavy as that for other similar structures at this loca-tion. The stones around the floor area, that is, those still in posi-tion, indicated that the wall height for the enclosure had averaged1 foot (30.48 cm).STRUCTURE 16 AND ASSOCIATED GRANARIESThe group corxsisting of No. 16 and its adjacent storage structures(fig. 39), represented an earlier stage in the sequence than the No.15 group. Some of the features noted in the latter were presentin the Nca 16 assemblage in cruder and less developed form andclearly showed certain steps in the evolution of characteristic ele-ments in the unit-type structure. The granaries were built in arow to the west and south of the house and differed to some extent68764?39 11
150 BUREAU OF AMERICAN ETHNOLOGY [Bull. 121
??'; c
10 FEET
( I I I I I J
FiGUKB 39.?Plan of No. 16 house and granary assemblage.
Roberts] ARCHEOLOGICAL REMAINS IN ARIZONA 151iiom those described for the precedmg group. The main interesthere, however, is in connection with the mode of entrance to thedwelling structure. This was not through the roof, as in most ofthose previously described, but through the opening at the south-east side that generally is considered as the ventilator.Structure No, 16 was of the semisubterranean type with a roughlycircular pit and bench (pi. 20, h). Both pit and bench were duginto the earth. As in the case of No. 15, the excavation for thebench was shallow and its back wall irregular, so much so, in fact,that the top of the bench was much wider in some places than inothers. The walls of the main pit were not lined with slabs; theywere simply covered with a thick coating of plaster. The super-structure was presumably of the truncated pyramidal type supportedby a rectangular framework resting on uprights. There had beenfour main supports, but one had been reinforced by another smallerpost. Two of the posts, one being that with a secondary brace, wereset in the floor some distance from the wall of the main pit, whilethe other two, at the compartment side, practically touched the faceof the bench. A few of the sloping side poles had their ends em-bedded in the earth at the back of the bench to hold them in posi-tion. Others merely rested in the angle formed by the back walland the top of the bench. The portion of the periphery where thetimbers were set in holes was that where the back wall of the benchwas very low and the additional feature of the embedded ends mayhave been a necessary precaution against possible slipping when thesuperstructure was covered with earth and plaster. Only a few ex-amples of the setting of sloping side poles in this manner werefound at this site, although it was a common practice in otherlocalities.Interior finishings were not as complex as in some of the struc-tures. The fire pit consisted of a simple circular basin (fig. 40, e)near the center of the floor space. No stones were used in its con-struction and the sloping sides and bottom were covered with plaster,a continuation of the layer that formed the floor. There was noladder box, for reasons associated with the difference in the modeof entrance, and no basin for ashes. There was no separate de-flector, the compartment wall functioning in that capacity. Therewere only a few holes in the floor. One of these (fig. 40, c) was thesipapu, one was for storage of small articles (fig. 40, &), and thetliird, more of a depression or small basin than an actual hole,was probably a pot rest (fig. 40, d). It was so placed that a round-bottomed vessel set there would not only maintain an upright posi-tion but it was close enough to the fire to keep its contents warm.The sides of the sipapu and the storage hole were vertical and bothsides and bottom in each were carefully plastered. In the case of
]^52 BUREAU OF AMERICAN ETHNOLOGY fBcLL. 121the storage hole this must have been difficult to accomplish as it wasof small diameter and it would have required a small hand to workinside and properly apply the plaster.There had been a compartment at the southeast side of the cham-ber, but most of the stones used in its wall had been removed. Thispossibly took place when the structure was abandoned. The moldswhere the stones stood were much in evidence, however, and the out-line of the wall was easily traced on the floor (fig. 40, g) . Six slabshad been used to form the wall and their surfaces and the spacesbetween were covered and filled with adobe plaster. At the west endof the wall a space large enough for an additional stone was entirelyblocked in with the mud mortar. The wall had been continuous andsince there was no break or opening near the center, between the firepit and the aperture to the vent passage, a deflector stone was notneeded. There was a small bin in the east end of the compartment.It was formed by cutting into the face of the bench and by blockingoff the end of the compartment by the use of additional stone slabs(fig. 40, /). The floor in the bin was considerably higher than eitherthat of the compartment or the main part of the chamber. The levelwas raised through the use of stones and mud mortar. The bottomof the bin was partially paved with small slabs set in the plaster.There was a shallow basin in the compartment floor directly in frontof the vent-passage opening (fig. 40, h). In this instance it wasnot a post placement of the type described for some of the otherstructures, yet it could have served as a rest for a brace leaned againsta stone slab blocking the aperture. It also might have functioned asa rest for a round-bottomed vessel, though its position would seem tobe somewhat inconvenient for such use. What its actual purpose wasis not known.A stone was set in the floor in the opening into the vent passage(fig. 40, i) . Wliether this was a sill placed there to protect the edgeof the passage floor, which was at a higher level than that of the com-partment, or had risen some distance above the floor as a sort ofsecondary deflector could not be determined. The stone was brokenand there was nothing to indicate its height at the time of occupancy,although the edge of the passage floor indicated that it had at leastbeen flush with its surface if not somewhat higher. As previouslymentioned, this passage had constituted the entrance to the dwelling.At its outer end it was wider than at the opening into the pit and itsfloor sloped upward from the room. Embedded in the floor at theouter end were the decayed butts of ladder poles (fig. 40, k) and forthis reason it is concluded that the feature was an actual entrancerather than a ventilator. The side and end walls of the passage didnot rise vertically. They sloped outward so that the passage waswider at the top than at the bottom. This would have facilitated
Roberts] ARCHEOLOGICAL REMAINS IN ARIZONA 153
FiGOUH 40.?structure 16. a, holes for main roof supports ; i, small storage hole ; o,sipapu ; d, pot rest ; e, fire pit ; f, storage bin ; g, groove for compartment wall ; h, basinin floor ; i, sill iu passage opening ; j, passage ; k, holes for ladder poles ; I, placements forside poles ; m, present surface ; n, old surface ; o, sand fill.
-.^^ BUREAU OF AMERICAN ETHNOLOGY [Bull. 121movement, as greater space would be provided for the body andshoulders of the person using it. The ladder had sufficient slant tomake ascent and descent easy. The passage was of the trench typeand had been covered with small poles along that part of its lengthwhere it cut through the bench, thus forming a short tunnel. Thecross timbers were probably covered with adobe plaster, possibly acontinuation of the surface extending along the bench top. Theouter end was in the highest point of ground in the vicinity of thestructure and surface water would have drained away from it on allsides. Everj^thing considered, this construction undoubtedly madean efficient and convenient means of access to the house.The fill in structure 16 had a number of significant features. Fromthe floor to just above the bench top it was a natural accumulationconsisting of wind-blown sand with a slight admixture of ashes andsporadic 'pieces of charcoal. There were no potsherds, stone chips,nor bone splinters in this portion of the deposit, showing that itwas not refuse material. Just above the bench level were severalstreaks of water-washed sand and silt similar to those noted in thefill in some of the other pits. In this instance they were at a con-siderably higher level and if they were ptroduced by the samephenomena as in the others would indicate a greater age for thisstructure. This was borne out by other factors suggesting that suchwas the case. Above the water-deposited layers was one of reddish-colored sand, wind-blown material. This in turn was overlain bysand mixed with ashes and charcoal. Numerous potsherds, stonechips, and bone fragments were scattered through this portion ofthe deposit. This was dump-heap material that had drifted downacross the pit from the nearby refuse mound (fig. 25) subsequent tothe abandonment of the house. The remains of a double burial, twochildren, were above the bench at the southwest side of the cham-ber (fig. 25, No. 30). The interment was made after the pit hadfilled to the bench level, as was shown by the fact that the boneswere in the red sand above the layers of water-deposited material.Furthermore, it was prior to the drifting of the refuse material fromthe adjacent mound, as the latter extended across the burial pit inan unbroken line. The skeletons were poorly preserved and had nomortuary offerings with them, so it is not possible to place them inthe general sequence for the site on the basis of associated objects.They may represent the stage of the No. 15 unit, but such cannot beestablished. When the grave was dug a portion of the back wall ofthe bench was removed, which accounts for the greater width of thefeature at that part of its periphery.There were no timbers in No. 16 of sufficient worth to givedendrochronological dates; the butts of the ladder poles were too
RoBEUTsj ARCHEOLOGICAL REMAINS IN ARIZONA J55decayed, and the actual age of the structure is not known. A fewpotsherds found on its floor (similar ones came from one of thenearby granaries) were of the type present in the lowest level in, alltest sections. Hence relatively the structure may be considered asearly in the sequence, even though the definite years represented arenot known.Structure 16 had a diameter of 9 feet 7 inches (2.921 m) on thesipapu, fire-pit line beloAv the bench. Above the bench it measured14 feet 6 inches (4.419 m). Below the bench, at right angles to thefirst measurement, the diameter was 9 feet 9 inches (2.971 m), andon the same line above the bench 14 feet 5 inches (4.394 m). The topof the bench ranged from 1 foot 9 inches (53.34 cm) to 3 feet 5^inches (1.054 m) in width. At the passage side of the room thebench top was 3 feet 10 inches (1.168 m) above the floor and theback wall of the bench was 4 inches (10.16 cm) high. The floor was4 feet (1.219 m) below the old ground level and 6 feet 2 inches(1.879 m) beneath the present surface. At the opposite side the topof the bench was 3 feet 8 inches (1.117 m) above the floor and itsback wall was 5 inches (12.7 cm) high. The floor of the chamberwas 3 feet 4 inches (1.016 m) beneath the original ground leveland 5 feet 4 inches (1.625 m) below the present. On the otherdiameter, at the west side of the room the bench top was 2 feet9 inches (83.82 cm) above the floor and the back wall of the benchwas 1 foot 4 inches (40.64 cm) high. The floor was 4 feet 1 inch(1.244 m) below the original surface and 7 feet 7 inches (2.311 m)beneath the present. At the east side the top of the bench was 2feet 6 inches (76.2 cm) above the floor and the back wall was 4inches (10.16 cm) high. The floor was 3 feet 2 inches (96.52 cm)below the old surface and 4 feet 9 inches (1.447 m) beneath thepresent. The holes at the back of the bench, for the slanting sidepoles, were from 3 to 4 inches (7.62 to 10.16 cm) in diameter.The holes for the support posts were slightly irregular in outlineand in the case of one the edges were broken at the floor line. Thehole at the north side of the chamber had diameters of 6l^ inches(16.51 cm) and 10 inches (25.4 cm) and a depth of 8Y2 inches (21.59cm). The hole was 9 inches (22.86 cm) from the face of the bench.The hole at the east side of the chamber had diameters of 6 inchesand 7 inches (15.24 and 17.78 cm) and a depth of IO14 inches (26.04cm) . It was at the edge of the wall. The hole for the south post wasalso against the wall and had diameters of 7% inches (19.05 cm)and 10 inches (25.4 cm). Its depth was 1 foot I14 inches (33.65 cm).The west hole, the double one, measured 8 inches (20.32 cm) by 11inches (27.94 cm). It was 1 foot 4 inches (40.64 cm) from the wall.The fire pit (fig. 40, e) had diameters of 1 foot 5 inches (43.18 cm)and 1 foot 31/2 inches (39.37 cm). The depth of the pit was 51/2
-.rg BUREAU OF AMERICAN ETHNOLOGY [Bull. 121inches (13.97 cm). The sipapu (fig. 40, c) was 81/2 inches (21.59 cm)from the fire pit. It had diameters of 6 inches (15.24 cm) and 7inches (17.78 cm) and a depth of 10 inches (25.4 cm). The smallstorage hole (fig. 40, h) was 3 inches (7.62 cm) from the sipapu. Thediameter was 3 inches (7.62 cm) and the depth 4 inches (10.16cm). The pot rest (fig. 40, d) was 21/2 inches (6.35 cm) from the firepit. It was 6 inches (15.24 cm) in diameter and 31/2 inches (8.89 cm)deep at the center.The bin at one end of the compartment measured 2 feet 3 inches(68.58 cm) by 1 foot 5 inches (43.18 cm). The floor was 1 foot 1/2inch (31.75 cm) higher than that of the room and the bin had adepth of 1 foot 4 inches (40.64 cm) on the room side. At the back,however, the top of the bench was 2 feet 1 inch (63.5 cm) above thebottom of the bin. The stones remaining in the bin, in order fromtop to bottom in the diagram (fig. 40), measured 1 foot 214 inches(36.83 cm) long, 2 feet 21/2 inches (67.31 cm) high, 4 inches (10.16cm) thick; 41/2 inches (11.43 cm) long, 2 feet 3 inches (68.58 cm)high, 3 inches (7.62 cm) thick; 51/2 inches (13.97 cm) long, 1 foot(30.48 cm) high, 1 inch (2.54 cm) thick; 6 inches (15.24 cm) long,10 inches (25.4 cm) high, % inch (1.58 cm) thick; 1 foot 11/2 inches(34.29 cm) long, 2 feet 2 inches (66.04 cm) high, and 2% inches (6.98cm) thick.The basin in the floor of the compartment (fig. 40, h) had diam-eters of 6V2 inches (16.51 cm) and 9 inches (22.86 cm) and a depthof 4 inches (10.16 cm). The basin was 8 inches (20.32 cm) from theopening into the entrance passage.The entrance to the passage Avas 1 foot 3% inches (39.37 cm) wideat the bottom and 1 foot 6% inches (46.99 cm) wide at the top. Ihetop part of the trench broadened to 2 feet (60.96 cm) a short distanceback of the doorway. The top of the bench was 3 feet 6 inches(1.066 m) above the floor of the passage at the opening. The stonesill (fig. 40, ^) was 1 foot li^ inches (34.29 cm) long, 2 inches (5.08cm) thick, and 2i^ inches (6.35 cm) high. The floor of the passagewas 4 inches (10.16 cm) higher than the floor of the compartment.The passage trench was 4 feet 8 inches (1.422 m) long at the bottomand 5 feet 8 inches (1.727 m) at the top. Wlien in use, the tunnelportion was 3 feet (91.44 cm) long and the manhole was 2 feet 6inches (76.2 cm) by 2 feet 8 inches (81.28 cm). The holes wherethe ladder ends rested (fig. 40, k) were against the end wall. Onehad diameters of 4 and 6 inches (10.16 and 15.24 cm) and the other7 and 9 inches (17.78 and 22.86 cm). The bottom of the trench atthe ladder placement was 3 feet 9 inches (1.143 m) below the oldground level and 5 feet 10 inches (1.778 m) below the present surface.The trench was 1 foot 81/2 inches (52.07 cm) wide at the bottom ofthe ladder and 2 feet 7 inches (78.74 cm) at the top.
Roberts] ARCHEOLOGICAL REMAINS IN ARIZONA 157The row of granaries was 9 feet 6 inches (2.895 m) west of struc-ture No. 16. Midway between the surface structures and the dwell-ing was a surface fire pit lined with stone slabs (fig. 39). Therewere no traces of a shelter or arbor like that in front of the No. 15granaries, A and B. The fire was merely built in the open. Thegranaries differed from those in the group near structure 15 in thatthey were roughly oval or circular in form. They were of the same
'^^ -,^
sreer
FIG0EE 41.?Granaries A and B of the No. 16 group, a, mano or hand milling stone;h, metate stone ; c, fire pit ; d, storage basin ; e, present surface ; f, original groundlevel.general construction, however, in that the walls of the pits were facedwith stone slabs and there were a few examples of the use of hori-zontally laid stones above the vertical ones. Not all of the slabslining the walls extended to the floor levels. Some of them wereset into the wall so that their tops were flush with the ground, whilethe bottoms were several inches above the pit floor. The granarieshad pole, brush, and plaster superstructures. The latter, in two ex-amples, seem to have been pyramidal rather than flat-topped. The
1 gg BUREAU OF AMERICAN ETHNOLOGY IBcll. 121evidence for this was not as clear-cut as could be desired and it ispossible that the other form was used. The other two (fig. 39, A, 7?)were of the truncated type, described in the discussion of the No. 15group, as indicated by the timber molds at the edges of the pits.Two of the granaries (fig. 39, A, B) were connected by a smalldoorway. This is unusual but is probably attributable to the factthat they were used, at least temporarily, as a dwelling. One of thepits {B) had a metate and mano lying on the floor at the center ofthe chamber and the other {A) had a small fire basin in the floor nearthe opening between the two pits (fig. 41). This basin was dug intothe floor and had simple earth walls. It had sufficient use to burnthe soil a deep red. Close to the fire pit and nearer the center of thepit was a depression in the floor that was almost filled with woodashes. Whether the feature was originally intended for such use orhad another purpose before serving as a depository for ashes couldnot be determined. It seems rather curious that so large a plnceshould be provided for an ash receptacle in so small a structureunless the residue was definitely being retained for some specific pur-pose. Parts from two large storage jars were also in this room.Pieces from one of them were found on the floor in No. 16, a factorthat is good evidence for the contemporaneity of the structure.Granary C (fig. 39) had the same general shape and size as B.The floor had been paved with small, thin slabs of stone. The chiefmatter of interest in connection with the structure was in the factthat it overlay a circular pit containing a triple burial (fig. 25, Nos.26, 27, 28). A young adult female and two cliildren were interredthere before the granary was created- There were no accompanyingmortuary offerings to provide a clue to the stage to which theybelonged. Only one of the skulls was in a condition to give anyindications of its general characteristics and since it was undeformed,that fact suggests that the group was among the earlier settlers atthe site. The burials were only 9 inches (22.86 cm) below the floorof the storage bin, but there was clear evidence that the pit for thebin was dug at some time subsequent to the placing of the bodiesthere.An almost identical situation was discovered in connection withgranary D. It also was above a burial, the only difference beingthat two instead of tliree individuals had been placed there. Theskeletons were m a fairly good state of preservation and representedan aged female and an adolescent child (pi. 21, a). They had beenplaced side by side, in tightly flexed positions, in an oval pit. Therewere no mortuary articles in the grave and as in the case of the otherburial there was nothing definite to show the phase of culturaldevelopment that they represented. Both were long-headed indi-
BUREAU OF AMERICAN ETHNOLOGY BULLETIN 121 PLATE 19
n. Pit am! floor of stnictuie 1.'
mm - .^m "'~^,
^. ^J- ''' "^k
b. Comiiartmem wall and ventilator of strueture 15.
BUREAU OF AMERICAN ETHNOLOGY BULLETIN 121 PLATE 20
aUREAU OF AMERICAN ETHNOLOGY BULLETIN 121 PLATE 21
a. Double burial beneath yrauary Ooor in structure 16 group.
h. View across Developmental Pueblo village.
BUREAU OF AMERICAN ETHNOLOGY BULLETIN 121 PLATE 22Ji iil3^
a. Sepond unit in Developmental village.
b. AVall construction in second unit.
ROBERTS] ARCHEOLOGICAL EEMAINS IN ARIZONA I59viduals, however, and in view of the fact that they had been buriedbefore the construction of the granary it seems logical to concludethat they belonged to the oldest stage or the Modified BasketMakers. It is possible that in both this burial and that beneath theother granary baskets or woven materials were mterred with thebodies as offerings and that all traces of them had disappearedthrough the course of subsequent centuries. The use of baskets andwoven bags for such purposes was quite common in the ModifiedBasket Maker horizon, as has been demonstrated by burials found incaves and protected places where the complete dryness of the sur-rounding earth has preserved the materials. The bones of these indi-viduals were 1 foot (30.48 cm) beneath the paved floor of the gran-ary. They were in clean sand and unquestionably were buried beforethe construction of the granary. The latter was not in as good astate of preservation as the others in the group- Most of the flag-stones for the floor were in position but a number of the wall slabshad fallen in and a few were missing. It was possible to trace themain outlines of the pit, however, and establish the fact that it wasroughly oval in outline and approximately the same size as some ofthe others.The main significance in this group of storage structures, as wasthe case for those associated with No. 15, is in their position withrelation to the semisubterranean dwelling and the evidence that twoof them had been used as living quarters. The complex again dem-onstrates the beginning of the unit type, but in slightly cruder formthan in the No. 15 group.The surface fire pit was approximately midway between granaryG and structure No. 16. It was 3 feet 6 inches (1.066 m) from Cand 4 feet (1.219 m) from the back of the bench of No. 16. The pitwas a simple rectangular box lined with stone slabs. A single stonesufficed for each of three sides, but two were used for the fourth.The pit was 1 foot 10 inches (55.88 cm) by 1 foot 9 inches (53.34 cm)and 10 inches (25.4 cm) deep. The old ground level, the top of theslabs in the pit, was 1 foot 7 inches (48.26 cm) below the presentsurface.Granary A was 8 feet (2.438 m) long, 6 feet (1.828 m) wide, 1 foot6 inches (45.72 cm) deep at the north end and 2 feet (60.96 cm) deepat the south. The depression containing ashes was oval in outlinewith diameters of 3 feet 1 inch (93.98 cm) and 2 feet 3 inches (68.58cm) . The depth was 4i/4 inches (11.43 cm) , This basin was 6 inches(15.24 cm) from the east wall of the granary pit. The fire pitwas only Zy^ inches (8.89 cm) from the ash basin. The pit was101/2 inches (26.67 cm) by 1 foot II/2 inches (34.29 cm) on its di-ameters and had a depth of 3 inches (7.62 cm). The pit was 1 foot1 inch (33.02 cm) from the east wall, 1 foot 11 inches (58.42 cm)
j^gQ BUREAU OF AMERICAN ETHNOLOGY [Bull. 121from the west, and 9 inches (22.86 cm) from the doorway. It wasfar enough to one side not to have interfered with use of the opening.The doorway was 1 foot 9 inches (53.34 cm) wide.Granary B was also 8 feet (2.438 m) long. The width varied. Atthe north end it was 4 feet (1.219 m) and at the south, just beforethe walls began to curve to form the end, 4 feet 6 inches (1.371 m).The depth at the north end was 2 feet (60.96 cm) and at the south1 foot 91/2 inches (54.61 cm).Granary C abutted B and was built subsequently to it. G had alength of 8 feet (2.438 m), was 3 feet 91/2 inches (1.155 m) wide atthe end near B and 5 feet 11/2 inches (1.562 m) wide at the other.It was much shallower than either A ov B. At the north end, nextto B, it was 8 inches (20.32 cm) deep and at the south end 1 foot li/^inches (34.29 cm).Granary D was 4 feet 7 inches (1.397 m) from C. The wall slabswere either entirely missing or had fallen from their positions, sothat the measurements are for the pit and for that reason are some-what larger than they would have been with the stones in place.The pit had a length of 9 feet 3 inches (2.819 m) and a width of 6feet 5 inches (1.955 m). The depth was shallow, being 7 inches(17.78 cm) at the north end and 8V2 inches (21.59 cm) at the south.ADDrnoNAi, Pit RemainsSTRUCTURE 17Pit structure No. 17 was located on the floor of the valley belowthe main site. It was uncovered during the excavation of the smallDevelopmental Pueblo village investigated in the 1933 season. Thestructure had been a typical pit dwelling, roughly circular in formwith somewhat more irregular walls than noted in many of thosedescribed in preceding pages (fig, 42). The walls were of earthcovered with a thick layer of adobe plaster. There was no encirclingbench of the excavated type, but the sloping roof poles had been setfar enough back from the top of the wall to supply the equivalent ofsuch a feature. The superstructure appeared to be of the same typeas that described for other pit structures. Four posts placed nearthe walls at approximately the cardinal points of the compass car-ried the main framework for the flat roof and slanting sides.The interior features compared for the most part with those ofother structures, except that there were more recessed subfloor wallpockets in this dwelling and a storage bin formed from stone slabsat one side of the chamber. There was no compartment in front ofthe ventilator opening. The usual combination of sipapu, fire pit,and ladder basin was present. The sipapu (fig 42, d) was fartherfrom the fire pit than in most cases. The diameter was approximate-
ROBEBTS] ARCHEOLOGICAL REMAINS IN ARIZONA 161ly the same as noted for a majority of the examples, although thehole was not deep. The sides and bottom were carefully plastered.
; m
'^
FEET
FIGURE 42.?Structure 17. a, holes for support posts ; 6, subwall pocket ; c, storage pit
;
a, sipapu ; e, fire pit ; 1, basin for ashes and ladder ; g, deflector ; /(-, sill in ventilatoropening; i, subwall pocket; j, storage bin; li, pot rest; m, ventilator shaft; n, presentgi-ound level ; o, original surface.The fire pit was dug into the floor. It was D-shaped with thestraight side toward the ladder pit. The face of the pit at that side
^Q2 BUREAU OF AMERICAN ETHNOLOGY [Bull. 121was reinforced with a stone slab (fig. 42, e) and a small stone wasset into the wall of the arc at the south corner. The remainder wasof adobe plaster applied to the earth wall of the basin.The ladder pit served a double purpose. The lower end of theladder rested in it and it was used as a depository for ashes. Thisbasin also was D -shaped (fig. 42, /). The straight side was towardthe ventilator and was reinforced by the deflector stone (fig. 42, g) .The latter was embedded in the floor and wall of the basin and roseabove the floor of the chamber to a height sufficient to prevent aircoming through the opening from blowing directly on the flames.The butts of the ladder poles rested in shallow depressions in thebottom of the basin at the ends of and against the stone. Wlietherthese "cups" were intentional or merely the result of the abrasiveaction of the timbers could not be determined. Depositing ashes inthe same basin with the ladder was not common. They must havebeen placed there when cold, otherwise they would have ignited thewood.The two recessed subfloor wall pockets were at opposite sides ofthe room (fig. 42, &, i) . When they were first noted, in the processof excavation, it was thought that they were placements for supple-mentary roof braces like those described for some of the houses inthe first group. Careful inspection of the cleared holes, however,indicated that such was not the case, as their sides and bottoms werecarefully plastered and bore no timber imprints. The recessed por-tions in the bench face sloped at approximately the angle that a roofbrace would have followed, but the main part of the holes in thefloor had vertical sides. Slightly more than half of one pocket (&)was beneath the wall, while practically all of the other (^) occupiedthat position. These features certainly provided convenient placesfor the storage of small objects.One end of the roughly oval-shaped basin in the floor near thenorth corner of the chamber (fig 42, c) also extended beneath thewall. It was recessed in much the same way as the two smaller holes.The sides of this depression were vertical, the bottom was level, andboth sides and bottom had been plastered. There was no border-ing rim on the floor as in the case of some of the basins in otherstructures. The purpose of this feature in structure 17 was notlearned, as there was nothing in it when the room was excavated.It could have served as a repository for corn, baskets, jars, or otherarticles used by the people.The bin at the southeast side of the chamber (fig. 42, j) was con-structed of three upright stone slabs augmented by large quantitiesof adobe plaster. One end was joined to the wall of the pit, whilethe other stopped short of a juncture and a space was left open toform a small doorway. The offset in the wall of the pit at the
Roberts] ARCHEOLOGICAL REMAINS IN ARIZONA ]^Q3point where the partition forming the bin joined it produced a largerarea inside that feature than would have been the case had the wallfollowed its normal arc along that section of the periphery. Thismight be considered as indicating that the builders had the bin inmind when the main pit was dug and provided tlie concavity for thatpurpose, although it is possible that the jog in the wall suggestedthe construction of the bin and it was a subsequent addition to thestructure. There was no evidence to show which was the case. Oneof the subfloor storage pockets (i) was in the back of this bin.The only additional interior feature was a pot rest (fig. 42, k) atthe southeast side of the chamber. This shallow depression in thefloor was circular in contour and its sides curved downward to thecenter of the hole. The diameter and curve of the basin were of theproper proportions to accommodate one of tlie large culinary orstorage jars typical of the stage represented by this type of structure.The ventilator opening was too small to have functioned in anycapacity except that of an inlet for air. The aperture was severalinches above the floor level and had a stone sill (fig. 42, h). Thesides and top of the ojDening, which was arched, were covered withplaster. The passage was of the tumieled form with a circular shaftat the outer end. The latter had been damaged by flood waters atsome time after the abandonment of the house and it was difficultto determine its original size.Structure No. 17 was filled with refuse and dump material thatwas deposited there soon after the abandonment of the dwelling. In-dications were that this accumulation came from the unit in the vil-lage that was north of the pit structure and that the latter was builtduring the occupancy of the semisubterranean structure. It is pos-sible that the group using No. 17 had joined in the construction ofthe nearby unit and that they had moved into it when it was com-pleted. The absence of all artifacts and the fact that the refusematerial was lying on the floor, without any intervening layer ofsterile material, suggests that the occupants deliberately moved outand took their belongings with them, while the remainder of the sitecontinued to be inhabited. The position and stratigraphy of the fillwas such that the deposition of material in an open pit was indicatedand this leads to the conclusion that the superstructure was removedat the time of the abandonment. Perhaps the timbers were neededin new construction work and were salvaged from the old structure.There was no decayed wood in the holes for the main supports norwere there any traces of timbers throughout the fill. The holeswhere the uprights stood were broken around the edges, particularlythe one at the north, as though the timbers had been pulled out.At least two granaries had been associated with No. 17. Theywere located several feet west of the structure in a position approxi-
jg^ BUREAU OF AMERICAN ETHNOLOGY [Boll. 121mating that noted for some of the other dwellings. One was largelyobliterated as the result of subsequent activities, but the other wasintact. Both were of the rectangular form with slab-lined pits.The one that remained was found beneath the occupation level ofthe second unit in the village. One end of the pit extended underthe foundations and part of the floor of one room in the surfacedwelling. The main factor of interest in this connection is the addi-tional evidence that the above-ground structures were of later datethan the pit-granary form. This particular granary had a centralbasin in the floor similar to those described for some of the others.Whether it was originally intended for use as a fire pit or had someother purpose could not be determined, but there was evidence thatsmall fires had been lighted at one end of the depression. Traces ofthe timber placements for the superstructure were meager, althougha few scattered holes along the walls indicated that the covering overthe pit was of the type described for the rectangular granariesassociated with structure No. 15. The location and relationship be-tween this granary and the later unit dwelling is shown in the planof the Developmental village (fig. 44) in a subsequent section ofthis report.Structure 17 had a diameter of 10 feet 7 inches (3.226 m) on theventilator, fire pit, sipapu line. At right angles to this measure-ment the diameter was 11 feet 11 inches (3.632 m). At the ven-tilator side of the room the floor was 4 feet 10 inches (1.473 m) be-low the level of the old surface of occupation and 6 feet 10 inches(2.082 m) below the present top of the ground. At the opposite sideof the chamber the depth was 5 feet 2i/^ inches (1.587 m) below theold surface and 7 feet 5^/2 inches (2.272 m) below the present.The holes for the main support posts were somewhat larger thanthose in a number of the structures. That at the north side of thechamber had diameters of 1 foot (30.48 cm) and 11 inches (27.94cm) and a depth of 2 feet (60.96 cm). The back edge of the holetouched the wall. The hole for the east post had a diameter of 10inches (25.4 cm) and a depth of 1 foot 4l^ inches (41.91 cm). Itwas 3 inches (7.62 cm) from the wall. The south hole was 9 inches(22.86 cm) in diameter and 1 foot 6 inches (45.72 cm) deep. It was6 inches (15.24 cm) from the wall. The hole for the upright atthe west side of the room was 11 inches (27.94 cm) in diameter and1 foot (30.48 cm) deep. It was located near the end of a low par-tition forming the storage bin and was 4 inches (10.16 cm) fromthe wall.The fire pit (fig. 42, e) was 1 foot 7 inches (48.26 cm) long, 1 foot4 inches (40.64 cm) wide, and 11 inches (27.94 cm) deep at the center.In front of the stone facing at the ventilator side of the pit thedepth was only 7 inches (17.78 cm). The large stone in the pit
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA lg5Jiniiig was 11 inches (27.94 cm) long and 2 inches (5.08 cm) thick.The smaller stone was 5 inches (12.7 cm) long and 1 inch (2.54 cm)thick. The ladder box (fig. 42, /) was 5 inches (12.7 cm) from thefire pit. It measured 1 foot 3 inches (38.1 cm) by 1 foot 5 inches(43.18 cm). It had a depth of 9 inches (22.86 cm). The deflectorslab (fig. 42, g) formed the face of the ladder basin at the ventilatorside. This stone was 1 foot 4l^ inches (41.27 cm) high, above thefloor level; was 1 foot 2A/2 inches (39.37 cm) wide at the base, 10inches (25.4 cm) wide at the top, and I14 inches (3.17 cm) thick.The sipapu (fig. 42, d) was 4 feet 1 inch (1.244 m) from thefire pit. The hole had a diameter of 5 inches (12.7 cm) and a depthof 4 inches (10.16 cm). It w^as closer to the wall than in most cases,being but 1 foot (30.48 cm) removed from it.The irregular-shaped depression in the floor (fig. 42, c) had amaximum length of 2 feet 4 inches (71.12 cm) and a width of 1 foot9 inches (53.34 cm). It projected 5 inches (12.7 cm) under the wallof the room. The average depth of the basin was 10 inches (25.4 cm).The subfloor wall pocket at the north side of the chamber (fig. 42,h) was 51^ inches (13.97 cm) wide, extended under the wall for 6inches (15.24 cm), and was 5 inches (12.7 cm) deep. The recessedupper portion in the face of the wall began 5 inches (12.7 cm) abovethe floor level. The second recessed pocket (fig. 42, i) was larger. Itextended under the wall for a distance of 9 inches (22.86 cm) and was11 inches (27.94 cm) wide at the opening. The lower portion, thesubfloor pocket, was 11 inches (27.94 cm) in diameter and 4 inches(10.16 cm) deep. The recessed portion in the face of the wall started1 foot (30.48 cm) above the floor level.The bin at the southwest side of the room (fig. 42, j) was 3 feet6 inches (1.066 m) long and 1 foot (30.48 cm) wide. The opening atone end was 10 inches (25.4 cm) wide. The three stones used in thewall varied in size. The first, that at the north end of the partition,was 8% inches (21. 59 cm) wide, 9 inches (22.86 cm) high, and 1 inch(2.54 cm) thick. The top of the stone indicated that it had beenbroken off, so the original height was probably greater than the meas-urement given. The second stone was 1 foot (30.48 cm) wide, 1 foot7 inches (48.26 cm) high, and 3 inches (7.62 cm) thick. The thirdhad a width of 1 foot (30.48 cm), a height of 1 foot 4 inches (40.64cm), and a thickness of 1 inch (2.54 cm).The pot rest (fig. 42, k) was 8 inches (20.32 cm) in diameter and 2inches (5.08 cm) deep at the center. The sides curved downward tothe center from the edges of the basin.The ventilator opening was 1 foot (30.48 cm) high and 1 foot 1inch (33.02 cm) wide. The sill was 4% inches (11.43 cm) above thefloor level. The stone forming the sill was 11^ inches (29.21 cm)68764?39 12
^QQ BUREAU OF AMERICAN ETHNOLOGY LBolu 121long, 5 inches (12.7 cm) wide, and 1 inch (2.54 cm) thick. The llooi-of the tunnel Wcos 1 inch (2.54 cm) below the top of the siU. The tun-nel was 3 feet 51/2 inches (1.054 m) long. Where it entered the shaftit was 1 foot 2 inches (35.56 cm) wide and 1 foot 3 inches (38.1 cm)high. The exact dimensions of the shaft could not be determined,but it closely approximated a diameter of 2 feet (60.96 cm). Thebottom of the shaft was 4 feet 5 inches (1.346 m) below the old groundlevel and 6 feet 3 inches (1.905 m) below the present surface.The single intact granary had a length of 5 feet 3 inches (1.600 m)and a width of 4 feet 3 inches (1.295 m). At one end it was 2 feet(60.96 cm) deep and at the other 1 foot 10 inches (55.88 cm). Thefloor was 3 feet 6 inches (1.066 m) below the present ground level.The basin in the center of the floor measured 2 feet 8 inches (81.28cm) long, 1 foot 9 inches (53.34 cm) wide, and 41/2 inches (11.43 cm)deep. The granary was 5 feet 9 inches (1.752 m) from the structure.STRUCTURE 18The second of the pit structures not located on top of the ridgeat the main site was associated with the unit-type ruins 1 mile(1.609 k) up Whitewater Valley from the village where No. 17 wassituated. No. 18 was a fully developed structure of the same gen-eral type as those described in preceding pages. The pit morenearly approximated the circular form than many of the others, butin the main it differed little from them. There were no indicationsof a definite bench. The sloping side poles of the superstructurewere set back from the edge of the wall, however, so that thereactually was a shelf around the pit between the top edge and theroof timbers. The superstructure was supported on four uprightsset close to the wall of the chamber. This arrangement placed therectangular framework of stringers close to the periphery of the pitand as a result the sloping portion of the ceiling was at a sharperangle than in some examples; the side walls more nearly approxi-mated the vertical.Interior features were few. Near the center of the floor space wasa circular fire pit (fig. 43, /) that was merely a basin dug into theearth. No stones reinforced the walls; they were simply coveredwith plaster. There was no ladder box and no deflector. The plas-tered floor midway between the fire pit and the ventilator openingwas slightly depressed and the surface was abraded in such a wayas to suggest that the lower end of a ladder rested there. Therewas nothing to indicate whether it had been of the pole-and-rungtype or of the notched-log variety. Because there was good evidencefor the former in so many of the structures it is probable that thesame kind was used here.
RoBEBTs] ARCHEOLOGICAL REMAINS IN ARIZONA 167
T^ -h 5 FEET
FiGDEB 43.?Structure 18. a, holes for roof supports; h, storage pit in floor; c, storagebasin ; d and e, sipapus ; f, fire pit ; g, ventilator shaft ; h, present surface ; i, oldground level.
jl^gg BUREAU OF AMERICAN ETHNOLOGY [Boll. 121At the north side of the chamber was an oval basin in the floor(fig. 43, c) comparable to others noted in various structures. Thesides of the basin were vertical and the bottom was approximatelylevel. Around the edges the floor level had an almost imperceptibleraise suggestive of a rim, but it was not actually such a feature. Thisbasin was rather shallow. It probably was similar in function tothose noted in other structures, although its real purpose was notlearned. Between this basin and the wall was an oval trough (fig.43, h) with curved sides and rounded bottom. The feature may haveserved as a rest for a series of round-bottomed vessels, or it may havehad some special function in connection with the larger bin.The only other holes in the floor were those in the position of thesipapu (fig. 43, d, e). Here, as in other examples previously cited,either hole could have been the sipapu or both may have served inthat capacity. There is a possibility that they were not contem-poraneous ; that one was used, then filled in for some reason or otherand the second one provided. If that was the case, any evidenceof such was missed in the digging. The holes were approximatelythe same size, the same depth, their sides were vertical, and bothsides and bottoms were covered with plaster.The ventilator opening was trapezoidal in shape. The sill andlintel were stone slabs but the sides were plaster. The sill wasraised above the general floor level. A plaster offset, encircling theaperture at the back, probably served as a rest for a cover stone.The passage was tunneled, the floor sloping upward toward the shaft.The latter was roughly oval in form.Structure 18 had been abandoned and the roof timbers were eitherremoved or had completely decayed, because there were no indica-tions of them in the pit. The lack of timber prevented any dendro-chronological dating of the structure. Potsherds found on the floorwere of the same types as the material from the houses on the mainsite and the structure unquestionably belongs to the early part of theDevelopmental period. The pit was filled with wind-blown mate-rial from the top of the wall to the floor. It had not been used as adumping place for refuse. The accumulation was wholly a naturalone.The main feature of interest in No. 18 is in the fact that it givesindependent evidence for the greater antiquity of the semisubterra-nean type of dwelling as compared to the surface type of building.The proof is in the stratification at the site. The pit remains didnot underlie any part of the later unit. Their earlier existence, how-ever, was clearly demonstrated in another way by conditions as con-vincing as an actual superposition. After the pit structure wasabandoned, its roof decayed and fallen in, and the pit filled with drift
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA JgQmaterial, a layer of reddish clay was washed down from the talusslope above and swept across the site. The foundations of the nearbyunit house rested on top of the clay stratum, while the filled pit wasbeneath it. As a matter of fact the latter was only accidentally dis-covered when a workman, following: the top of the red layer, brokethrough into the older structure. When the hole for the kiva of theunit dwelling was dug it passed through the clay layer, which alsodemonstrates the later horizon for the unit. There probably weregranaries accompanying No. 18, but they were not definitely located.A few upright slabs were found in scattered positions beneath thefloors in two of the rooms in the nearby unit house. These probablywere remnants of granaries, although activities associated with theerection of the surface building had so modified the area that their ac-tual outlines could not be traced. They were in proper location withrespect to the pit structure and they extended below the red clay, sothat they seem to date from the same period as that house.Structure 18 had a diameter of 10 feet 7 inches (3.225 m) on theventilator, fire pit, sipapu line. The measurement at right anglesto this line was 10 feet 9 inches (3.276 m). At the ventilator sideof the chamber the floor was 3 feet 3 inches (99.06 cm) below the oldground level and 7 feet 5 inches (2.260 m) below the present surface.At the opposite side of the room the floor was 2 feet 10 inches (86.36cm) below the former surface and 7 feet 4 inches (2.235 m) belowthe recent ground level. The floor at the north side of the room was3 feet 1 inch (93.98 cm) below the old surface and 6 feet 8 inches(2.032 m) below the present. The measurements at the south wallwere slightly greater, with the depth below the old surface 3 feet2 inches (96.52 cm) and below the recent ground level 7 feet 10 inches(2.387 m).The holes for the support posts showed that the timbers used hadapproximately the same diameters, as the differences were not great.The hole at the north side of the room had diameters of 9 inches(22.86 cm) and 10% inches (26.67 cm) and was 1 foot 3 inches(38.1 cm) deep. Tlie hole was li^ inches (3.81 cm) from the wall.The east hole had diameters of 8 inches (20.32 cm) and 8I/2 inches(21.59 cm) and a depth of 1 foot 4 inches (40.64 cm). It was 3inches (7.62 cm) from the wall. The south hole was oval in shapeon the floor level but tapered to a circular form near the bottom.At the floor the diameters were 1 foot (30.48 cm) and 6 inches (15.24cm). From midway of the depth to the bottom the diameter was6 inches (15.24 cm). The hole was 1 foot 1 inch (33.02 cm) deep andwas 2 inches (5.08 cm) from the wall at its east end and 7 inches(17.78 cm) away at the west. The hole at the west side of the chamberwas 9 inches (22.86 cm) in diameter and 1 foot (30.48 cm) deep. Ittouched the wall.
l^Q BUREAU OF AMERICAN ETHNOLOGY [Boll. 121The fire pit (fig. 43, /) was 1 foot 6 inches (45.72 cm) in diameterat the floor level and 1 foot 4 inches (40.64 cm) at the bottom ofthe basin where the sloping sides reduced the measurement. The pithad a depth of 5 inches (12.7 cm).The two holes at the location of the sipapu (fig. 43, d, e) were1 foot 3 inches (38.1 cm) and 1 foot 9 inches (53.34 cm) from the firepit and 2 feet 1/2 inch (62.23 cm) and 2 feet 6 inches (76.2 cm) fromthe wall. Hole d was 4 inches (10.16 cm) in diameter and 6 inches(15.24 cm) deep. Hole e was II/2 inches (3.81 cm) from d and had adiameter of 3 inches (7.62 cm) and a depth of 6 inches (15.24 cm).The large oval-shaped basin (fig. 43, c) had diameters of 2 feet 41/0inches (72.39 cm) and 1 foot 7 inches (48.26 cm). The average depthwas 5 inches (12.7 cm). The adjacent elongated depression (fig.43, h) was 21/2 inches (6.35 cm) from o. It had a length of 1 foot 6inches (45.72 cm). For 5 inches (12.7 cm) of its length it had awidth of 41/^ inches (11.43 cm), while for the remainder it was 6inches (15.24 cm). The depth was 4i/^ inches (11.43 cm). The sidesand ends curved downward to form the rounded bottom. The endsof the trough were 5 inches (12.7 cm) and 8 inches (20.32 cm) fi'omthe wall of the room.The ventilator aperture was 1 foot 5I/2 inches (44.45 cm) wideat the bottom, 1 foot 2 inches (35.56 cm) wide at the top, and 1 foot21/2 inches (36.83 cm) high. The sill was 2 inches (5.08 cm) abovethe floor of the chamber. The slab of stone used in the sill measured1 foot 4 inches (40,64 cm) in length, 6 inches (15.24 cm) in width,and 11/2 inches (3.81 cm) in thickness. The lintel was 1 foot 3inches (38.1 cm) long, 6 inches (15.24 cm) wide, and 1 inch (2.54 cm)thick. The plaster offset, for a cover stone around the opening, wasset back 3% inches (9.52 cm) from the edge. The ridge was 2 inches(5.08 cm) wide and I14 inches (3.81 cm) high.The ventilator passage was 2 feet II/2 inches (64.77 cm) long. Theopening into the shaft was 1 foot 6 inches (45.72 cm) wide and 1foot 7 inches (48.26 cm) high. The diameter of the shaft was 3 feet(91.44 cm) on the north-south axis and 2 feet 6 inches (76.2 cm) onthe east-west. The bottom was 2 feet 6 inches (76.2 cm) below theold surface and 6 feet 10 inches (2.082 m) below the present.Unit-Type StructukesThe remains of three definite unit-type structures were investigated.Two of these comprised the early Developmental Pueblo village atthe foot of the talus below the main site and the third, a lateDevelopmental form, was one mile (1.609 k) up the Whitewater onthe south side of the valley. These units are interesting from severalstandpoints. They represent three stages \n the growth of the type

tLLS WITH SLJ'FOUNDATIONS\ \ '"nr\ ^"^""^'^ / SHELTER H
'*'-A,,.i?,,--?"'"''

:)ivi wm'/ffl/'A WALLS Wi I H SLABFOUNDA TfONSWALLS ENTIRELYOF MASONRY68764?39 (Face p. 17J
)
FiGriRH 44.?Flan <
Roberts! ARCHEOLOGIOAL REMAINS IN ARIZONA ^7]^and show the general trend of development. In addition, theydemonstrate clearly certain changes taking place in specific featuresduring the progress from one form to another. Each unit consistedof a number of above-ground enclosures and a semisubterranean orsubterranean chamber. The general plan was much the same for allthree. The surface buildings were located northwest from the under-ground structures and the surface area between the two forms showedextensive use for general occupation purposes. All three units weresituated on sloping ground, so that there was drainage away fromthe structures on at least three sides. This was an essential con-sideration for below-ground chambers and also had decided ad-vantages as far as the surface buildings were concerned. This wasparticularly true of the two units in the one village because thefloors of the rooms in both of the surface houses were depressed belov;the general ground level. Even in the case of the dv>^elling in thethird unit, where the floors were on the surface level, it preventedstanding water from seeping into the mud mortar and softening thefoundations.The village containing the two units (fig. 44; pi. 21, h) was of par-ticular significance because it recorded a shift in function in thesubterranean structures and a definite change in the status of the sur-face rooms. The culmination of this transition and the subsequentperfection of the features involved were exemplified in the third unit.The latter, in many respects, e"\ddenced so marked an improvementin some constructional phases that it seems as though there musthave been additional stages between it and the second in theseries. The elapsed time from the date of the first to that of thethird, as well as the character of some of the lesser objects of thematerial culture, also suggests that there may have been interveningsteps. The differences, however, are in the nature of refinementsrather than major changes, and for that reason do not have anymarked effect on the main outline of the growth of the unit type.There are additional small house rums in the valley located be-tween the two groups herein described, and it is quite likely thatsome of them may contain evidence of the evolutionary processes.Surface indications, the i3otsherds particularly, are that at leasttwo of these sites represent a time intermediate between the secondand third units. UNIT NO. 1The first unit in the series was located at the northeast end of theDevelopmental village (fig. 44). It consisted of a surface structurecontaining three rooms, a portico, and a small court; several outsidefire pits, the remains of brush shelters, and a semisubterranean cham-ber, marked Kiva A on the plan. There is some question about
j^^2 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 121the correctness of this designation, as there is no doubt but whatthe structure did sei-ve as a dwelling during a part of the unit's his-tory. Some of its features, however, were indicative of a ceremonialchamber, and for that reason it was so named. As a matter of factthe structure represents the evolutionary stage when the shift fromsecular to ceremonial functions was taking place and because of thatit is difficult to assign it to either category.The refuse mound for the unit was situated across an old water-course some distance east of the buildings. This channel, indicatedby the dotted lines in the diagram (fig. 44) carried the run-off ofsurface water from the higher slopes followmg heavy rains. Sometime after the first unit was built the channel was blocked and thewater diverted to the valley bottom south and east from the dwell-ings, possibly for use in flood-water irrigation of the fields, and theold bed gradually filled with an accumulation of blow sand andrubbish from the settlement. At a much later period, after the aban-domnent of the entire village, one or two large floods swept acrossthe southeastern half of the site and a new channel was cut througha portion of the refuse mound belonging to the second unit. Thelatter has no bearing on the site so far as any archeological featuresare concerned.The surface structure was approximately rectangular in plan, al-though its walls were not straight and the rooms were somewhatirregular in form. The three definite rooms (1, 3, and 5, fig. 44)were built in a single row. In front of them were the portico andthe small court. The rooms were about the same size and theiroriginal floors were 6 inches (15.24 cm) below the level of the sur-rounding ground. None of the tliree contained fire pits and thereis some question as to whether they should be regarded as livingquarters or storage places. In view of their size, which was con-siderably greater than that of the usual granary, it seems as thoughthey must have served, on occasions at least, as domiciles. Many ofthe rooms in later Pueblo buildings do not have fire pits, hence theirpresence or absence is not necessarily a requirement for a secularchamber. Because of the several fire pits in the portico and court,as well as those associated with shelters, it may be that all of theculinary tasks were performed there and that no provision was madefor such activity in the actual rooms. Under these conditions therewould be no need for fire in the rooms, except in winter weatherwhen the people may have retired to the nearby semisubterraneanstructure. The latter possibility will be considered in greater detailin the discussion of the kiva.There was little fallen material in the fill in the rooms and notmuch in the accumulation of earth around the outside walls. Thereare two possible explanations for this condition. One is that the
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA 173upper walls were mainly of adobe mud that had melted down andwashed away, hence there was little indication of their former ex-istence. The other is that the house was stone-robbed to supply ma-terial needed in the building of the second unit. The latter seemsmore logical. Even when mud walls are entirely eroded away suffi-cient material usually collects on the floors and around the founda-tions to suggest their former presence. Since there was almost noslumped plaster in the rooms the walls probably consisted for themost part of masonry and the upper courses of stone were removedafter the structure was abandoned. Enough of the walls remained,however, to show the type of foundation and the nature of the con-struction directly above. The shallow rectangular pit forming thebase of the rooms was lined with upright stone slabs placed aroundthe periphery in the same manner as those used in the granary andsome of the house pits. Large quantities of adobe plaster were em-ployed in conjunction with the slabs. Above these were severalcourses of horizontally laid stones. The outside portion of themasonry rested on the surface of the ground, while the inside wassupported by the tops of the lining slabs. Where the latter pro-jected above the ground level they were incorporated in the bottomcourses of the masonry. Little attempt was made to dress or shapethe blocks of stone employed in the walls and there was a minimumof chinking. There was no definite nor consistent breaking of jointsand such bonding as occurred between the two rows of stones form-ing the thickness of the wall was purely accidental. The buildersapparently relied on generous applications of mud plaster tostrengthen the construction.Only one of the rooms, No. 5, had the remains of a doorway. Itwas near one corner and opened to the court at the southeast side ofthe building. The other two rooms probably had openings in theirwalls at the portico side. There was not sufficient construction leftto demonstrate this fact, but their existence was indicated by uprightstones placed close to the wall in positions suggesting steps. Thelatter were present in the second unit under conditions clearly show-ing that they undoubtedly were to aid people stepping into or out ofthe chambers and a similar purpose is postulated for those in thisseries of rooms. There were no evidences of windows because of themeager height of the wall remnants, but an item that raises an in-triguing line of thought was found in room 3. A slab of translucentselenite measuring 8 inches (20.32 cm) long, 3 inches (7.62 cm) wide,and % inch (1.90 cm) thick was picked up from the floor midwayof the outer wall. In Pueblo structures of late periods small win-dows were glazed with pieces of this material and the question arisesas to whether or not the builders of this structure used the examplefound there in some comparable way. Thus far no evidence has
174 BUREAU OF AMERICAN ETHNOLOGY tBoLL. 121come to light indicating that the practice was known in pre-Spanishtimes and until a slab of selenite is actually found in place over anopening in an early wall it probably is assuming too much to sug-gest that such was the case in this instance. The possibility is merelymentioned to call attention to the fact that some architectural fea-tures in fully developed Pueblo buildings may have had a compara-tively early origin.There was nothing to indicate what type of roof covered the rooms.In view of the fact that there were no post molds and no signs ofembedded poles around the walls it seems logical to suppose that theroof timbers were supported by the walls and that the covering wasflat. From the standpoint of construction that type of roof wouldnot have caused the builders any great structural difficulties. Largebeams placed across the short way of the rooms would have supportedsmaller poles, brush, and a plaster upper surface that would besatisfactory.The floor of room 1 was paved with stone slabs and room 3 origi-nally had a similar feature, but a second floor was placed in thelatter chamber at some subsequent date and the level raised severalinches. The later floor was smoothed plaster supported by a foun-dation of stones resting on the old floor. These stones were large,unworked blocks fitted together as closely as their imeven surfaceswould permit. Small stones were placed in the interstices as chinking.The plaster above the stones was 2 inches (5.08 cm) thick. The floorin room 5 was hard-packed mud plaster applied to the native earthof the bottom of the shallow pit.There was a slight offset in one wall of room 3 that had no ap-parent purpose. From the southwest corner of the chamber for adistance of approximately 3 feet (91.44 cm) the wall was muchthicker than for the remainder of its length. It is possible that thebuilders decided a thinner wall would suffice, or that the structurewas erected from the ends toward the center, and some of the build-ers were more generous in their use of material than others. On theother hand, there may have been a certain amount of reconstructionand remodeling in the house that produced the feature and this notbe apparent in the remaining ruins of the structure. "Whatever thecause, there does not seem to be any definite function for the jog.The portico (fig. 44, ?) was suggestive of the arbor associated withtwo of the granaries in the No. 5 pit-structure group. The floorspace was enclosed by a low wall that formed a rectangular enclosure,although a wide opening was left at the southeast corner towardthe semisubterranean structure. A roof or shade of some kind waserected over this space. Three upright posts set near the front oroutside wall of the enclosure supported this covering. The otherends of the main beams probably rested on the walls of rooms 1 and
RoSBBTs] ARCHEOLOGICAL REMAINS IN ARIZONA J^753, as there were no indications of support posts along the wall. Onepost was placed some distance from the wall in the south half ofthe arbor as an additional brace. Five fire pits are shown in thedrawing (fig. 44). All were not contemporaneous, however. As inthe case of room 3, the portico had two different floors. Four of thepits belonged to the original level and one to the later. There wasno evidence to show it, but the upper level may correspond to thesecond one in room 3 and have been laid at the same time. Fourof the pits were slab lined, but the fifth was a simple plaster-coveredbasin. The fill between the two floors was earth containing somerefuse material, except for a small area underlying the later fire pitwhere a foundation of stone was provided to furnish support for thestone sides of the pit and also to serve as a j^aved bottom. The mud-plaster surface of the upper floor was applied directly to the topof the fill. There were two storage bins in the southwest cornerof the portico. One was built in the corner and the other was alongthe wall adjacent to it. The corner example had a slab-coveredexterior surface, a thick wall of adobe, and a smootlily plasteredinterior. Its bottom was below the floor level of the room. Threeslabs that were part of the original wall formed the back of the bin.The other consisted of upright slabs covered with plaster. Theenclosure in the corner would have been an excellent place for stor-ing grain, while the other would have satisfied a number of needs.The court (fig. 44, ^) was somewhat comparable to the portico,only not as elaborate. It was entirely open on the side toward thesemisubterranean structure and may not have been covered. Therewere no indications of upright posts to support a roof, althoughtimbers extending over the space could have rested on the walls atthe ends if the latter were liigh enough to function in that capacity.There was no evidence to show that either the north or south wallahad risen to the required height, but they were not as low as thenorth and east walls of the portico. The north wall of the court waserected at the same time as the walls for rooms 3 and 5. This wasdemonstrated by the fact that it was tied into them in the sense thatit was all part of the same construction. The wall at the southend of the court was built after room 5 was completed and abuttedthe corner of the building. The same type of construction was usedin it, although the floor was not depressed. The bottoms of theupright foundation slabs were sunk in the earth for several inchesand the horizontal stones laid around them, so as to suggest an all-masonry wall.There were two fire pits in the court, one near the front at thenorthern end and the other near the back corner at the south end.Both were on the same level and appeared to have been used con-temporaneously. Stone slabs served as lining in both pits, but some
I'JQ BUREAU OF AMERIOAK ETHNOLOGY [^dlu 121were missing from one wall in the pit near the corner. The onlyadditional feature was an upright stone slab near the doorway toroom 5. This stone was not properly placed to form a deflector orscreen for the doorway, and in addition to that was too low to haveserved in that capacity. No step was needed at that place and sincethere was nothing to indicate its purpose no explanation can be givenfor its presence.Just outside the portico was a surface fire pit that gave the appear-ance of being a double pit. Investigation showed that a slightlysmaller and somewhat older pit had been replaced by a larger one.The later example partially cut into the side of the older one and someof the stones were removed from it, possibly for use in the new pit.Charcoal and ashes filled the bottom half of the new pit. On top ofthis deposit were a number of charred fragments of deer bones anda lower jaw from a dog. The latter bore no traces of fire and musthave fallen or been thrown into the pit after the embers were cold.Furthermore, no fires were lighted there subsequent to its deposition.The deer bones no doubt represent scraps from a meal. Whether ornot the dog jaw has a similar connotation is open to question. ThePueblos probably ate dogs at times of great need, but the practicedoes not seem to have been as common with them as with some otherIndian groups. It certainly was not a general custom in this districtbecause numerous articulated dog skeletons were discovered underconditions showing that they were intentionally buried. Stray boneswere so rare that they constitute a practically negligible percentagein the kitchen-midden material. In no case were examples fomid thathad been split for the marrow, a consistent characteristic of the bonesfrom other kinds of animals. The chances are that the jaw in ques-tion had been picked up from the refuse mound where it had eitherwashed out or been micovered by digging, perhaps for a burial, andwas tossed into the pit. The sides of this pit were faced with stoneslabs and the bottom was paved.The few slabs at the south end of the building just outside of thewall of the court, ^, gave little indication of their purpose. Theyprobably were the remnants of a granary or storage enclosure. Notenough was left of the construction, however, to be certain just whatthe}^ rej)resent. There was some alteration of the surface attendanton the erection of the building and the short wall forming the en-closure. This activity probably was responsible for the lack ofevidence relating to the significance of these particular stones.Along the ridge southwest from the surface building and the semi-subterranean structure were the remains of several shelters. They areshown on the plan of the village (fig. 44) in the area between thetwo units designated Court. Every indication was that the group offire pits and holes for arbor posts nearest to Kiva A and the buried
ROBEETS] ARCHEOLOGICAL REMAINS IN ARIZONA 177watercourse were contemporaneous with the first unit. There is somequestion about those just below the legend, Court, and it is possiblethat they date from the period of the pit structure, No, 17, althoughthey may have bridged the transition from that to the first unit.Accumulations of wind-blown sand and some refuse material on thesurface of occupation indicated that they did not correlate with thesecond unit group, as potsherds identical with those from the latteroccurred above the thin deposits. These shelters seem to have beenmuch like those previously described. There were holes for uprightposts to support a shade or covering and fire pits for cooking. Therewere no evidences of side walls. Two of the shelters were separatedby a row of upright slabs extending north from enclosure No. 6, andthe two on the edge of the watercourse were partially delimited in likemanner. Further evidence for use of such places as general outdoorfoci of domestic activity is noted in the metate located adjacent to oneof the fire pits. This grinding stone was not in the position noted as aresult of fortuitous events. It was definitely placed there and rested ona base of stones similar to the metate placement described for struc-ture 15. All of the fire pits associated with these remains were linedwith stone. Some of the pits were round, others rectangular. Anumber of the holes where support posts stood were reinforced bysmall stone slabs.The structural remains marked 6 in the plan of the site (fig. 44)present a problem that cannot be solved satisfactorily from themeager evidence obtained from the excavations. There was nothingto show definitely its relationship to other elements in the village andthere is some question about its function. In many respects it seemsas though there should be a correlation between it and the nearby pitstructure, No. IT, yet there were slight indications that it could beregarded as more closely associated with the first unit. The onlydefinite factor is that it antedated the second unit and seemingly wasabandoned and covered over by accumulated debris and rubbish beforethe second unit was built. In the fill in the structure there were nopotsherds of a type that was characteristic of the second unit, despitethe fact that it was very near that unit. Since potsherds of thatparticular form were exceedingly abundant about the site their ab-sence from enclosure 6 is a good indication of its antecedent status.There were two stages or two phases in the occupation of 6. Origi-nally it had had a fairly large shallow pit. Later the size wasreduced by the erection of an inner wall of stone slabs surmountedby rough masonry. The space between the old and new walls wasfilled in with debris and refuse. Both the large and small forms hadbeen covered with some sort of superstructure supported on threeuprights. This feature was unique for that district. Whether theuprights merely supported a flat shade or arbor such as suggested
]^78 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 121for the brush shelters or a three-sided truncated roof comparableto the four-sided type described for the pit structures could not bedetermined. There were no traces of slanting side poles, althoughthey could have been used without leaving marks around the edgesof the pit.Near the center of the floor space was a fire pit of the generalizedD -shape. The edge of the basin was reinforced at one side by astone slab, but the remainder of the periphery was native earth sur-faced with a coating of plaster. The most interesting interior feature,however, was a subfloor cache pit at the west side of the enclosure.The circular opening into this storage place was against the wall.The cache pit proper was jug-shaped, although the opening was de-cidedly off center, and extended under the floor and beyond thewall of the enclosure. The location and size in relation to the mainpart of No. 6 are shown by the dotted lines in the drawings (figs. 44and 49). At one side of the storage pit, just below the opening, wasan upright slab of stone so placed that it formed a step to aid ingetting down into it. Some such provision was necessary becauseit was not possible to reach the farther limits from the top of theopening. To place corn or other materials around the wall, or toremove objects located there, would require a person's being in thepit and a step of this type was practically essential because the smallsize of the opening and the depth of the pit made access withoutsome form of footrest a difficult process. The pit was carved outof the earth. The upper walls sloped down and away from theopening and met the vertical lower walls 1 foot 6 inches (45.72cm) above the floor. Except for the step, no stone was used in itsconstruction. The walls and bottom were covered with plaster.A cover of some kind was probably provided for the opening. Nonewas found, but it does not seem likely that the pit was used withoutone. Stone slabs were frequently employed in that capacity, par-ticularly when openings were in the floor of an enclosure. Absenceof such a feature is not conclusive evidence that nothing of thekind was used. A cover stone might have been removed to servesome purpose in another structure when No. 6 was abandoned.The cache pit in No. 6 was the only example found inside a struc-ture in this district. A somewhat similar pit was present at the siteof the third unit, but it was removed some distance from the build-ings. Comparable granary or storage pits occur at various places inthe Southwest. In one of the units of a Developmental village insouthern Colorado there was a large storage pit in the floor of alean-to or shed that offers an analogous combination of structure andgranary.^^ In most cases, though, they are outside and do not con-
^B Roberts, 1930, p, 9, fig. 5.
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA ^79stitute a part of any structure. An interesting factor in connectionwith these storage pits is that they appear to occur in the earlier hori-zons in different areas in the Southwest rather than in the well-estab-lished phases. Just why this should be, or what its significance is,has not yet been determined. It is possible that need for them wasremoved in the later type of structures with their various lower backrooms available for storage purposes. Not enough work has beendone in sites representing Developmental stages to warrant attemptsat definite conclusions on the storage-pit feature or its importancein the cultural complex.The cache pit in No. 6 was filled with some refuse and considerablewind-blown sand. There were no potsherds in the material. This israther curious when their occurrence on the floor and in the fill ofIhe associated enclosure is considered. Their absence may be eitherwholly accidental or due to the fact that the lower pit had becomecompletely filled before any waste matter found its way into theupper structure. No great importance can be attached to the cir-cumstance because there is no question but that the cache pit andNo. 6 were contemporaneous.The problem of what No. 6 represents is a puzzling one. Whetherit should be regarded as a variant form of pit dwelling, a brush shel-ter, or an enlarged granary is a difficult question to answer. Particu-larly since there is so little satisfactory evidence. The general fea-tures are more suggestive of the brush-shelter type of construction,yet the depressed floor level and slab facing on the walls of the pitare more like those of the granaries or dwellings. The pit is largerthan the general run of granaries and not quite as large as the major-ity of those for domiciles. The presence of a true fire pit bespeaksa dwelling place but the shallow pit, the lack of any form of venti-lator, and the peculiar triangular frame for a superstructure set itapart from the usual form for such structures. Wlien all of the vari-ous phases of the problem and the numerous ramifications of eachare considered it seems that the most logical conclusion is that theenclosure was an adjunct to the nearby pit structure. It was a moreelaborate development of the shelter than customary, possibly wasused over longer periods of the year, and for that reason exhib-ited more of the characteristics of a true house. The shallow depthfor the pit was not altogether unusual ; it will be recalled that both13a and 14 were characterized by a similar condition, but the absenceof a ventilator in such a house is rare and for that reason the shelteridea is given preference. Shallow pits of approximately the samesize and depth were present at the Long H Ranch, but the houses didhave ventilators.^" There is no good proof for the foregoing conclu-sion, however, and it is presented solely in the form of a questionable
? Roberts, 1931, pp. 43-48, houses A and B.
IgO BUREAU OF AMERICAN ETHNOLOGY [Boll. 121explanation. The granary function of No. 6 was served by the sub-floor cache pit.Kiva A was similar in some respects to the semisubterranean struc-tures discussed in preceding pages, but in others it was quite distinct.The pit did not approximate a circular form as closely as most ofthose in this area, as it tended to the rectangular with roundedcorners. The upper portion of the pit was encircled by a narrowbench. Both pit and bench were dug into the earth. The entireperiphery of the chamber at the floor level was encircled by uprightslabs. Horizontally laid masonry rested on top of the vertical stones,which were not uniform in height, and carried the wall to a higherlevel. The stones used in the courses above the slabs were of differentsizes and shapes. Some were large, others small, some tabular inform, and others cubical. They exhibited very little dressing or shap-ing and apparently were laid in the wall just as they came from thequarry. On the whole tliey were larger and more variable than usualfor kiva construction. A few small pieces of stone were employed aschinking, but the latter was not extensive. The top of the masonryat the bench level was covered with a good coating of plaster. Thetop of the earthen part of this feature was not so treated, however,and was quite irregular in comparison to the narrow ring of thewall. The main part of the wall, from the bench to the floor level,was originally well plastered. At the time of excavation, though,large patches had fallen off.The superstructure that covered the pit was of the truncatedpyramidal type supported by four upright posts placed near thecorners of the room, but set out from the wall. There was a series ofsmall holes around the back of the bench where the butt ends of theslanting poles rested. The post pockets at the corners were largerthan the others, indicating, presumably, that heavier timbers wereused in those positions. One of these corner placements, that at thesoutheast, was reinforced by a stone slab set in the front side of thehole. The structure had one unusual feature in the presence of a fewirregular pockets in the back wall on the bench level between thesloping poles. These served for the storage of small objects, as wasshown by the presence in them of some bone awls and punches andtwo small pottery ladles.The fire pit near the center of the room (fig. 45, g) was D -shaped.A stone slab was set in the side toward the ash box and ventilator.The remainder of the basin was surfaced with mud plaster. The pitwas encircled by a plaster ridge or rim that was a continuation and apart of a ridge that extended from wall to wall and separated thefloor space into two approximately equal parts (fig. 45, /). Similarplaster ridges are frequently found extending from the fire pit to thebottoms of the main support posts at the ventilator side of the
Roberts] ARCHEOLOGICAL REMAINS IN ARIZONA 181chamber/^ but the occurrence of such a feature passing straight acrossthe approximate center of the room is rare. The nearest comparable
??0 /
FlGURH 45.?Kiva A. a, holes for roof supports ; 6, storage pit ; c, storage hole ; d and e,sipapus ; t, plaster ridge on floor ; g, fire pit, h, old ladder pit ; i, ladder pole depressions ;;, storage pocket ; h, pot rest or storage pocket ; I, ventilator aperture ; m, poles rein-forcing ventilator opening; n, ventilator shaft; o, slanting timber placements; p, Q, r.Storage pockets on bench top ; s, present surface ; t, old ground level.
?i Judd, 1926, p. 113; Roberts, 1929, pp. 25, 29-50, 54, 57-58; 1931, pp. 26-29.68764?39 13
1^2 BUREAU OF AMERICAN ETHNOLOGY [BnLU 121example was in a small house belonging to the Modified Basket Makerstage located in the Chaco Canyon.*^^ xhe function or significance ofsuch ridges is not known. In the Modified Basket Maker houses theyappear to be derived from the placement for the slabs that formed thecompartment on the ventilator side of the chamber and seeminglyrejiresented that feature in delimiting a certain portion of the floorarea. In some cases logs were buried in the floor to form a foundationfor these radiating ridges. Undoubtedly there was a good reasonfor their inclusion in houses, but what it was has not been learned.The ridge in Kiva A was the onlj^ example found in the Allantowndistrict. No wood was used in its construction ; it consisted wholly of])laster.Adjoining the fire pit in the position usually occupied by theladder box was a pentagonal pit that could have served such apurpose, although it apparently did not (fig. 45, h). The pit wascovered with a stone slab. When the room was excavated there wasa layer of fine wood ashes on top of the stone, yet only a smallamount in the pit. The latter seemingly had sifted through thecracks between the stone and the edges of the pit. The bottom andsides of the latter were not stained in the manner common to thosewhere ashes have remained for any length of time, hertce it didnot serve as a depository for them. The ladder stood on the floorof the chamber between this pit and the ventilator opening, as shownby two small basins worn in the plaster where the butts of thepoles rested (fig. 45, ^). The pit originally may have functionedas a ladder rest, as a slight abrading of the bottom at the ventilatorend indicated. Then for some reason the position of the ladderwas changed and the pit covered with the stone. Why it was notused as an ash depository as in some other houses is not known. Itmight have served as a storage place on occasions, particularly forsmall objects, but that is purely conjectural, as none were foundin it. There was no deflector and no signs that one had been presentduring the occupancy of the structure.There were several holes at various places in the floor of thechamber. At the location of the sipapu there were two, d and e,either one of which could have represented that symbolic feature.Both were shallower than usual and somewhat smaller in diameter.Their sides and bottoms were carefully plastered. Close to the wallnorth of them was another small hole, c, that evidently was for thestorage of minor objects. Its sides and bottom were also plastered.Near the base of the support post at the northeast corner of the roomwas an irregularly shaped basin, b. It was deeper than many ofthose occupying similar positions in other structures. The walls
"2 Roberts, 1929, pp. 57-58.
ROBBBTS] ARCHEOLOGICAL REMAINS IN ARIZONA lg3were vertical and the bottom comparatively level. The walls andbottom did not form a sharp angle, rather a curved one. Both wallsand bottom were covered with plaster. The basin near the post atthe southeast corner, j, was at the base of the wall. One side of thedepression was bordered by part of the slab forming the wall atthat point on the periphery. The basin was oval in contour andquite shallow. The sides curved downward from the edge on thefloor level to the maximum depth near the center. Perhaps thiswas a pot rest. The shallow basin, h^ near the support post at thesouthwest corner also seems to have been for a pot rest. The diam-eter and depth suggest that it was the placement for a large culinaryor storage jar. When the room was excavated a thin-slab metate ofthe open-end form was leaning against the wall near the basin andthe, support post.The ventilator passage and shaft were dug as a trench and thencovered over. No tunneling \vas used in its construction. The endof the trench where it penetrated the wall of the room widened to amarked degree, but it w^as reduced in size by the erection of a crudemasonry framework for the opening. This stonew^ork cut off moreof the trench at the west side than at the east, as may be seen fromthe dotted Imes in the drawing (fig. 45). The aperture proper wasframed by plaster reinforced by poles. Two small poles were setin the wall at either side of the opening and a short timber was laidacross their tops on a level with the top of the bench. The cross-piece was above the lintel of the opening, which was a stone slab.The construction was covered with plaster so that at the timethe chamber was occupied the timber was not exposed to view.The sill of the opening was above the floor level and was pavedwith a stone slab. The walls of the passage were lined with up-right slabs; the ceiling consisted of slabs and the floor was pavedwith them. The walls and floor were covered with plaster forabout half the length of the passage. The bottom of the ventilatorshaft was lined with upright slabs surmounted by horizontally laidstones. The latter were drawm in course by course as the masonryextended upward until the opening on the ground level was quitesmall. The work was not as good as it might have been and thesmall hole at the top was somewhat off center. The floor of thepassage sloped slightly downward from the aperture in the cham-ber to the bottom of the shaft. This was not the case in many ofthe structures, although such a slope would have merit, in thatrain or surface water falling into the shaft would collect at the outerend and not run into the room. There was little danger of thelatter, however, as the ground sloped sharply away from the top ofthe shaft toward the old watercourse.
2g4 BUREAU OF AMERICAN ETHNOLOGY [Bull. 121There were several interesting features in the fill in Ejva A.When the pit was excavated a pillar 3 feet (91.44 cm) square wasleft at the center, like those described for some of the other struc-tures, and a careful check made of various phenomena recordedthere. Through the central portion of the pillar the layers werethinner than those at the top or bottom and contained more refusematerial, particularly stone chips, charcoal, bone fragments, andpotsherds. Indications were that an mterval of some length passedbetween the abandonment of the structure and the use of the pitas a dumping place. After the refuse had filled it to a level ap-proximately midway between the top of the bench and the oldground level it was no longer used for that purpose and subsequentlayers resulted from natural agencies. Indications were that thepresence of the old pit had been marked by a depression for a con-siderable length of time and that water had gathered and stoodthere at intervals. Subsequent to the abandonment of the villageall traces were obliterated by wind-blown sand and drift from thehigher levels.On two occasions during the period when the pit was functioningas a depository for refuse some structure in the village was damagedby fire. This was shown by the layers of sand mixed with charcoal,bits of burned stone, and fragments of burned plaster, debris re-sulting from reconstruction activities. The levels at which tliismaterial occurred were sufficiently separated to suggest that a num-ber of years elapsed between the conflagrations. One of the lastlayers laid down prior to the discontinuance of dumping wastematter in the pit consisted almost entirely of stone spalls, brokenbuilding stones, and pieces of mud plaster. This debris probablycame from the nearby surface building and may indicate the periodwhen it was dismantled for the material used in its construction.If such is the case it suggests that the second unit was occupied forsome time before the first was robbed of its stone and timbers. Thefact that it also occurred just above the last layer of debris from aburned structure carries the implication that the abandoned firstunit supplied the material needed to replace the portion of the sec-ond unit damaged by the flames. That such actually took place isnot known, of course, but in view of the evidence the conclusion thatit did does not seem to be unwarranted.One strange find in the fill was not included in the stratigraphicpillar. It consisted of a human skeleton lying along the bench atthe west side of the chamber. A portion of the skull rested on thebench. The remaining bones were just below that level and some-what scattered, but not sufficiently disarticulated to indicate theburial of an already decomposed body. There were no leg or footbones, however. The pelvic bones were present and in good state of
ROBEBTS] ARCHEOLOGICAL REMAINS IN ARIZONA 185preservation. They were normal in every respect. The acetabula,the cup-shaped articular cavities in the hip bones where the ends ofthe thigh bones rest, were undamaged and showed that the individ-ual originally had legs. All of the remaining fill was carefullysifted but not one additional bone was recovered. There are severalintriguing questions concerning the remains that cannot be an-swered in a satisfactory manner. Was the skeleton buried wherefound, or did it tumble into place after the pit was partially filled?If the latter, where did it fall from and what happened to the legs ?The bones were not in a definite grave dug into the fill. They wereon a former surface and overspread with refuse. Indications werethat the legless body was deposited there and covered with materialscraped from the surrounding area. Why it had no legs is notdefinitely known, but what became of them can be answered in partby finds in the second unit and therein lies a possible explanation fortheir absence.A left femur or thigh bone of proper size for this individual wasuncovered in the corner at the outside of the building where thewalls of Nos. 7 and 9, figure 44, meet. One end of the bone wasgnawed. The teeth marks were too large to be those of a rodent andpossibly were made by a dog. At the back of the recess betweenNos. 9 and 11, in the northwest corner formed by the walls 9 and 10,a complete left foot, including the ankle bones, was uncovered. Thefoot had been sealed in the corner, after being covered with earthfrom the refuse mound, by a thick coating of plaster. The bones,except for two in the ankle, were articulated and showed that themember had been interred while held together by flesh and ligaments.These two occurrences open an interesting field for speculation.The teeth marks on the femur suggest that the body in the pit ofKiva A had not been covered as thoroughly as the situation demandedand that some of the village dogs may have dug into the grave andpulled out the legs and torn them apart. Then some of the grislyfragments were gathered up by the inhabitants of the village andreburied in the spots close to the building instead of in the refusemound or old house pit. Perhaps they thought that by so disposingof them they would be protected from further violation. This is,of course, pure conjecture and may not be an even approximate re-construction of what happened. The leg bone and the foot may havebelonged to entirely different individuals. The burial of parts ofthe body occurred sporadically throughout the Southwest and it ispossible that these represent some such happening rather than thatthey are missing members from the legless skeleton. Since they arethe proper size for that individual, however, and the leg bone showsmarkings from animal teeth, the conclusion that they belong togetheris not unreasonable. Unfortunately the condyle on the femur that
IgQ BUREAU OF AMERICAN ETHNOLOGY [Boll. 121would rest in the socket in the pelvic bone was the end that had beengnawed off. As a consequence it is not possible to obtain the evi-dence that would settle the question, that of fitting the one into theother.An almost identical situation was encountered at the opposite sideof the chamber just above the floor level where a partial dog skeletoncame to light. Both hind legs and one front leg were missing fromthese remains. The skull and other bones were present and articu-lated. The same questions raised by the human remains apply tothose of the animal, but even less convincing explanations can beoffered in answering them. No stray bones that could be attributedto the dog were found elsewhere. It is possible that after thecreature's demise and the disposal of the body in the old house pitrodents made away with the legs. There was nothing to show that,however. In view of the number of dog burials about the site it doesnot seem likely that the remains were just tossed into the pit. Thenature of the earth around the bones was such that a gradual accu-mulation was indicated. Hence intentional burial probably was notpracticed in this instance. One plausible explanation is that the dogfell into the pit, possibly while the superstructure or portions of itwere still in place, died there and before drifting sand had coveredthe remains they were mutilated by other animals. The problemis not an important one and has no definite bearing on the culturalstatus of the community, but it is an example of one of the interestingincidentals occasionally noted in an archeological site. The completeskeleton of a dog was found on the floor of a pit structure in theChaco Canyon under conditions suggesting that the animal wastrapped and perished in an abandoned dwelling.^^ The somewhatanalogous occurrence at this site is an additional indication that theprowling proclivities of the Indian dogs were likely to lead theminto a combination of circumstances that were fatal.The status of Kiva A, whether dwelling or ceremonial chamber, isdifficult to establish. If it was associated with a cluster of simplegranary remains, like those near structures 15 and 16, it unquestion-ably would be called a domicile. On the other hand, if found witha well-developed surface house there would be no hesitancy in con-sidering it a ceremonial structure, although one in which certain fea-tures were not as highly specialized as in most examples. The sig-nificance in this seeming paradox is the evidence that it gives fora transition in function in the semisubterranean structure. Alsothat at the present stage of knowledge of the Southwest it is notpossible to tell from the internal nature of a structure, dating fromthis period in the growth of the cultural pattern, whether it was houseor kiva. Conclusions must be aided and governed by the associatedM Roberts. 1929. p. 66.
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA 1?7remains. The fact that the nearby surface building had features thatqualified it for use as a dwelling tends to argue for a ceremonialfunction for the underground chamber. The latter, however, was sosimilar to pit structures on the site that definitely were dwellingsthat it is difficult to disregard that aspect of the problem. The trun-cated pyramidal superstructure was characteristic of the form foundon dwellings, but the horizontally laid wall masonry is a featuremore commonly present in kivas. The floor depth below the top ofthe bench and original ground level was somewhat greater thancustomary for houses, although the chamber was not as deep as theaverage kiva. Considered from all points of view the structureis intermediate in type between a dwelling and a ceremonial cham-ber. There is no definite proof to show that such was the case, butit seems probable that the pit structure was built first and usedas a habitation, then the surface rooms were constructed and therewas a gradual shift from one to the other. The few scattered evi-dences that regular slab-pit granaries occupied the site of the sur-face building prior to its erection suggest that possibility. Evenafter the surface rooms were occupied the people may have revertedto the old underground house during the winter season. As waspointed out in the discussion of the surface building, there was noprovision for fires in the rooms and it is doubtful that those blazingin the pits in the floor of the portico or the court would have givenoff enough heat to warm the house in cold weather.Charcoal suitable for dendrochronological studies was scarce in thefill in the kiva. Sections from two of the slanting roof poles weresalvaged from holes at the back of the bench. One of these couldnot be correlated with the ring chart and a number of outside ringswere missing from the other. The closest approximation to a dateis 845. Due to the range in the timbers for other houses it is evi-dent that too much stress cannot be placed on the 845. There is noway of knowing whether it is at the earlier or later end of a seriescovering a long span of years, such as 31 in No. 15 or 74 in No. 12,or falls midway in some comparable sequence. There is also thepossibility that the particular pole was salvaged from the nearbyabandoned pit structure. No. 17, and hence older than the kiva. Itwill be recalled that the former structure gave evidence of havingbeen dismantled. On the basis of evidence from some types ofpotsherds the kiva appears to be roughly contemporaneous with thefinal stage in No. 15, yet other types indicate that the correlationshould be with the first stage described for that structure. If syn-chronous with the final stage the date for the kiva would be some40 years later.Several interesting questions are raised by the 845 date and it istantalizing not to know what interpretation to place on the struc-
188 BUREAU OF AMERICAN ETHNOLOGY [Bull. 121ture. If 845 is correct for the year when the house was built it sug-gests that the people in this unit were more advanced in architec-tural development than those in No, 15. The implications are thatthey not only had adopted a surface structure with contiguous roomsat a time when the others were still building the old style granariesand only occasionally using them as habitations, but that they pre-ceded them in this advance by two generations. Considering theproximity of the two structures this does not seem reasonable. Ofcourse, as pointed out in the discussion of the kiva, it is fairly evi-dent that the surface building was erected subsequent to the semi-subterranean chamber and allowing for a lapse of a number of yearsbetween their construction the discrepancy would not be as great,although still of marked character. On the other hand, when com-pared with the 867 of No. 3, the 845 does not seem as far out of line.Especially when it is recalled that some of the timbers there showedcutting dates of 842 and 852 and that No. 3 apparently was a kivafor a surface structure quite comparable to that in this unit. Hereagain is the indication that No. 15 was a laggard unit. This hasfurther significance in the fact that the dominant type of potsherdsin the 15 assemblage was a form centering to the north and westfrom the AUantown district, the western type of Developmentalpottery generally called Kana-a black on white. From this it seemsthat 15 may represent an addition to the community of a group thatwas not as progressive in house building and was somewhat con-servative in the adoption of nevr forms. Other evidence about thesite showed that the form of pottery in question was later in itsappearance in this district than other Developmental forms.The 845 date, if considered as being indicative of No. 17 ratherthan the kiva, correlates quite well with other aspects of the remainsin the district in that the pit structure compares favorabh' with othersclosely approxim.ating the same age. Considered from the broadpoint of view, the import of the whole matter probably is that the45-year interval represented by these structures covers the periodwhen the transition in house types and other cultural traits was atits maximum. The combination of circumstances and implicationsraised by this particular piece of charcoal emphasizes the fact thattoo much importance must not be placed on a house date based ona single specimen of wood. As an indication of the approximateperiod represented it is helpful, but it is not to be considered con-clusive.The surface building in this unit had a maximum length of 32feet (9.754 m) and a width of 18 feet 6 inches (5.638 m). Insidemeasurements for room 1 were: length 8 feet 6 inches (2.591 m),width 6 feet 6 inches (1.981 m). Room 2 (the portico), length 18feet 10 inches (5.740 m), width 8 feet 9 inches (2.667 m). The open-
ROBESTS] ARCHEOLOGICAL REMAINS IN ARIZONA J?9ing at the southeast corner was 4 feet (1.219 m) wide. Room 3 was11 feet (3.352 m) long and 7 feet (2.133 m) wide. The offset at thesouthwest corner was 3 feet (91.44 cm) long and 9 inches (22.86 cm)wide. Room 4 (the court) was 10 feet 3 inches (3.124 m) long. Thewall at the north end measured 5 feet 9 inches (1.752 m) m lengthand that at the south end 4 feet 9 inches (1.447 m). The doorwaybetween 4 and 5 was 1 foot 6 inches (45.72 cm) wide. Room 5 hada length of 9 feet 6 inches (2.895 m) and a width of 6 feet 3 inches(1.905 m).The fire pits in room 2 in order from north to south measured asfollows: The first, diameters 2 feet (60.96 cm) by 1 foot 9 inches(53.34 cm), and depth 6 inches (15.24 cm); the bottom was 5i/^inches (13.97 cm) above the old floor. The second diameters 1 foot4 inches (40.64 cm) by 1 foot 3 inches (38.1 cm) and depth 5i/^inches (13.97 cm). The third, diameters 1 foot 4 inches (40.64 cm)by 1 foot 6 inches (45.72 cm) and depth 6 inches (15.24 cm). Thefourth, diameters 1 foot 8 inches (50.8 cm) by 1 foot 6 inches (45.72cm) and depth 7 inches (17.78 cm). The fifth, the plaster pit, diam-eters 2 feet 3 inches (68.58 cm) by 1 foot 7 inches (48.26 cm) anddepth 514 inches (13.97 cm).The holes for the support posts for the arbor over the portico didnot show a marked range in size ; that at the northeast corner had adiameter of 6 inches (15.24 cm) and a depth of 8l^ inches (21.59cm) ; the post stood 1 foot (30.48 cm) from both walls. The holemidway along the front wall had a diameter of 6 inches (15.24 cm)and a depth of 10 inches (25.4 cm). It was 1 foot 6 inches (45.72cm) from the wall. The two holes near the fire pit at the southeastopening had diameters of 7 inches (17.78 cm) and 6 inches (15.24cm) and depths of 9I/2 inches (24.13 cm) and 4 inches (10.16 cm).The larger hole was 11 inches (27.94 cm) from the wall and 9 inches(22.86 cm) from the fire pit; the smaller hole was 4 inches (10.16cm) from the larger and 3 inches (7.62 cm) from the fire basin.The hole for the post near the back of the south end of the roomhad a diameter of 101/2 inches: (26.67 cm) and a depth of Q% inches(17.14 cm). It was 1 foot 10 inches (55.88 cm) from the back wall.The rectangular-slab bin along the wall near the southwest cor-ner measured 2 feet 6 inches (76.2 cm) by 1 foot (30.48 cm). Theslabs were 1 foot 4 inches (40.64 cm) and 1 foot 9 inches (53.34 cm)high. The granary in the corner of the portico had inside meas-urements of 1 foot 4 inches (40.64 cm) by 1 foot 9 inches (53.34 cm),with a depth of 2 feet 2 inches (66.04 cm) . The plaster rim or wallhad an average thickness of 6I/2 inches (16.51 cm).The fire pits in the court (No, 4, fig. 44) were approximately thesize of those in the portico. That at the north near the front of thespace had diameters of 2 feet 2 inches (66.04 cm) and 1 foot 101/2
IQQ BUREAU OF AMERICAN ETHNOLOGY [Boll. 121inches (57.15 cm) at the top, and 1 foot 3 inches (38.1 cm) and 1foot 8 inches (50.8 cm) at the bottom. The depth was 5 inches(12.7 cm). The pit was 9 inches (22.86 cm) from the north wall.The pit at the southwest corner had diameters of 1 foot 9 inches(53.34 cm) and 1 foot 10 inches (55.88 cm) at the top and 1 foot(30.48 cm) and 1 foot 4 inches (40.64 cm) at the bottom. Thedepth was 7 inches (17.78 cm). The pit was only 3 inches (7.62 cm)from the west wall and 1 foot 2 inches (35.56 cm) from the south.The stone slab set in the floor of the court was 1 foot 3 inches (38.1cm) long, 6I/2 inches (16.51 cm) high, and 31/2 inches (8.89 cm)thick. It stood 9 inches (22.86 cm) from the wall. The old pitbeneath the floor close to the stone had diameters of 1 foot 3 inches(38.1 cm) and 1 foot 8 inches (50.8 cm), with a depth of 1 foot 2inches (35.56 cm).The fire pit in front of the building, just outside of the portico,measured 1 foot 8 inches (50.8 cm) by 2 feet (60.96 cm) and had adepth of 10 inches (25.4 cm). The older pit, which it replaced,measured 1 foot 6 inches (45.72 cm) by 1 foot 9 inches (53.34 cm)and was 1 foot (30.48 cm) deep. The old pit was 9 inches (22.86 cm)from the portico wall and the later one was 2 feet (60.96 cm) away.The remains of the brush shelter with the two fire pits and themetate, located on the edge of the old watercourse, were 12 feet (3.657m) from the corner of the surface building. The rectangular pitmeasured 1 foot 9 inches (53.34 cm) by 1 foot 7 inches (48.26 cm)and was 3i/4 inches (8.89 cm) deep. The metate was 1 foot (30.48cm) from the fire pit. The metate was 1 foot lli/^ inches (59.69 cm)long, 1 foot 4 inches (40.64 cm) wide at the open end and 1 foot(30.48 cm) wide at the closed end. The trough was 1 foot 2 inches(35.56 cm) long, 10 inches (25.4 cm) wide, and 4 inches (10.16 cm)deep. The stone was 4^/^ inches (11.43 cm) thick. The second firepit measured 1 foot 7 inches (48.26 cm) by 1 foot 9 inches (53.34 cm)and had an average depth of 7% inches (19.05 cm). The standingstones that partially separated the floor area around the two pitshad a combined length of 1 foot lOi^ inches (57.15 cm) and stood111/2 inches (29.21 cm) above the floor. The holes for the supportposts for the arbor erected over the pits ranged between 6 and 7inches (15.24 and 17.78 cm) in diameter. Their depths were 1 foot(30.48 cm).The fire pit south of the above group measured 1 foot 2 inches(35.56 cm) by 1 foot 3 inches (38.1 cm) and was 1 foot (30.48 cm)deep. The wall of upright slabs to the west of this pit, separatingthe general area of occupation marked "Court" on the diagram(fig. 44), had a length of 12 feet (3.657 cm). The fire pit on thewest side of this wall measured 1 foot 2 inches (35.56 cm) by 1 foot3 inches (38.1 cm) and was 8 inches (20.32 cm) deep. The sur-
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA JQ]^rounding holes for posts ranged from 71/2 inches (19.05 cm) to 9inches (22.86 cm) in diameter and from 10 inches (25.4 cm) to1 foot 6 inches (45.72 cm) in depth.The original enclosure at No. 6 had diameters of 9 feet (2,743 m)and 10 feet 6 inches (3.200 m). The smaller room was 6 feet(1.828 m) by 9 feet 3 inches (2.819 m). The floor of the originalroom was 2 feet 1 inch (63.5 cm) below the old ground level at thesouth side. The stone wall for the reduced enclosure rose 2 feet1014 inches (86.99 cm) above the floor at the south side, 2 feet 9inches (83.82 cm) at the east, 3 feet 5% inches (1.047 m) at thenorth, and 2 feet 5 inches (73.66 cm) at the west. The single holefor a support post in the corner of the old room was 8 inches(20.32 cm) in diameter and 1 foot 7% inches (49,53 cm) deep. Thehole at the south corner in the smaller room was 5i^ inches (13.97cm) and 6% inches (16.51 cm) in diameter and 6 inches (15.24 cm)deep. The hole near the east corner was 4 inches (10.16 cm) and5 inches (12.7 cm) in diameter and 4% inches (12.06 cm) deep.The hole at the north corner was 5 inches (12.7 cm) in diameterand 61/2 inches (16.51 cm) deep. The fire pit measured 1 foot 3inches (38.1 cm) by 1 foot 7 inches (48.26 cm) and was 7 inches(17.78 cm) deep. The stone that faced one side was 1 foot 1 inch(33.02 cm) long, 3 inches (7.62 cm) thick, and projected 2 inches(5.08 cm) above the floor.The opening into the cache pit or granary beneath the floor ofNo, 6 was 1 foot 9 inches (53.34 cm) in diameter. The bottom ofthe pit was 3 feet 8 inches (1.117 m) below the floor of No, 6. Thebottom of the cache pit was oval in contour with diameters of 4 feet(1.219 m) and 4 feet 10 inches (1.473 m). The wall was approxi-mately vertical for a distance of 1 foot 6 inches (45.72 cm) abovethe floor, then curved inward to the circular opening in the upperfloor. The stone slab set in the pit floor just below the opening was1 foot 6 inches (45.72 cm) high, 81/2 inches (21.59 cm) wide, and31/2 inches (8.89 cm) thick,Kiva A had a diameter of 15 feet 10 inches (4.826 m) on the sipapu,fire pit, ventilator line above the bench and 13 feet 5 inches (4.089m) below it. Along the direction of the plaster ridge on the floorthe diameter was 16 feet (4.876 m) above the bench and 13 feet 9inches (4,191 m) below it. On the ventilator side of the chamber thefloor was 8 feet 71/2 inches (2.628 m) below the present ground leveland 7 feet 71/2 inches (2.323 m) below the old surface. At the north-east end of the plaster ridge the floor was 7 feet 71/2 inches (2.323 m)below the present surface and 6 feet 7 inches (2.006 m) below theold ground level. At the southwest end of the ridge the depth was7 feet 21/2 inches (2.196 m) from the present ground level and 6feet 21/^ inches (1,891 m) from the old surface.
J92 BUREAU OF AMERICAN ETHNOLOGY IBclu 121The holes for the support posts were larger in this structure thanin many of the others. That near the north corner had diametersof 10 inches (25.4 cm) and 1 foot 2 inches (35.56 cm) and a depthof 2 feet 34 inch (62.86 cm). It was 1 foot 81/2 inches (52.07 cm)from the wall. The hole for the east post was 1 foot (30.48 cm)and 1 foot 4 inches (40.64 cm) in diameter and 2 feet i/4 inch (61.59cm) in depth. The hole was 6 inches (15.24 cm) from the wall.The hole for the support at the south corner was 8 inches (20.32 cm)and 9 inches (22.86 cm) in diameter and 1 foot 8 inches (50.8 cm)deep. It was 6 inches (15.24 cm) from the wall. The western postwas set in a hole with diameters of 9l^ inches (24.13 cm) and 1foot 2 inches (35.56 cm). The hole was 1 foot 6 inches (45.72 cm)from the wall.The fire pit in the floor of the kiva (fig. 45, ^), had diametersof 1 foot 101/2 inches (57.15 cm) and 1 foot II1/2 inches (59.69 cm).The bottom of the pit was 8% inches (22.22 cm) below the floor level.The surrounding rim of plaster increased the depth, however, to101/^ inches (26.67 cm). The stone in the edge of the pit was 1 foot 4inches (40.64 cm) long, 31^ inches (8.89 cm) thick, and its topprojected 1 inch (2.54 cm) above the plaster rim. The rim rangedfrom 4 inches (10.16 cm) to 7 inches (17.78 cm) in width and hadan average height of 1% inches (3.81 cm) above the floor. Theladder box or pit covered with the stone slab (fig. 45, h) was sep-arated from the fire pit by the stone slab. This pit measured 1 foot214 inches (36.83 cm) by 1 foot 4 inches (40.64 cm) and had a depthof 1 foot 21/^ inches (36.83 cm). The cover slab was 1 foot 41/2 inches(41.91 cm) by 1 foot 5 inches (43.18 cm) and 1 inch (2.54 cm) thick.At one side of the room the plaster ridge (fig. 45, /) was quite regu-lar in width and did not vary greatly from the 4i/2-inch (11.43 cm)average. The continuation at the opposite side, however, ranged be-tween 21/^ inches (6.35 cm) and 4 inches (10.16 cm). The heightwas quite consistent along the whole length and deviated only slightlyfrom the li/2-inch (3.81-cm) average.The cache or storage pit in the floor near the north support post(fig. 45, &) had diameters of 1 foot 3 inches (38.1 cm) and 1 foot6 inches (45.72 cm) . The depth was 1 foot 2 inches (35.56 cm) . Thehole was 9 inches (22.86 cm) from the hole for the support post and1 foot (30.48 cm) from the wall. The small pocket (fig. 45, c) haddiameters of 21/2 inches (6.35 cm) and 31/2 inches (8.89 cm) and adepth of 21/2 inches (6.35 cm). It was 81^ inches (21.59 cm) fromthe wall. The two holes occupying the position of the sipapu wereof slightly different size. The larger (fig. 45, d)^ was 4 inches(10.16 cm) and 41/2 inches (11.43 cm) in diameter and 2% inches(6.03 cm) in depth. The hole was 2 feet 5 inches (73.66 cm) fromthe wall. The smaller hole (fig. 45, e), had diameters of 3 inches
ROBiaiTS] ARCHEOLOGICAL REMAINS IN ARIZONA J93(7.62 cm) and 31/0 inches (8.89 cm). The depth was 2 inches (5.08cm). Hole e was 3 inches (7.62 cm) from d and 3 feet 5 inches(1.041 m) from the fire pit.The two cuplike depressions in the floor where the ends of theladder poles rested (fig. 45, %) were 1 foot 1 inch (33.02 cm) and1 foot 3 inches (38.1 cm) from the end of the slab-covered pit inthe position of the ladder box. The depressions were 1 foot 1 inch(33.02 cm) apart. They had diameters of 3I/2 inches (8.89 cm), sug-gesting that 3-inch (7.62-cm) poles were used in the ladder. Thedepressions were 1 foot 6 inches (45.72 cm) from the base of the ven-tilator opening, giving ample foot space for a person using theladder.The oval-shaped basin (fig. 45, jj) at the base of the wall near theeast support post had diameters of 1 foot 3 inches (38.1 cm) and 7inches (17.78 cm). The sides sloped downward from the rim of thebasin to a depth of 2I/2 inches (6.35 cm) near its center. The basintouched the wall along one side. It was only 3 inches (7.62 cm)from the edge of the hole for the post. The similar floor feature(fig. 45, h) near the south support post had diameters of 8 inches(20.32 cm) and IOI/2 inches (26.67 cm). Its depth was greater, being41/^ inches (11.43 cm). The hole was 31/2 inches (8.89 cm) fromthe edge of the hole for the post and 6 inches (15.24 cm) from thewall.The ventilator opening (fig. 45, I) was 9I/2 inches (24.13 cm) wideand 1 foot 4 inches (40.64 cm) high. The sill was 1 foot (30.48 cm)above the floor of the chamber. The original trench where it cutthrough the wall of the room was 2 feet 7 inches (78.74 cm) wide.The stonework reducing this size and forming a frame for the aper-ture was 1 foot (30.48 cm) wide at the west side of the opening and9 inches (22.86 cm) at the east side. On the west side the construc-tion extended back into the tunnel 1 foot 8 inches (50.8 cm), wherethere was a 5-inch (12.7-cm) jog. On the east side the masonry extended only 1 foot 2I/2 inches (36.83 cm) along the tunnel wall. Thejog was the same size as that on the opposite side, 5 inches (12.7 cm).The poles employed to reinforce the framework filling in the trenchopening (fig. 45, m) had diameters of 2i^ inches (6.35 cm) and 3inches (7.62 cm) . The stone slab used as a riser from the floor of thechamber to the sill of the aperture was 1 foot % inch (31.75 cm)high. The slight projection of its top above the sill was compen-sated for when the structure was occupied by the adobe plaster thatcovered the sill. The passage, from the aperture to the place whereit entered the bottom of the shaft, was 3 feet 4 inches (1.016 m) long.The width at the outer end was 1 foot 5 inches (43.18 cm). Theheight of the opening was 1 foot 7 inches (48.26 cm) . The bottom ofthe shaft measured 1 foot 9 inches (53.34 cm) by 2 feet li/^ inches
]^94 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 121(64.77 cm). The bottom of the shaft was 7 feet 9 inches (2.362 m)below the present ground level and 5 feet 11 inches (1.803 m) belowthe old surface. The opening on the ground level (fig. 45, n) haddiameters of 1 foot 1 inch (33.02 cm) and 1 foot IVs inches (34.29cm).The bank of the buried watercourse was only 1 foot 6 inches(45.72 cm) from the outside edge of the ventilator opening. Theside of the old channel dipped sharply to the bottom, which was4 feet (1.219 m) below the present ground level and 2 feet 2 inches(66.04 cm) below the former surface of occupation.The pockets on the bench between the base ends of the slantingroof poles were of different sizes. The circular one (fig. 45, o) at thewest corner was 5% inches (13.97 cm) in diameter and 3 inches (7.62cm) deep. The back edge of the pocket was against the back wallof the bench. Pocket p, figure 45, undercut the back wall of thebench to some extent. About half the pocket was beyond the lineof the wall. The pocket was 9 inches (22.86 cm) long, 6% inches(16.19 cm) wide, and 4 inches (10.16 cm) deep. It cut into the walla distance of 3I/2 inches (8.89 cm). The largest pocket of the groupwas q, figure 45, which was entirely in the wall at the back of thebench. It had a length of 1 foot 9I/2 inches (54.61 cm) and a widthof 81/2 inches (21.59 cm). The bottom of the pocket was 2 inches(5.08 cm) below the top of the bench. The opening into the pocketwas 1 foot 2 inches (35.56 cm) long and 4 inches (10.16 cm) high.The pocket near the south corner (fig. 45, r) was smaller. It alsowas entirely subwall in its form. The total length was 8 inches(20.32 cm) and the width 41/2 inches (11.43 cm). The opening was5 inches (12.7 cm) long and 3 inches (7.62 cm) high. The bottomof the pocket was only 1 inch (2,54 cm) below the level of the benchtop. The diameters for the holes for the butts of the roof timberswere consistently in the range of from 3 inches (7.62 cm) to 41/2inches (11.43 cm). The majority varied only slightly from a 314-inch (8.89-cm) diameter. UNIT NO. 2The second unit, also a part of the Developmental village, waslarger and more complex than the first one (pi. 22, a). It consistedof a surface structure containing six rooms, the remains of severalshelters, and a subterranean chamber designated Kiva B on theplan of the village (fig. 44). There is little question in this casebut that the circular structure had a ceremonial function, particu-larly in the final phase of its occupation. The refuse mound waseast of the house remains. A portion of it was beyond the water-course. Originally all of it had been separated from the housesite by the channel, but as it grew in size some of it contributed
RoBEitTs] ARCHEOLOGICAL REMAINS IN ARIZONA 195to the fill in the old stream bed and ultimately spread across it.The mound was much larger than that for the first unit. It cov-ered a larger area and also was deeper. The north end probablyconsisted in part of material deposited by people using the sheltersof the first unit and of that deposited by the one-time inhabitantsof the pit structure. Slightly more than the southern half unques-tionably was the accumulation of refuse from the second unit. Thispart of the mound was by far the deepest and, even making allow-ances for an increase in the size of the group living there, indi-cates a longer period of occupation.One noticeable feature in connection with the second unit is thatof the outside fire pits. They not only were around the house, be-tween it and the kiva, but were scattered over the surface of therefuse mound. In one or two cases there were indications that somesort of shade had been erected nearby. Most of them, however, wereused without any such accompanying construction. Whether theywere used for cooking fires or were simply places where peoplegathered, in the manner of modern Indians, for warmth and lightin periods of relaxation and sociability could not be learned. Asa matter of fact they probably served both purposes, although it islikely that most of the culinary activity centered around those wherethe walls of the surface building served as a windbreak or the userswere protected by an arbor.The buried watercourse passed much closer to the second unit thanto the first. Part of the kiva was even constructed in and below thechannel. The blocking of the stream some distance above the unitand diversion of the run-off from the higher slopes, as previouslymentioned, made possible the placing of the kiva in that location.Otherwise the structure would have been exposed to danger of dam-age from flood waters and a seepage of moisture into its pit. Theslope of the bank and the bed of the chaimel no doubt reduced theamount of digging necessary for the kiva pit, but also required someprovision for compensation for the lack of depth at that side. Thelocation was satisfactory from the standpoint of the surface waterdrainage because the slope was sufficient to carry away the ac-cumulations from falling rain and melting snow. As long as thediversion wall above the village held there was little likelihood of in-undation from the old channel. After the abandonment of the sitethe wall fell into disrepair and water began flowing toward the kivaalong the former course. The accumulations of refuse and otherdebris were such, however, that the flow was shifted to the right anda new channel cut. On one or two occasions, probably followingcloudbursts, enough water came down to overflow the banks andsweep sand and gravel across the ventilator portion of the kiva.
jgg BUREAU OF AMERICAN ETHNOLOGY [Bdll.12iThe surface structure in this unit approximated a rectangular blockin plan, although there was a small wing at one end and a niche orrecess at one side. The rooms were in a double tier with four in onerow and two and the niche in the other. The rooms were not uni-form in size. They fall roughly into three groups, however, withNos. 7, 10, and 12 corresponding quite closely in general measure-ments, 9 and 11 being approximately the same, and 8 standing alone.Three of the rooms were definitely equipped for dwelling purposesin that they contained fire pits and some form of storage bins. Theother three had no such features. They were the smaller chambersand it is quite possible that they were storage places rather thanactual living quarters. Granting such to be the case it is interest-ing to note the proportion of one dwelling room to one store room.On the basis of comparison with practices among some of the mod-ern village dwellers in the area the division in this structure suggeststhat the group here may have consisted of three units or families,a family consisting of husband, wife, and children. The relationshipin general possibly was that of mother and father with unmarriedchildren in one apartment and a married daughter with her husbandand children in each of the other two. The likelihood of some suchstatus is indicated by the evidence that two of the dwelling rooms,9 and 11, as well as two of the storage rooms, 7 and 12, were subse-quent additions to the original nucleus of 8 and 10. While it cannotbe proved definitely that such was the order, foundation and floorlevels, in relation to the old surface, indicate that 12 was added first,then 11, followed by 7, and finally by 9. Whether a part of the re-modeling process when 9 was built or the result of some other oc-currence could not be learned, but there was considerable reconstruc-tion in the wall separating 8 and 9.Construction methods and wall types were much the same in thesecond unit as those in the first. There was more horizontally laidmasonry in the second. The floors in the rooms were below thegeneral ground level and the foundations were laid along the wallsof the shallow room pits. The central portion of the structure, par-ticularly rooms 8 and 10, was characterized by the use of large, un-worked blocks of stone (pi. 22, h). This was true for both theslab foundation and the horizontally laid walls. In the later addi-tions smaller stones were employed, the proportion of mud mortarto building blocks was greater, and small spalls were inserted aschinking. There was abundant use of mortar and small stones ina sort of rubble finish on the exterior of many of the walls. Thiswas i^articularly noticeable on those with the upright slab founda-tions. The probabilities are that the necessity of compensating forthe uneven tops of the slabs and providing a comparatively level baseled to the development of that style finish. Added to tliis was the
ROBEBTS] ARCHEOLOGICAL REMAINS IN ARIZONA ]^97uneven nature of a wall made up of heterogeneous shapes and sizesof stones and the tendency to lay them so that the inside face wassmooth. Alining them along one side would make the irregulari-ties more pronounced on the other and the use of mud and smallfragments of stone would furnish a satisfactory means of obtaininga more regular surface. The practice of chinking with small tab-ular bits of stone may have developed out of tliis constructionalfeature. There does not appear to be any marked significance in thedifference in wall types. The fact that there was a blending andmixture of all three variations in the same wail and that the stepsleading from one form to another were obvious in the house suggestsa local development rather than an influence from the outside or achange resulting from alien increments to the village.Koom 7 was a simple enclosure with a flagstone floor. The onlyinterior feature was an upright stone set in the floor midway of thelength of the wall on the side toward the kiva side of the building.This is one of the stones that is thought to represent a step. Thisfunction was suggested not only by the position and height of thestone, but by the worn surface on the top as well. The ridges andrough spots were smoothed in a mamier quite indicative of thetread of feet. All other surfaces on the stone w^ere rough. Occur-ring midway of the wall it suggests that the opening or doorway intothe chamber was at that point and that the center of the wall wasthe customary place for a doorway. No evidence of that featureremained in the wall itself, although it seems likely that the sill couldnot have been much above the level of the remaining wall. Oneend of room 7 partially overlay the pit of one of the granaries thatis thought to be correlated with the remains of pit structure No. 17.The floor of room 7 cut into the fill in the granary to some extent,but its floor was 1 foot 3 inches (38.1 cm) above the floor of thegranary. The original ground level dropped away rather abruptlyat this end of the building and the north wall of the granary pro-jected above the old surface to a greater degree than usual for suchstructures. Three of the walls in room 7 were of the slab foundationand horizontal masonry type. The fourth was wholly of horizontalmasonry. It was the end of the original building, the outside ofroom 8. The two side walls of No. 7 abutted it.Room 8 was one of the dwelling chambers. The enclosure waslong and narrow. There was sufficient fallen material in the fill toshow that the walls had risen to the height of a single story, approx-imately 6 feet 6 inches (1.981 m) above the old ground level. Thisadded to the depth of the floor would give ample head room for theinhabitants. The upper part of the fill was mainly hard-packed sandand adobe mixed with wall stones. The adobe probably was slumped68764?39 14
j^gg BUREAU OF AMERICAN ETHNOLOGY [Boll. 121plaster from the wall and roof and the sand wind-deposited mate-rial. From the floor to approximately 1 foot (30.48 cm) above thefill was sand with some admixture of ashes. The nature of this layersuggested that the room was unoccupied for some time before thewalls fell in, the sand being blown in through the doorway. On thefloor were numerous potsherds and stone implements. The walls ofthe room were built of large, unshaped blocks of stone laid in courses.There was little chinking in three of them but the fourth, that sep-arating 8 and 9, did have small spalls in addition to the largerblocks. So many of these were used in places that they becamecourses of small stones rather than chinking. This wall, as pre-viously mentioned, appeared to be of later construction than the othersin this room and probably correlated with the remainder of room 9rather than with 8. This wall rested in part on the base of theoriginal wall for room 8, but the faces of the two did not coincide,and as a result there was a narrow oft'set along that side of the roomjust above the top of the stone step (pi. 23, a).The step in room 8 was similar to those in the other chambers. Itwas a rectangular block, unworked, set in the floor approximatelymidway of the eastern wall. The top was slightly worn and smoothedfrom use. The fire pit for the room was nearby. From the positionof the step, it would appear that the fire pit had been almost in frontof the doorway. The pit was partially lined with stones and partiallyfaced with plaster. A bin or storage box once stood in the north-west corner. The slab walls for this enclosure were missing, but thestone paving for the floor was still intact. Traces of the wall slabswere noted in the floor along the edges of the paving stones. Rec-tangular, corner bins of this type, made by using upright slabs, werecommon in Zuiii in comparatively recent times and occasionally oneis built even now. They served for the storage of beans, corn, and thelike.*** There was a small stone at the end of the room that had noapparent purpose. Its top was flush with the floor. It was not apart of the bin. There was nothing beneath it. The stone hadsimply been set in the floor at that place for some purpose notindicated by the remains. There were no other features in theenclosure.Room 9 was one of the later additions to the house. It was thelargest enclosure in the building and had more internal features thanthe others. There were a number of interesting items in its construc-tion. Three of the walls consisted, for the most part, of large, hori-zontally laid blocks of stone alternating with small, thin stones likethe construction shown in the upper part of the wall in the photo-graph (pi. 23, a) . The south end of the room presents a problem in
?? Mindelefif. 1891, p. 210.
BUREAU OF AMERICAN ETHNOLOGY BULLETIN 121 PLATE 23
a. Stone step and wall construction in room S.
h. ('orner bin in room 11.
BUREAU OF AMERICAN ETHNOLOGY BULLETIN 121 PLATE 2i
a. Corner of buil'iing outside room 11 and bin.
6. Original floor in Kiva B
BUREAU OF AMERICAN ETHNOLOGY BULLETIN 121 PLATE 25
a. Holes for slanting side poles at back of original bench top in Kiva B.
b. Katcina niche in wall of Kiva B.
BUREAU OF AMERICAN ETHNOLOGY BULLETIN 121 PLATE 26
1'"*'''^' IHMt^.
*?miNK^j
n. Suhwall storage bin in Kiva H.
6. Ventilator openings in Kiva B.
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA X99that there is a question as to whether or not it actually had a com-plete wall. The enclosure was finished, at that end by a row of up-right stone slabs. What form of construction had risen above the topsof these slabs, if any, is a matter of some doubt. From the amount ofdebris composed of small stones and chunks of mud mortar foundalong the bases of the slabs on both sides of the row, it would seemthat a rubble wall of the type previously described once stood there.Yet the tops and sides of the slabs were so completely devoid of anyadhering material of this nature that there was no actual evidenceto show that it once rested on and around them. For that reason astatement that there was a wall of such construction is not whollywarranted. On the other hand, it is questionable that the end of theroom was left open, except for the low barrier of standing stones, andas a consequence the best conclusion seems to be that there was arubble wall. It is possible, of course, that the enclosure was a shedrather than a room.Support posts for the roof were set at three places in the room. Onewas incorporated in the outside wall not quite midway of its lengthand the other two were along the inside wall at approximately one-third its length from each end. These uprights may have been placedthere to carry the main weight of the roof at the time the room wasbuilt or they may have been installed later as braces for a saggingceiling. The fact that the one on the outside was built into the wallis good evidence that it was a part of the original construction, or atleast was set there early in the occupancy of the room. Several possi-bilities are suggested by the combination of factors here indicated.There was a slight variation in the tj^e of masonry in the outer wallon each side of the post. That extending to the northeast corner andthence to the corners of rooms 7 and 8 was all of one form, while thatbetween the post and the end of the upright slab wall was noticeablydifferent, although conforming to the same general style. This sug-gests that the original room may have terminated at the post andthen subsequently was enlarged by the erection of the shorter lengthof wall and the slab construction at the end. Certain irregularitiesin the floor and below it, on a line between the outer post and the oneopposite near the inner wall, may have resulted from the razing of anend wall, which would substantiate the above postulation.The reinforcing of corners by the use of timbers was not a commonpractice in this district and it is curious that two examples shouldbe found in this room. One explanation for the occurrence is thatin the beginning room 9 was a portico like No. 2 in the first unitand that it was later built into a room, the support posts for the roofbeing included in the rising walls. In its final stage it unquestionablywas a room, because the fallen wall material was too extensive fora court. Another possibility is that a bit of misfortune was re-
200 BUREAU OF AMERICAN ETHNOLOGY [Boll. 121sponsible for the condition found, including the new wall betweenrooms 8 and 9. Because of the additional roof weight that would bedirectly correlated with the larger size of room 9, the original wallfor room 8 may have buckled. The attendant reconstruction wouldaccount for the type of masonry in the wall between 8 and 9, alsoits thickness and the use of uprights to aid in supporting the roof,the enlarging process taking place at the same time. There is nodefinite evidence that any of these explanations are correct. Theyare postulations offered merely to call attention to certain possibilitiesin the history of the room.There were two fire pits in room 9. The larger was in the approx-imate center and the smaller at one side. Both were partially linedwith stones and partially faced with mud plaster. At the time whenthe debris was cleared from the floor the smaller pit was filled to thebrim with fine wood ashes. Practically no charcoal was present inthe ash. The larger pit was only half filled with ashes and charcoal.The situation suggests that the smaller pit was functioning in thecapacity of an ash depository, while the large one was actually serv-ing as a place where fires burned. Just why there was the definitelyintentional saving of ashes as noted in so many instances in thisdistrict is not known. There were several uses for them, however.The adobe plaster occasionally was mixed with ashes, which gave ita harder, more cohesive quality. They also may have been neededin other pursuits for which there is no evidence. The modernZuiii, in preparing hominy, mix a quantity of wood-ash paste in a potof cold water and then put corn, removed from the cob, in it. Thepot is set over a fire and after the corn has come to a boil it is stirredwith a stick.^^ The stirring and boiling remove the hulls. Wood-ashlye was also used in the preparation of the well-known Pueblowafer bread. Similar culinary practices may have prevailed inearlier times and some of the ashes have been preserved for suchpurposes in this village.The storage bin or granary in the southwest corner of the roomwas built of stone slabs set on end. They were placed in a curvedrow that cut off the corner. Five stones were used in the construc-tion. They no doubt were covered with plaster at the time of occu-pancy of the house. The bottom of the bin was below the floorlevel, as in the case of that in No. 2 in the first unit. The shape ofthe feature, when considered in the light of the jog in the wall atthat end of the room, suggests that it might have been in existenceprior to the completion of the chamber and then incorporated in itduring the remodeling. There was no evidence either for or againstthis possibility, but the general form is indicative of it. When the
?= Stevenson, 1904, p. 367.
ROBBETS] ARCHEOLOGICAL REMAINS IN ARIZONA 201slabs in the recess along the wall of room 11 are included it looksvery much as though there had been a large circular granary infront of rooms 8 and 10 and that it was partially dismantled whenroom 9 was erected. In fact, the end V7all of 9 may have cut acrossthe granary instead of the smaller bin being incorporated in it.To provide a level floor in room 9, because of the natural slopeof the ground wliich dropped away rather abruptly from in front ofthe original structure, it was necessary for the builders to fill inalong the outside wall. Approximately one-third of the floor spacewas treated in this way. Refuse-mound material and broken stoneswere employed for that purpose. The mud-plaster flooring waslaid on top of this fill (fig. 49). The outer wall was close to the wallof the pit for the subterranean structure, but with the roof in placeand covered with earth the latter would have had no effect on the sur-face house. As a matter of fact the area was probably leveled off sothat the roof of the kiva formed a dooryard in front of the dwelling.Room 9 did not have one of the stone steps like those in the otherrooms. For that reason there is no indication of the location of thedoorway. The most suitable place for an opening would have beenat the south end of the room next to the corner storage bin. If itwas placed there a step would not have been needed, as the buildingup of the floor level at that side of the room brought it to about theground level and a person entering the chamber would not be re-quired to step down into it as on the others. That may explain theabsence of the step.Room 10 was a simple chamber of the storage type. Its wallswere similar in construction to those of room 8, except for the oneat the south where a slab foundation type was used. This wall ap-peared to be a part of the original structure, although it is possiblethat it was rebuilt when room 12 was added. Digging the pit forthe floor of the latter enclosure could have produced some settlingin the masonry, with the result that a new piece of construction wasdeemed advisable. The only additional feature of interest in themasonry was in the northwest corner of the room, where the normaljuncture of the walls was modified by a short, diagonal series ofcourses that gave the floor plan a pentagonal instead of a quad-rangular form. There was nothing to indicate whether this thick-ening of the walls was intentional or accidental. Midway of thewall on the side toward room 11, in the position occupied by thestone step in other rooms, a small post had been set in the floor. Itsoriginal height is not known. Because of its location it is possiblethat it was also a step, although one of wood rather than stone. Onthe other hand, it may have been a brace to hold a covering in placeover the doorway. Since it stood some distance from the wall thestep interpretation seems more logical. The floor in this room was
202 BUREAU OF AMERICAN ETHNOLOGY [Bvhh. 121mainly smoothed plaster, but there were two slabs embedded in thesurface at the north end of the enclosure.When the interior of room 10 w^as being cleared of its accumulateddebris an unusual mass of beads was found on the floor. They wereall in one group, as though they had been in a pouch or container ofsome perishable substance that had disintegrated after being coveredwith earth. There was no apparent reason for their presence in theroom. Probabilities are that they either were dropped and not re-covered or that the bag was hidden there and then for some reasonor other was not retrieved. The beads are all made from shell, somepink, some w^hite, some red, and a few a slight orange shade. Therewas a variety of sizes from the standpoint of diameter, but none werevery large. Most of them were a simple disc in form. A few wereshaped like a figure 8 w4th the perforation for stringing through thesmaller end. There was a marked consistency in thickness and theyaveraged 20 beads to the inch. There was no way of telling whetherthey had been strung or what kind of necklace, if strung, theyformed. When threaded on a single strand they made a string 37feet 4 inches (11.379 m) long. There were approximately 9,000 beadsin the group. In addition to the beads, a number of pieces of azuriteand malachite and a few fragments of turquoise were present in theassemblage. They, too, seemed to have been in the same containeras the beads.The recess or niche between rooms 9 and 11 gave no evidence thatit had ever been roofed. It would have been a simple matter, how-ever, to lay timbers across from the end of one room to the end ofthe other and thus provide a covering for the space. Under the cir-cumstances no traces of such a roof would remain and it is impossi-ble to say what the nature of the enclosure was when the buildingwas occupied. It may have been open to the heavens or it could havebeen topped with timbers and brush. With a roof it would havemade a serviceable adjunct to the house. Various objects could havebeen kept there out of the weather. The indication of a bin at theback of the recess has been mentioned already in connection withroom 9. If the upright slabs placed here were not a part of a formergranary they at least partially enclosed the southwest corner at therear of the recess and contributed to its qualifications as a place forkeeping things. It was in the corner in this enclosure that the humanfoot, mentioned in the discussion of the skeleton in Kiva A, wasfound sealed behind adobe plaster.Room 11 was similar to room 9 in its general size and shape. Itdid not have any support posts, however, and the wall constructionwas mainly of the slab foundation rather than the horizontal blocktype, although there was some of the latter in two places. A littlemore than half of the north wall was built in that fashion and a
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA 203section near the southeast corner also consisted of large blocks laidhorizontally from the ground up. The slab-founded walls in thisenclosure had a larger percentage of rubble exterior finish than anyother part of the structure. This was particularly true of the oneat the south end. The fire pit in the room was merely a small cir-cular basin in the floor. It was not as deep nor as large in diameteras those in rooms 8 and 9 and it had no stone facing. Close to thefire pit, in almost the center of the room, was a metate or millingstone. Midway of the outside wall was a hole in the floor where apost had been placed. It occupied a position similar to that of thepost in room 10 and may also have been a wooden rather than astone step. On the opposite side of the wall, outside of the room,a stone similar to the steps in several of the rooms was set in theground close to the base of the wall. It unquestionably was a stepand indicates that the doorway to the chamber was in that portionof the outer wall. A person could easily have stepped from thestone through the opening and onto the short wooden post inside.The southwest corner of the room was approximately bisected by alarge stone slab. It appeared to be all that remained of a bin orstorage place comparable to those in the corners of 2 and 9. Therest of the enclosing wall probably was of rubble construction.There was sufficient debris, composed of small irregularly shapedrocks and chunks of mud plaster, on the floor in that vicinity tosuggest that such had been the case. The actual limits of the bincould not be determined. The mud and rock part of the wall appar-ently fell before the remainder of the room had collapsed and thefloor was so damaged that traces of the wall's footing were notdiscernible.The most interesting feature in room 11 was the granary or stor-age place incorporated in the wall near the southwest corner. Whilein a broad sense it was comparable to those in rooms 2 and 9, itactually was unique for the district. Although smaller, its generalform was much like that of the regular detached granaries. A shal-low pit was lined with slabs placed upright around its edges. Theslabs projected some distance above the ground level and were toppedwith courses of horizontally laid masonry. The actual wall wasthicker than the slabs and from the ground level to the beginningsof the horizontal courses the additional "body" was supplied by rub-ble construction. There were a few scattered places where two orthree stones were laid in a semblance of horizontal masonry ; but thebulk of the wall around the slabs was of the rubble type. A door-way connected the granary with the room (pi. 23, h). The sill ofthe opening consisted of a stone slab set on edge. The jambs ateach side were single upright slabs to a height of 1 foot 6 inches
204 BUREAU OF AMERICAN ETHNOLOGY [Boll. 121(45.72 cm) and 2 feet (60.96 cm) above the floor. Above these thesides of the opening had been the horizontal block masonry. Thefloor of the closet was 9 inches (22.86 cm) below the floor of the roomand was paved with thin slabs of stone. The fallen wall materialin the granary and along its outside walls indicated that it had risento approximately the same height as the room. Corner closets orgranaries of this type are not common. A good example of one wasfomid in a Great Pueblo ruin on the Zmli Reservation.^^ The lat-ter is probably a century or more later in date and it is curious thatso advantageous a feature in house form was not more widely used,particularly since it was known for so long a period. Constructionalcomplications and the tendency to a weakening of the walls may haveoverbalanced the convenience of such a storage place to the extentthat it did not attain to popular use.Room 12, like 7 and 10, had no interior features of note. Thewalls were of the slab-foundation type and that on the side towardroom 11 was very irregular. There was no apparent reason for theuneven surface or the unusual thickness of the construction. Thewall at the end of the building rested upon and incorporated in itsfoundations a few slabs from an older granary. It is possible that asimilar condition prevailed along part of the wall between rooms 11and 12 and that the variability of both surface and thickness weredue to the inclusion of parts of an older structure. This was notapparent on the exterior surfaces of the wall, however, and the con-struction was not disturbed.There was no step in room 12. Wliether there ever had been one,or one was present and subsequently removed, could not be deter-mined because the enclosure had been used as a burial place and theinterment had caused some disturbance of its original status. Thesurface of the floor was broken in several places along the base of thewall in the approximate location for a step, judging from the posi-tions of those in other chambers, and any traces of its placementthat might have survived were obliterated. The burial was that, ofan adult, probably a male, and was peculiar in that the body hadbeen placed front down with the face turned to one side. The legswere tightly flexed. The left arm extended along the side and theright was crossed over the back. The positions of the arm bonessuggested that the person had been bound, his hands tied behind hisback. The individual appeared to have been dumped on the floorand then covered with refuse, stones, and mud plaster. The latterwas poured around and over the stones but not smoothed. It set andheld the covering firmly in place over the remains. Why the personwas buried in this manner would make an interesting story. The
?? Roberts, 1932, p. 31, flg. 1.
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA 205fact that there were no accompanying mortuary offerings and thatthe individual apparently was tied suggests a captive or prisoner ofwar. Being disposed of in a room rather than interred in the refusemound, as was the case with most of the burials, indicates some ex-ceptional circumstances, possibly a desire to conceal the fact tliatthe person had been disposed of or had succumbed. There v?asnothing to show how long before the abandonment of the house theinterment took place. The people may have lived there for sometime subsequent to that event or it is possible that it was a contribut-ing factor in the group's moving. One other skeleton exhibiting asimilar position, sprawled face down with the arms crossed over theback and the left foot drawn up as though tied to the lashings thatheld the wrists, was uncovered in a shallow pit beyond the limits ofthe refuse mound (fig. 44, No, 4). This burial also lacked funeraryofferings and the implication is that it was the final resting place ofan unwelcome alien or a vanquished foe.At the southwest corner of the building, adjacent to room 11 andits granary, was a partially enclosed area containing a fire pit. Thissheltered spot, in the lee of the house, probably served as an outdoorkitchen. A masonry wall extended from the granary several feettoward the southeast, forming a protection on that side. At thecorner formed by the juncture of the slab-founded and horizontalmasonry walls in room 11 a single large stone slab was set on endto aid in delimiting the area. First impressions were that thishad been an additional room, but no evidence was found for theother two walls or that the construction had ever been more exten-sive. A stone placed near the horizontal masonry wall was anal-ogous in position and shape to the steps in the various rooms. Itdoes not seem to have had this purpose, however. It would have beenserviceable as a seat or a place to set objects used in cooking. Thefire pit was roughly circular in form and faced with stones (pi. 24, a).Close to the fire pit, on the north side, was a long stone slab, em-bedded in the floor of occupation. The top of the stone was flushwith the level of the surrounding area. Underneath the stone werethe remains of a young child. The body had been placed in a shal-low pit. The arms were folded across the chest and the kneesdrawn up close to the body; it was a tightly flexed burial. Smallpoles were laid across the top of the pit and the slab placed onthem. The stone was large enough to extend beyond the rim of thepit and it is likely that the grave was sealed by the use of mudplaster around the edges of the stone. Children were often buriedclose to a hearth in the Southwest and there undoubtedly was somesignificance in the custom. The actual reasons for the practice arenot known, but by analogy from a similar custom among the modernIndians in the area several suggestions can be offered. One explana-^toto^
206 BUREAU OF AMERICAN ETHNOLOGY [BiJi-l. 121tion is that when buried near the center of domestic activity the littleone will be less lonesome. Another is that it is an effort to keep thefamily circle unbroken, a custom that has been world-wide in itsdistribution among peoples of less developed cultures. Or it mightbe that the mother wanted the remains kept as near to her aspossible as the result of some purely maternal reaction to the lossof a cherished offspring. Another youngster was interred in thearea on the opposite side of the wall, not as close to a fire pit asthe former, but well within the boundaries of all household activi-ties. Both of these burials belonged to the final stages of occupancyof this unit because the mortuary offerings of pottery consisted ofvessels of the type correlating with the final phase of the Develop-mental village.The area bounded by room 11, the granary, the extension wallfrom the granary, and the kiva was also a center of domestic activ-ity. There was a slab-lined fire pit in the approximate center ofthe space and the old surface was packed and stained from thetramp of many feet. There were several standing stones near thefire pit. Their purpose was not apparent, but they undoubtedlyhad some connection with the pit. Possibly they were merelyprovided as places on which to put pots or baskets so that theywould be off the ground. There were no traces of support postsand the area probably had no shade or covering erected over it.There was a similar center for outdoor cooking in the corner formedby rooms 7 and 9. The slab-lined fire pit located there was morethan half filled with ashes and small fragments of charcoal. Nu-merous splinters of burned bone occurred in the pit and on the sur-face around its rim. An arrowhead and a bone scraper were foundat the base of the wall of room 9 near the corner.Southeast from the surface structure, across the old channel andat the edge of the present watercourse, were the remains of a brushshelter. This example was more elaborate than many and quitereminiscent of the one located at the main site on top of the ridgeclose to structures 12 and 13. This shelter was suggestive of asmall pit dwelling in some respects but it was not deep enough,although the floor was below the surrounding ground level, andhad no form of ventilator. The shade or arbor erected over theshallow pit was supported by four upright posts set near the cor-ners of the depression. There were no indications that side wallswere provided. Slanting poles may have been used to form a wind-break on two or more sides, but there were no traces on the groundaround the edges of the floor pit that could be regarded as anindication of the presence of such timbers. The fire pit was at oneside of the floor space close to the north support post. The pit wasD -shaped. The flat side was faced with a stone slab and the
RoBEHTs] ARCHEOLOGICAL REMAINS IN ARIZONA 207curved face was covered with plaster. A plaster rim encircled thepit on the floor level. Two stones were placed near the west cornerof the pit. One was set in the floor 6 inches (15.24 cm) from thewall in a position suggesting that it might have been a step. Theother was lying on the old ground level and its location in relationto the first was such as to indicate that it may have been a treadto protect the edges of the pit from damage when people steppedup or down. There were two storage places along the southwestwall. One was a circular pit in the floor (fig. 46, d) and the other asubwall recess (fig. 46, e). The storage pit w^as merely a hole in thefloor with vertical sides and approximately level bottom. Both thesides and the bottom were covered with plaster. The pocket inthe wall was close to the circular pit. It was cut in the nativeearth, the back wall sloping down to form a curved back. The bot-tom was a continuation of the general floor level of the main partof the shelter. The pocket was not deep nor high but would havebeen a convenient place to keep various objects.One question in connection with this shelter and that on the ridgeabove concerns the possibility that they actually were the basalportions of houses rather than mere shelters. The depressed floorsand the definite provision of storage places suggests something morepermanent than an ordinary summer kitchen. As pointed out inthe discussion of both places, there were no evidences of slantingside poles from the framework at the top of the main posts to theground back of the edges of the floor pits. The character of theground is such, however, that any traces left by the butts of tim-bers placed in that way could be obliterated without much diffi-culty. Hence the possibility that faint indications of poles were missedin the digging. Both structures could have had flimsy jacal wallsand all traces of them have vanished. If such was the case themore elaborate floor features would not be out of place. Jacal con-struction was not unknown in the region, as evidences of housesmade in that way were found at the Long H Kanch in a horizon ap-proximately contemporary with that in this village. *'^ As a conse-quence, some form of it could have been used here, although therewas nothing to show that such had been the case. The most logicalconclusion to draw from the evidence is that the places were brushshelters, but the possibility of their being remnants of another formof house should be borne in mind,Kiva B contained a number of interesting features and is thefirst of the underground structures thus far considered to exhibitmost of the characteristics of a ceremonial chamber. The struc-ture had two distinct phases with a marked difference in certain
" Roberts, 1931, pp. 86-88.
208 BUREAU OF AMERICAN ETHNOLOGY [Bull. 121
5 FEET
Figure 46.?Brush shelter, a, holes for roof supports ; 6, fire pit ; c, stone step ; A, stor-age pit in floor ; e, subwall pocket ; t, stone tread ; g, present surface ; h, old groundleveL
RoBEBTs] ARCHEOLOGICAL REMAINS IN ARIZONA 209constructional elements in each. As far as this locality is concerned,Kiva B probably marks the transition stage in which there was adefinite shift from the older pit-type dwelling to the strictly cere-monial chamber. This was evidenced by a change in the type ofroof construction and a modification of some other features to theextent that they produced a complex comparable in all respects tothe circular ceremonial rooms of later horizons.The pit for Kiva B was roughly circular in form and was in-closed by a definite bench (fig. 47). The bench was partially dugand partially built up. At several places around the periphery theexcavation for the lower portion of the pit had cut back too far intothe bench and it was necessary to fill in with refuse and rubblebehind the masonry wall that formed the face of the bench. On alittle more than half the circumference of the upper and back wallof the bench masonry construction was used to compensate for thelack of depth resulting from the slope of the ground toward theold watercourse. This stonework was not a part of the originalstructure. It was put there during the remodeling process whenthe character of the chamber was changed. During the first stagethe addition to the back wall was not needed, as the low earth facewas sufficient. When the alterations were made, however, the benchlevel was raised and the builders had to add to the wall height.The face of the bench or wall of the main part of the chamber wasof stone construction. On about three-fourths of the periphery thewall consisted of horizontally laid stones resting on the tops ofupright slabs (pi. 24, &) ; the remainder was horizontal masonryfrom top to bottom. Some of the slabs used in the facing were sotall they extended almost to the bench top and had only one or twocourses of stones above them; others were very short and weretopped by many layers. The horizontal masonry in this structurewas better than that in Kiva A. Smaller stones were used for themost part and there were some indications of attempts at shapingand dressing a number of the blocks.The covering over the pit in its final stage of use consisted of acribbed roof, the form of superstructure characteristic of later-dayceremonial rooms. None of the timbers were present when theremains were uncovered, but the top of the bench clearly showed theimprint of logs lying along it in the positions that the bottom rowof timbers would occupy in a cribbed roof. Furthermore, there wereno holes for upright support posts in the floor that correlated withthis phase. The original roof, however, was of the truncated pyram-idal type with sloping side poles and flat center supported on fouruprights. The holes for the uprights were found in the old floorbelow the plaster of the upper floor and when the addition to thetop of the bench was removed to the old bench level some of the
210 BUREAU OF AMERICAN ETHNOLOGY [Bull. 121placements for the slanting side poles were uncovered (pi. 25, a).These occurred sporadically around the bench. In some cases the
^ODI'?ct:oQ(
PiGDBE 47.?Kiva B, upper floor level, a, storage hole in floor ; 6, Katcina niche ; c, stor-age hole in floor ; d, sipapu ; e, fire pit ; f, storage hole ; ff, ash pit ; h, stone in floor
;
i, stone in floor ; j, subfloor ventilator trench ; k, ventilator shaft ; m, post incorporatedin bench ; n, present ground level ; Oj original surface.butts rested in the angle formed by the back wall and the top of thebench ; in others they were set in the shallow holes.When the old style superstructure was in place there was littleneed for a high back wall to the bench. As long as there was enough
Roberts] ARCHEOLOGICAL REMAINS IN ARIZONA 211of a face to hold the side poles in position, nothing more wasrequired. The brush, earth, and plaster covering would be carriedto the ground level and a low bench wall would be an advantage inthat all necessary provision for drainage would be supplied withoutthe use of too much fill material. Due to the domelike nature of acribbed roof and the practice of filling in with larger quantities ofdirt until the surface surrounding the hatchway was comparativelylevel, a much higher back wall was needed to hold the earth coveringthe logs. This reason, coupled with the fact that the original backwall was decreased in height by the construction of a new top onthe bench, made it essential to build up the back wall on the ven-tilator side and explains the courses of horizontally laid masonryfound there. On the side toward the house the pit was deep enoughto answer all requirements. AVhen the fill over the cribbed roof wasput in place, however, the workmen did not stop at the old groundsurface but carried the level higher, so that a comparatively flat areawas formed in front of the surface building. In all probability thedooryard, from the walls of rooms 9 and 11 to the kiva entrance,was approximately level during the final period of occupancy in theunit.Interior features varied somewhat in the two stages of occupancy.Those belonging to the later period, that of the cribbed roof, con-sisted of the fire pit, sipapu, a Katcina niche, several small holes inthe floor for storing minor articles, two stones embedded in the floor,and a ventilator. The fire pit (fig. 47, e) was D -shaped and linedthroughout with small stone slabs. The stones projected above thefloor and were incorporated in an encircling rim of adobe plaster.This pit was built over and around the pit belonging to the olderfloor level. In comparison with those in many of the houses it wasshallow. The stones and plaster rim were colored from the fires thatburned there and the lower part of the pit was filled with ashes andsmall fragments of charcoal. None of the latter were of sufficientsize to furnish material for dendrochronological studies. The posi-tion normally occupied by the ladder box or ash pit, in the pit houses,was taken over by the end of the subfloor ventilator. At one side ofthis feature, however, was a small rectangular pit (fig. 47, g) thatcontained some ashes. The pit was shifted to one side, but it hadapproximately the same relative position with respect to the fire pitas in structures where there was no ventilator opening of the typefound here.The sipapu (fig. 47, d) was a simple circular hole with verticalsides. The sides and bottom were carefully plastered. The holesa, (?, and / were mere plain, plastered storage places of the kinddescribed for the other structures. One of them, /, was in the loca-tion of one of the support posts for the superstructure of the original
22^2 BUREAU OF AMERICAN ETHNOLOGY [Bull. 121chamber. It was not as large in diameter as the hole for the postand its bottom penetrated only a short distance into the fill. The flatstone embedded in the floor at one side of the fire and ash pits(fig. 47, A) was in approximately the same position and about thesame size as others noted in the discussion of various structures.As previously suggested, these stones may have been rests for adrum or basket serving as a drum, like those in some of the kivas inmodern pueblos. This function would be in keeping with the indi-cation of ceremonial character for the chamber. Of course, it mayhave had some use entirely distinct from religious observances. Thesecond stone (i) stood several inches above the floor. Its lower endpassed entirely through the upper floor and was embedded in thehole for the support post that formerly stood there. There wasnothing to suggest the purpose of this stone.The Katcina niche was located in the wall at the side of the roomopposite from the ventilator (fig. 47, h). It was placed high up inthe masonry, just above the top of one of the upright facing slabs(pi. 25, &). The niche was a rectangular box lined with small stoneslabs. Similar niches are frequently found in the walls of later-date kivas at the same side of the chamber. They are also present inthe kivas of today in the Hopi country and at Acoma. The nameis taken from the Hopi term for them. They may or may not havehad a comparable designation in olden times. The full Hopi name isKatcina Kihu, the house of the Katcinas, the Katcinas being imper-sonations of the gods. During the progress of a ceremony, the masksworn by certain dancers are placed in the niche when not in actualuse. The significance among the Acoma is slightly different; it isconsidered as the "doorway" through which the spirits of the godsenter and leave the kiva. Also, when prayers are offered up to thedeities of the Sun and the Moon, and of the northern, eastern, andwestern mountains, they are made into that opening. Occasionallyin ruins, pipes, pieces of turquoise, and fetishes are found in theniches. There was no way of telling whether or not the niche in thisstructure was present in both stages of occupancy. It may have beenprovided when the alterations were made or it could have been a partof the original chamber. They are rare in the older pit dwellings,however, and it is more likely that it was a later addition.The ventilator in this structure is interesting because it is the onlyexample of the subfloor type found in the excavations. In additionto the usual form with an aperture in the wall, there was a trenchextending across the floor from the side wall to the fire pit. Thistrench started below the normal vent opening in the wall. It wasthe same width as the tunnel portion of the main ventilator and waslined with thin stone slabs. When the chamber was in use it hadbeen covered with a series of stone slabs that were plastered over to
Roberts] ARCHEOLOGICAL REMAINS IN ARIZONA 213conform with the general floor surface. Near the fire pit was arectangular opening that provided passage for the air coming in
:x. -^^Q
W'''''MW//'mj'M'/,wmmym
Figure 48.?Kiva B, original floor level, a, holes for main roof posts ; J), stone in floor ;0, storage box ; d, sipapu ; e, fire pit ; f, ladder box ; g, stone embedded in floor ; h, stor-age pit ; i, storage bole ; /, subwall storage box ; k, stone in floor ; m, stones embedded onbench top ; n, holes for slanting side poles ; o, ventilator shaft ; p, present ground level ;q, original surface.through the tunnel beneath the floor. This subfloor feature wasnot part of the original house. It was installed when the changeswere made at the time of putting in a new floor and adding to the68764?39 15 ,
214 BUREAU OF AMERICAN ETHNOLOGY [Boll. 121bench height. This was shown by the fact that the wall slabs of thetrench cut through the old floor and the plaster rim around the oldfire pit was broken out to make room for the end stone. There wassome reconstruction work around the base of the vertical aperturein the wall and the floor of the passage was dug out to the extent thatit was below the ends of some of the slabs lining the walls just backof the old opening. Subfloor ventilators of this type are not uncom-mon in kivas of later stages in Pueblo development. They rarelyoccur in the earlier phases, however, and the present example is one ofthe oldest to come to light thus far. The type is not unusual for thisgeneral region. Two kivas excavated by Hodge near Hawikuh,"^and two at another site on the Zuiii Reservation ^^ exhibited thefeature. One of the latter was quite comparable to the present inthat it exhibited the same double nature with a horizontal vent in thefloor and a vertical opening in the wall. There probably is somesignificance to the subfloor type of ventilator, but what that may behas not yet been determined. Thus far there have been no indicationsof causes or reasons for its development nor definite evidence of itsparticular group affiliations.The passage through the wall of the structure and the shaft atthe outer end were essentially the same for both ventilators. Theoriginal tunnel was built up. A trench was dug, lined with stoneslabs, roofed with a series of slabs, and then covered with a dirt fillto the level of the bench top. The original floor had sloped upwardfrom the sill of the vertical opening to the bottom of the shaft. Thelater floor sloped gradually upward from the edge of the rectangularpit just below the vent in the floor to a point near the outer end ofthe old passage and then turned sharply up to meet the bottom ofthe shaft. The shaft was more irregular in contour at the bottomthan in most of those in this district. Instead of having an ovalor circular form it was practically pentagonal. The lower portionwas lined with upright slabs. The remainder of the shaft had hori-zontally laid masonry resting on the tops of these standing stones.The space where the passage entered the shaft was bridged with aheavy lintel of stone which supported the wall of the shaft on thatside.The presence of two ventilators raises the question as to whetheror not both were used at the same time. It is possible that one or theother openings was stopped; in fact there is a suggestion that thereAvas a cover slab for the vertical opening. When the debris was re-moved from the room during the process of excavation a number ofstones were found on the floor in front of the vent aperture. Some
88 Hodge, 1923.?? Roberts, 1932, pp. 54-55, 67-
ROBEETS] ARCHEOLOGICAL REMAINS IN ARIZONA 215were part of the wall above the opening, that had fallen in, andothers were pieces from a larger slab that had been broken. Thelatter in its original form would have answered the purpose ofa covering for the hole. That it actually did so is a matter ofuncertainty.There were no indications of a deflector and it is probable thatnone was present in the chamber. As a matter of fact such an ar-rangement would not be necessary with a subfloor ventilator, as aircoming into the room through the horizontal opening in the floorwould not strike the fire but pass upward over it. Some kivas inlater ruins where that type of ventilator is present also have a de-flector, while others do not. The ruin on the Zuni Reservation, previ-ously cited, where two kivas contained the subfloor ventilator, had anexample of each, one with and one without the deflector feature.^"The placement of the ladder in this structure may have provedsomewhat of a problem. Because of the opening in the floor nextthe fire pit, it would need to rest farther away from the pit unless itsbase actually stood in the vent opening. The latter position wouldnot be entirely out of the question and may be the explanation forthe boxlike pit in the end of the subfloor passage immediately belowthe rectangular opening in the floor. The ladder poles would notinterfere to any great extent with the incoming air and the wall ofthe pit would prevent their slipping and damaging the stone cover-ing the passage and framing the opening at that side. The coverstone was thick enough to have supported a person using the ladder
;
in fact from the standpoint of ordinary activity in the chamber,there would be no difference between the ordinary floor and thepresent example after plaster had been applied over the stones roof-ing the ventilator trench. Hence there may have been a combinationladder box and vent opening. The only alternative would havebeen to set the ladder on the floor beyond the opening and beyondthe ash pit (fig. 47, g). There were no traces of pole abrasions in theplaster on either side of the trench to indicate such a placement.Of course the ladder could have rested on top of the covering overthe trench and all signs of its presence have been lost by the collapseof the roofing. This situation, however, would not be wholly satis-factory because there probably was a ridge in the floor above the ventpassage due to the failure to set the cover slabs entirely below itslevel, and some likelihood of the ladder slipping to one side or theother. Without definite evidence in the matter, the better conclu-sion seems to be that the lower ends of the latter poles did rest inthe box below the opening in the floor.
? Roberts, 1932, pi. 12.
2Jg BUREAU OF AMERICAN ETHNOLOGY [B^ll. 121The original floor in the chamber had a number of features notpresent in the later level of occupation. Before discussing them itis necessary to mention one item of note concerning the distinctionbetween the two horizons. That is, there was a definite line ofdemarcation between them in the form of a layer of purple-coloredclay. The latter was ordinary earth with a heavy admixture ofpurple hematite. A layer of refuse-mound material was first spreadover the old floor, filling all holes and depressions, and then thehematite mixture was laid down. After this was evenly distributedover the surface the plaster of the upper floor was applied. Whatsignificance the use of hematite in this fashion may have had is notknown. Occasionally a kiva or pit house is found where it was em-ployed in a similar way to that in this structure. Consequently itwas more than a restricted local development.Features on the old floor level consisted of a fire pit, sipapu, holesfor the support posts for the superstructure, storage pits, a subwallstorage cist, and stones embedded in the floor. The fire pit (fig.48, e) was of the same general style and sliape as that on the laterlevel. It was somewhat smaller, however, and did not have as largean encircling rim. The entire pit had been faced with small stoneslabs and the bottom had stone paving. Installation of the subfloorventilator obliterated all traces of a ladder box, if there was one.There is a possibility that the ventilator opening may have incor-porated the former ladder box and the rectangular pit at the end ofthe subfloor trench be the remains of such a box. There was nothingin the construction to show that that had been the case, but the con-tingency should be mentioned. The question of the former presenceof a deflector is another that cannot be answered. Any evidencerelative to that feature was destroyed by the remodeling activitieswhen the new ventilator was installed. The sipapu was not in thesame location as the later one, although it had similar general charac-teristics. It was merely a circular hole with vertical walls and sidesand bottom covered with plaster. The diameter and the depth werenot quite as large as the one on the first level.The stone set on the floor near the base of the wall (fig. 48, &) atthe north side of the chamber was similar in nature to the one lo-cated at the south side in the upper floor. It was of squarish shapewith a flat top and projected several inches above the floor. Thebase was firmly embedded in the plaster. There was no indicationof the purpose of the stone. At one side of the fire pit a large slabwas embedded flat in the floor, its top flush with the plaster, in ap-proximately the same position as the one described for the later level(fig. 48,^). This stone was larger than the first one, but in allprobability it was placed there for the same purpose. Another stonewas set in the floor on edge in a position suggesting that there
Roberts] ARCHEOLOGICAL REMAINS IN ARIZONA 217originally was a compartment on the ventilator side of the room(fig. 48, k). The stone as found did not represent the former totalheight, because the top showed that it had been broken off. Therewere no traces of additional stones on the opposite side of the venttrench, but the floor in that area had been sufficiently damaged toremove any indications of that nature. Hence it is possible that thestructure did have a compartment during its early stage. On theother hand, it may well be that there was only a storage bin at oneside and that the remaining stone is all that had been placed there.The storage places for this stage of occupancy were interesting be-cause they were more carefully constructed than most of those previ-ously described. At the base of a wall on the side opposite the ven-tilator, just below the Katcina niche, was a rectangular pit (fig. 48, c)with vertical sides and an approximately level bottom. The outsidewall was faced with a large stone slab. The inner or back wall wasformed by the slabs in the main facing of the wall of the chamber.The two ends were of earth covered with a thick coating of plaster.This is the only example of such a bin in the entire group of exca-vated structures. A smaller and better storage place, somewhat com-parable to the one just described, was situated at the base of the wallnot far from the south support post (fig. 48, j; pi. 26, a). This boxwas dug into the floor and extended below the bench so that approxi-mately half of it was under the wall. The pit was rectangular inshape and faced on all four sides with stone slabs. The edges of thestones projected slightly above the floor level, forming a rim aroundthe opening. A box of this type would have been very convenient forkeeping articles safe from accidental damage and out of the way ofthe occupants in the chamber. Both of those boxes or bins had beenfilled with refuse before the second floor was laid.The basin or depression in the floor midway along the wall betweenthe two bins just described was similar to many noted in other struc-tures. The front edge was curved and the back was formed by thebase of the wall of the room (fig. 48, h). The sides sloped down-ward from the edge on the floor level to the bottom. The interiorwas plastered. Nearby was a small, circular hole (i) that probablywas a pocket for small articles, such as awls, punches, and otherimplements.One interesting feature that occurred above the bench level was theniche in the back wall above the ventilator (pi. 26, h). Wliether itwas present in simpler form in the original construction or was intro-duced at tlic time of the remodeling was not apparent. From a gen-eral point of view the most logical supposition is that it was one ofthe elements added when the style of roof was changed. The sig-nificant factor in its occurrence is that it apparently was an adum-brant form of the recess in the kivas of later horizons. Several ex-
218 BUREAU OF AMERICAN ETHNOLOGY [Edll. 121amples of the feature were found in the ceremonial rooms in the largeruin at the Long H Ranch southwest from this site ;^^ it was presentin most of the kivas in the village of the Great Kivas on the ZuniReservation southeast from the Allantown district ; ^^ and is quitecommon in the regions to the north. It occurs at Mesa Verde," inthe ruin at Aztec," N. Mex., in the Chaco Canyon,^^ and in northeast-ern Arizona.'^'' The feature is particularly prominent in the unit-house ruins investigated by Prudden along the San Juan River,^'' andseems to be a northern characteristic. The purpose of the recess isnot known. They may be analogous to the spectator's bench in theHopi kivas,'^? the platform where members of the group not partici-pating in ceremonies gather to watch the performance of the rites.The present example was too small for any such use, but it has thesame general shape and location and can be considered as the fore-runner of the larger forms.One aspect of the niche raises a question that is difficult to answer.With a cribbed type of roof starting from the bench, provision of ameans of access to the niche would not be a simple matter. In thelater forms of kivas the roof timbers generally were supported bypilasters and raised high enough above the bench to leave a good-sized opening to the recess. That was not possible when the roofstarted from the bench. Hence the question concerning the open-ing, A small aperture could be provided by starting the lower ringof logs at each side of the break in the wall and the additional tim-bers could be placed in such a way that it would not be necessaryfor a beam to cross in front of the niche until the third or fourthtier was reached. This may have been done. There is the possi-bility that no opening was provided and that as far as the interiorof the chamber was concerned there was no recess. That such wasthe case does not seem reasonable, though, in view of the careful wayin which the niche was built. Some sort of doorway undoubtedlywas left there. The niche here is the earliest example thus far re-ported and indicates that the feature developed rather soon afterthe beginning of definite ceremonial chambers. Wliy it was presentin some kivas and not in others is a problem still to be solved.The fill in Kiva B had several features that threw some light onconditions subsequent to the abandonment of the structure. On theupper floor, the last one used, were several thin layers of water-deposited clay mixed with bits of charcoal. These indicate that
'i Roberts, 1931, pp. 93, 97, 100, 107.
?^ Roberts, 1932, pp. G7-69, 72-73, 81, 83-84."^sFewkes, 1909, pi. 1 ; 1911, b, pi. 8.
'* Morris, 1924, map of ruin.'Mudd, 1930, ground plan of Pueblo Bonito.? Kidder. 1924, fig. 12, p. 69." Prudden, 1914."Fewkes, 1911, p. 24.
Roberts ] ARCHEOLOGICAL REMAINS IN ARIZONA 219flood waters found their way into the pit while the superstructurewas still in place. Above these layers was a thick deposit of wind-blown sand containing some ash and sporadic pieces of charcoal.This stratum probably represents an appreciable interval, as it wasrather evenly spread throughout the pit and of greater depth thanaverage for this kind of material in the district. Above this was athick stratum of coarse clay streaked with lines of decayed wood.The level unquestionably represents the stage when the roof col-
-Sections in first and second units. 15 FEETlapsed and fell into the partially filled pit. The clay was themelted-down plaster and roof material. On top of this layer werehalf a dozen thinner strata composed of water-washed sand and claycontaining some small fragments of charcoal and occasional tracesof ashes, and two thicker layers of sand, clay, and charcoal. Theseunquestionably were carried into the depression by an intermittentseries of small floods sweeping down across the site. When the lat-ter had been deposited the pit was filled to the top of the masonrywall at the back of the bench. Then several layers of clean sandand gravel were left there and these in turn were covered by wind-
220 BUREAU OF AMERICAN ETHNOLOGY tBuLt. 121blown material. In fact a small dune was formed over the remainsof the structure and part of the fallen walls at the southeast sideof the surface building. The water action responsible for the layersabove the bench level in the pit also damaged the upper part of theconstruction in the ventilator shaft and washed out a portion of theback wall in the niche above the ventilator tunnel. The chief sig-nificance of the fill is in the evidence it gives that the site wasabandoned at the same time as the kiva and that neither was re-occupied. There was no dump material in the fill. The traces ofcharcoal and ashes, and the sporadic potsherds present were car-ried into the pit from the surrounding surface and were not thrownin as refuse from a habitation. Had people continued to live in thesurface house there would undoubtedly have been some trash on thekiva floor or in the layers between the floor and the stratum formedby the decay and collapse of the superstructure. Also, if the sitehad been occupied again at a later date there would have beensome signs of it in the higher strata in the pit.Koom 7 was 7 feet 6 inches (2.286 m) long and 5 feet 4 inches(l.C)25 m) wide. Three of the outside walls, those that were slab-founded, averaged 1 foot (30.48 cm) in thickness. The fourth, whichwas the end of the original building, averaged 7 inches (17.78 cm).The stone step was 1 foot ?> inches (38.1 cm) high, 6y2 inches (16.51cm) wide, and 214 inches (5.71 cm) thick. It stood 2 inches (5.08cm) from the wall.Room 8 was 16 feet (4.876 m) long and 4 feet 10 inches (1.473 m)wide. The north wall was 7 inches (17.78 cm) thick; the east, abovethe offset where it rested on the original wall, averaged 1 foot 5inches (43.18 cm); the south averaged 1 foot 4 inches (40.64 cm),and the west 1 foot 4 inches (40.64 cm). The step was 1 foot 4inches (40.64 cm) high at the front edge and 1 foot 3 inches (38.1cm) high at the side next the wall. The stone was 5 inches (12.7cm) wide and 4 inches (10.16 cm) thick. It stood 1 inch (2.54 cm)from the wall. The fire pit measured 2 feet 8 inches (68.58 cm) by1 foot 10 inches (55.88 cm) and was 7 inches (17.78 cm) deep. Thestorage bin at the corner of the room had a length of 4 feet (1.219 m)and a width of 1 foot 3 inches (38.1 cm). The height of the wallwas probably 2 feet 6 inches (76.2 cm) but this could not be deter-mined for certain. The offset along the east wall varied in width.Along a greater part of the length it averaged 4 inches (10.16 cm)but at one point toward the north end of the room reached a maxi-mum of 1 foot (30.48 cm).Room 9 was 14 feet 9 inches (4.495 m) long and ranged from 8feet (2.438 m) to 8 feet 9 inches (2.667 m) in width. The wall at thenorth end averaged 1 foot 5 inches (43.18 cm) in thickness. Theeast wall varied from 1 foot 4 inches (40.64 cm) at the northeast
ROBBKTS] ARCHEOLOGICAL REMAINS IN ARIZONA 221corner to 8 inches (20.32 cm) at the southeast. The south wall wasentirely slabs and their thickness averaged 314 inches (8.89 cm). Thewest wall, that between rooms 8 and 9, averaged 1 foot 5 inclies(43.18 cm) thick. The fire pit measured 2 feet 4 inches (71.12 cm)by 1 foot 9 inches (53.34 cm) and had a depth of 7 inches (17.78cm). The asii basin measured 1 foot 6 inches (45.72 cm) by 1 foot3 inches (38.1 cm) and was 6 inches (15.24 cm) deep. The bin inthe southwest corner was 4 feet 3 inches (1.295 m) long and 2 feet 5inches (73.66 cm) wide. The posts set at various places in the flooraveraged 6 inches (15.24 cm) in diameter.Room 10 was 7 feet 9 inches (2.362 m) long and 5 feet 4 inches(1.625 m) wide. The north wall averaged 1 foot 5 inches (43.18cm) in thickness, except at the northwest corner where it attaineda maximum of 2 feet 9 inches (83.82 cm). The east wall ranged from8 inches (20.32 cm) to 1 foot (30.48 cm). The south averaged 6inches (15.24 cm) and the west varied from 11 inches (27.94 cm) to1 foot 5 inches (43.18 cm) in thickness. The butt end of the postthat has been considered as representing a wooden form of the stepliad a diameter of 5 inches (12.7 cm). The post stood 2 inches (5.08cm) from the wall.Room 11 had a length of 13 feet 6 inches (4.114 m), was 8 feet 1inch (2.463 m) wide at the north end, 7 feet 6 inches (2.286 m) wideat the location of the doorway into the granary, and 5 feet 6 inches(1.676 m) wide at the south end. The north wall averaged 7 inches(17.78 cm) in thickness. The east measured 8 inches (20.32 cm) andthe masonry section near the granary 2 feet 2 inches (66.04 cm).The south wall was 1 foot 6 inches (45.72 cm) thick at the south-east corner and tapered to 1 foot 3 inches (38.1 cm) at the southwestcorner. The west wall was somewhat irregular, measuring 2 feet(60.96 cm) at the southwest corner, 2 feet 3 inches (68.58 cm) at theapproximal center, and 1 foot 3 inches (38.1 cm) at the northwestcorner. The fire basin had diameters of 1 foot 6 inches (45.72 cm),1 foot 2 inches (35.56 cm), and a depth of 4 inches (10.16 cm). Theslab at the southwest corner, that had formed a part of a storageibin, was 2 feet 5 inches (73.66 cm) long, 31,4 inches (8.89 cm) thick,and 2 feet (60.96 cm) high. The metate lying on the floor had alength of 2 feet (60.96 cm), a width of 1 foot 6 inches (45.72 cm),and a total thickness of 4 inches (10.16 cm). The trough was 1 foot9 inches (53.34 cm) long, 10 inches (25.4 cm) wide, and 214 inches(5.71 cm) deep. The step at the east side, that is the wooden post,had been placed 2 inches (5.08 cm) from the wall and had a diam-eter of 514 inches (13.33 cm). The stone step on the outside was8 inches (20.32 cm) wide, 5 inches (12.7 cm) thick, and 1 foot 4inches (40.64 cm) high.
222 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 121The granary opening off room 11 had an inside measurement of2 feet 11 inches (88.9 cm) by 1 foot 9 inches (53.34 cm). The door-way was 1 foot 3 inches (38.1 cm) wide. The top of the stone sillwas 3 inches (7.62 cm) above the floor of room 11 and 1 foot 3 inches(38.1 cm) above the floor of the bin. The granary floor was 1 foot(30.48 cm) below the floor of the room. The w'all averaged 9 inches(22.86 cm) in thickness.Room 12 was 7 feet 9 inches (2.362 m) long, 5 feet 4 inches (1.625m) wide at the north end, narrowed to 4 feet 3 inches (1.295 m) nearthe center, and measured 4 feet 4 inches (1.320 m) at the south end.The west wall was 9 inches (22.86 cm) thick and the south 10 inches(25.4 cm). The north was the same as the south for room 10 and theeast corresponded to the west wall of room 11.The wall extending toward the southeast from the outside of thegranary had a length of 6 feet 3 inches (1.905 m) on the north sideand 6 feet 9 inches (2.057 m) on the south. It had an average thick-ness of 1 foot 2 inches (35.56 cm), stood 2 feet ll^ inches (64.77 cm)high at the west end and 1 foot y^ inch (31.75 cm) at the east end.The stone slab extending from the corner of room 11 to delimit theother side of the area was 1 foot 10 inches (55.88 cm) long, 1 foot 1inch (33.02 cm) high, and 21^ inches (6.35 cm) thick. The fire pit inthis corner area measured 1 foot 10 inches (55.88 cm) by 1 foot 8inches (50.8 cm) and was 8 inches (20.32 cm) deep. The large stonenearby that covered the child burial was 3 feet 4 inches (1.016 m)long, 1 foot (30.48 cm) and 1 foot 7 inches (48.26 cm) wide, and 2^/^inches (6.35 cm) thick. The stone near the masonry wall was 10inches (25.4 cm) high, 7 inches (17.78 cm) wide, and 4 inches (10.16cm) thick. It stood 8 inches (20.32 cm) from the wall.The fire pit in the area on the opposite side of the wall was 1 foot(30.48 cm) long, 9 inches (22.86 cm) wide, and 9 inches (22.86 cm)deep. The stone in the floor close to it measured 3i/^ inches (8.89 cm)by 4 inches (10.16 cm) and stood 3/2 inch (1.27 cm) above the occu-pation level. The second stone, a little farther removed from thefire pit, was 4 inches (10.16 cm) long, 1 inch (2.54 cm) wide, and1% inches (3.81 cm) high. The fire pit in the corner formed by thewalls of rooms 7 and 9 measured 1 foot 4 inches (40.64 cm) by1 foot 7 inches (48.26 cm) and was 7 inches (17.78 cm) deep. It was3 feet 4 inches (1.016 m) from the wall of room 7 and 2 feet 1 inch(63.5 cm) from the wall of room 9.The brush shelter at the southeast corner of the site was 15 feet6 inches (4.724 m) from the end of the masonry wall outside thegranary of room 11 and 15 feet 4 inches (4.673 m) from the kiva.The floor depression of the shelter was 8 feet 6 inches (2.590 m)by 7 feet 414 inches (2.247 m). At the north side the floor was111/2 inches (29.21 cm) below the old ground level and at the south
Roberts] ARCHEOLOGICAL REMAINS IN ARIZONA 223side 1 foot 8^2 inches (52.07 cm) below it. The holes for the sup-ports for the arbor varied in diameter. That near the northeastcorner was 6 inches (15.24 cm). The depth was 1 foot 1 inch(33.02 cm). The southeast hole had diameters of 7 inches (17.78 cm)and 91/^ inches (24.13 cm) and a depth of 1 foot 2 inches (35.56 cm).The southwest hole was 7% inches (19.05 cm) by 10 inches (25.4 cm)on its two diameters and 11 inches (27.94 cm) deep. The northwesthole was 6 inches (15.24 cm) by 7 inches (17.78 cm) with a 1 foot1 inch (33.02 cm) depth. All four holes touched the side walls.The fire pit had diameters of 1 foot 7 inches (48.26 cm) and 1 foot8 inches (50.8 cm). The bottom was only 5^ inches (13.97 cm)below the floor level of the room but the encircling rim of plaster in-creased the depth to 8 inches (20.32 cm). The rim averaged 5 inches(12.7 cm) in width, except for a length of 6 inches (15.24 cm) nearthe stone where it measured 4 inches (10,16 cm). The rim had anaverage height of 2^/2 inches (6.35 cm). The standing stone nearthe northwest corner (fig. 46, c) was 6I/2 inches (16.51 cm) wide, 4inches (10.16 cm) thick, and 8l^ inches (21.59 cm) high. Itstood 1 foot (30.48 cm) from the wall. The storage pit (fig. 46, d)had diameters of 1 foot 4 inches (40.64 cm) and 1 foot 6 inches(45.72 cm) and a depth of 1 foot 6 inches (45.72 cm). At one pointon the periphery it touched the wall. The subwall niche (fig. 46, e)was 1 foot 7 inches (48.26 cm) long at the opening in the wall,1 foot 3 inches (38.1 cm) long at the back, and had an average depthof 5 inches (12.7 cm). The opening in the wall was 6 inches(15.24 cm) high.The fire pits scattered over the surface of the refuse mound didnot vary greatly in size, although there was a considerable rangein depths. There were four with diameters ranging from 2 feet3 inches (68.58 cm) to 2 feet 6 inches (76.2 cm) and depths from71/2 inches (19.05 cm) to 2 feet (60.96 cm). The two smaller pits,both of which were rectangular, measured 1 foot 2 inches (35.56 cm)by 1 foot (30.48 cm), with 6 inches (15.24 cm) depth, and 1 foot 7inches (48.26 cm) and 2 feet (60.96 cm), and 1 foot 10 inches (55.88cm) deep.Kiva B had diameters of 18 feet 9 inches (5.715 m) above thebench and 15 feet 4 inches (4.673 m), below the bench, on the sipapu,fire pit, ventilator line. Across the chamber in the opposite direc-tion the diameter above the bench was 19 feet 6 inches (5.943 m)and below it 14 feet lOi/s inches (4.533 m). At the ventilator sideof the room the upper or last floor was 10 feet 1 inch (3.073 m)below the present ground level and 4 feet 11 inches (1.498 m) belowthe surface at the time of occupation. The old floor was 3 inches(7.62 cm) below the later one at this point. The top of the benchwas 3 feet 10 inches (1.168 m) above the upper floor. In its original
224 BUREAU OF AMERICAN ETHNOLOGY [Boll. 121form the bench top was just 3 feet (91.44 cm) above the old floorlevel. At the opposite side of the chamber, in front of the Katcinaniche, the last floor level was 9 feet 7 inches (2.921 m) below thepresent ground and 6 feet 3 inches (1.905 m) below the old surface.The original floor was 6 inches (15.24 cm) below the second one atthis point. The bench top was 4 feet 5 inches (1.346 m) above thesecond floor level and for the original construction the old top was3 feet 1 inch (93.98 cm) above the old floor. At the east side of thechamber the floor was 8 feet 2 inches (2.489 m) below the presentsurface and 5 feet 3 inches (1.600 m) from the old level. The laterfloor was 3 inches (7.62 cm) above the original one. The top of thelast bench level was 3 feet 8 inches (1.117 m) above the later floor.The original bench top was 3 feet 1 inch (93.98 cm) above the oldfloor. At the west side of the room the last floor was 9 feet 9 inches(2.971 m) below the present surface and 5 feet 6 inches (1.676 m)below the old level. The old floor was 4 inches (10.16 cm) belowthe later one. The last bench top was 3 feet 9 inches (1.143 m) abovethe last floor and the original floor was 3 feet 3 inches (99.06 cm)below the top of the original bench.The holes for the support posts for the original superstructurewere less variable in size than in many of the structures previouslydescribed. That for the north support had a diameter of 6^/^ inches(16.51 cm) and a depth of 1 foot (30.48 cm). The hole was 1 foot6 inches (45.72 cm) from the wall. The hole at the east side of theroom had diameters of 6 inches (15.24 cm) and 7 inches (17.78 cm)and a depth of 1 foot 3 inches (38.1 cm). The hole was 5% inches(13.97 cm) from the wall. The hole for the south support post haddiameters of 514 inches (13.97 cm) and 6 inches (15.24 cm) and adepth of 1 foot 2 inches (35.56 cm). It was 8I/2 inches (21.59 cm)from the wall. The hole at the west side of the room had diametersof 6 inches (15.24 cm) and 7 inches (17.78 cm). The hole was 2 feet10 inches (86.36 cm) from the wall and had a depth of 1 foot 3 inches(38.1 cm). The holes at the back of the original bench top, wheretlie lower ends of the slanting roof poles rested, ranged from 21^to 3% inches (6.35 to 9.52 cm) in diameter. They were quite shallow,the depths varying from 2 inches (5.08 cm) to 3 inches (7.62 cm).The two pockets at the back of the bench measured 8I/2 inches(21.59 cm) by 5 inches (12.7 cm), with a 4-inch (10.16-cm) depth;and 91/2 inches (24.13 cm) by 61/2 inches (16.51 cm) with an 8-inch(20.32-cm) depth.The Katcina niche was 4 inches (10.16 cm) high, 6I/2 inches (16.51cm) wide, and 11 inches (27.94 cm) deep. It was 1 foot 9 inches(53.34 cm) above the final floor and 2 feet 7 inches (78.74 cm) belowthe last bench top. It was 2 feet 1 inch (63.5 cm) above the originalfloor and 8 inches (20.32 cm) below the old bench top.
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA 225The fire pit for the room in its final form had inside diameters of2 feet 21/2 inches (67.31 cm) and 2 feet 7 inches (78.74 cm). Thedepth at the center was 6 inches (15.24 cm). The plaster ridgeencircling the pit ranged from 4 inches (10.16 cm) to 6 inches (15.24cm) in width. On the sipapu side it had a height of 1 inch (2.54 cm)and on the side toward the ventilator was 4 inches (10.16 cm) high.The other, underlying fire pit had diameters of 2 feet 2l^ inches(67.31 cm) and 2 feet 9 inches (83.82 cm) and a depth of 4 inches(10.16 cm).The sipapu for the last or upper floor was 2 feet 9i^ inches (85.09cm) from the fire pit. The hole had a diameter of 5 inches (12.7 cm)and a depth of 6 inches (15.24 cm). The sipapu for the originalfloor was 4 feet (1.219 m) from the fire pit. This hole had diam-eters of 3 inches (7.62 cm) and 4 inches (10.16 cm) and a depth of5 inches (12.7 cm). The small storage hole (fig. 47, c) near the baseof the wall on a line with the sipapu of the upper floor level haddiameters of 3i/^ inches (8.89 cm) and 4 inches (10.16 cm). It was 8inches (20.32 cm) from the wall and had a depth of 3i/^ inches(8.89 cm). The storage hole placed inside the hole for the eastsupport post of the old level (fig. 47, /) had a diameter of 4 inches(10.16 cm) and a depth of 5 inches (12.7 cm). The ash pit or box-like pit in the floor near the end of the ventilator tunnel (fig. 47, g)measured 1 foot (30.48 cm) on the length paralleling the vent passageand 10 inches (25.4 cm) on the opposite direction. The pit was 4inches (10.16 cm) deep.The stone embedded in the floor (fig. 47, h) was 1 foot 2 inches(35.56 cm) long, 9I/2 inches (24.13 cm) wide, and 2 inches (5.08 cm)thick. The stone on the old floor, in approximately the same posi-tion, was somewhat larger (fig. 48, g). It had a length of 1 foot lli/^inches (59.69 cm) and a width of 1 foot 1 inch (33.02 cm). Thethickness was 214 inches (5.71 cm). The upper floor stone that wasplaced in the hole for the south support post (fig. 47, i) measured 2inches (5.08 cm) by 3 inches (7.62 cm) and was 3 inches (7.62 cm)high. The similar stone located at the opposite side on the originalfloor level (fig. 48, h) measured 2i/^ inches (6.35 cm) by 3 inches(7.62 cm) and was 2% inches (6.35 cm) high. It was 11/2 inches(3.81 cm) from the wall. The stone slab that suggested part of acompartment on the floor of the original chamber (fig. 48, k) was2 feet 5y2 inches (74.93 cm) long, 3 inches (7.62 cm) thick, and stood2 inches (5.08 cm) above the floor. That was not its original height,however, as the top of the stone had been broken off.The subfloor bin at the base of the wall below the Katcina niche(fig. 48, c) was 5 inches (12.7 cm) wide, 2 feet (60.96 cm) long, 10inches (25.4 cm) deep. The stone forming the face of the outside wallof the pit was 2 feet (60.96 cm) long and li^ inches (3.81 cm) thick.
226 BUREAU OF AMERICAN ETHNOLOGY [Bull. 121The top projected 1 inch (2.54 cm) above the old floor level. Thesubfloor, subwall storage box at the south side of the room measured1 foot 1 inch (33.02 cm) in length, 9 inches (22.86 cm) wide atone end, 8 inches (20.32 cm) wide at the other end, and 8 inches(20.32 cm) deep. The box extended 4 inches (10.16 cm) under thewall.The large storage basin at the base of the wall in the old floor atthe southwest side of the chamber (fig. 48, h) had a total lengthof 2 feet 5 inches (73.66 cm) and a total width of 111/2 inches (29.21cm) and a depth of 5 inches (12.7 cm). The small hole betweenthis basin and the subwall box (fig. 48, i) had diameters of 21/2inches (6.35 cm) and 4 inches (10.16 cm). The depth was 3 inches(7.62 cm).The horizontal ventilator opening in the floor near the fire pitof the upper level measured 1 foot 51/2 inches (44.45 cm) on thepassage line and 1 foot 4 inches (40.64 cm) in the opposite direction.The rectangular pit directly below this opening was 1 foot 31/2inches (39.37 cm) long and 1 foot 3 inches (38.1 cm) wide. Thebottom was 1 foot 5 inches (43.18 cm) below the upper floor leveland 5 inches (12.7 cm) below the sloping bottom of the subfloortrench. The subfloor trench was 4 feet 4 inches (1.320 m) long,from the end near the fire pit to the point where it penetrated thewall of the chamber. Its depth at the edge of the opening in thefloor was 1 foot (30.48 cm) and where it passed under the wall was10 inches (25.4 cm). The width ranged from 1 foot 21/2 inches(36.83 cm) to 1 foot 3 inches (38.1 cm). The point where it joinedthe old passage floor at the edge of the shaft at the outer end was5 feet 4 inches (1.625 m) from the edge of the wall.The vertical opening in the wall of the room, the aperture to theoriginal ventilator, was 1 foot 1 inch (33.02 cm) wide and 1 foot(30.48 cm) high. The sill was 6 inches (15.24 cm) above the floorlevel of the latest occupancy and 9 inches (22.86 cm) above that ofthe original room. The sill stone was 1 foot 3 inches (38.1 cm) long,61/2 inches (16.51 cm) wide, and li/o inches (3.81 cm) thick. Theoriginal passage was 5 feet 4 inches (1.625 m) long. Where it en-tered the shaft it was 1 foot II/2 inches (34.29 cm) wide and 2 feet2 inches (66.04 cm) high.The base of the shaft, that portion lined with upright slabs, was2 feet 7 inches (78.74 cm) long and 1 foot 5 inches (43.18 cm) wide.The bottom of the shaft was 3 feet 10 inches (1.168 m) below theold ground level and 7 feet 6 inches (2.286 m) below the presentsurface. The opening on the old ground level measured 1 foot 6inches (45.72 cm) by 1 foot 5 inches (43.18 cm).
Roberts] ARCHEOLOGICAL REMAINS IN ARIZONA 227UNIT NO. 3The third unit, located 1 mile (1.609 k) up the valley from thefirst and second, was a typical single-clan or unit type of the lateDevelopmental Pueblo period, the stage usually referred to as PuebloII. A surface structure with six rooms and a court, an exteriorstorage pit, one outside fire pit, and a subterranean chamber or kivacompleted the group. The above-ground dwelling was located ona slight knoll, just beyond the foot of the talus of the ridge lyingto the south, so that it had been sufficiently elevated above the sur-rounding terrain to assure proper drainage on all sides. An inter-mittent watercourse that carried the run-off from the higher groundpassed the site on its western side, not far from the southwestcorner of the dwelling. This stream bed was probably present atthe time of occupation, the ensuing years serving to broaden anddeepen the channel but not to change its location. At two placessome distance above the ruins were traces of masonry walls on thebanks of the channel. Their position and nature indicated thatsmall check dams had been placed there, probably to hold back someof the water flowing along the course after rains, and to formsmall reservoirs for the community. The refuse mound for the unitwas situated to the east of the structure. It was fairly deep, as themiddens for such units go, and had served as a cemetery. The ante-cedent pit structure. No. 18, previously described, was between thenortheast end of the surface building and the refuse mound.The surface structure was built on an approximation of the blockplan with a double tier of rooms, 2 in one row and 4 in the other(fig. 50). The structure was not erected as a complete building ona preconceived ground plan, but, as in the case of the second unit,grew as the result of additions to an original nucleus. The firstconstruction was that of rooms 2 and 4. This was followed by theerection of rooms 3 and 5, and subsequently by 1 and 6. There wasnothing to show the relation between 1 and G. Wliether one or theother was added first or both represent a single stage of activity wasnot evident in the remains. The general status, however, indicatedthat there was no great interval of time represented by the variousadditions. It is quite possible that 3 and 5 were started soon afterthe completion of 2 and 4 and that 1 and 6 were not long in follow-ing. The construction work was much better in this building than ineither the first or second units. The walls were of horizontally laidmasonry. One course of large blocks served for the foundations.Above these the stones were smaller, more tabular in shape, andformed a solid, compact wall. Many of the blocks were shaped anddressed. There was a greater percentage of stone to mud mortar inthis structure than in the preceding ones and more use of small
228 BUREAU OF AMERICAN ETHNOLOGY [Bull. 121
?.,...*'" "'o
.,../
< o
?$v*>"
a .2
'?''III,,,.,," '/(nil lillV*'' ''//ll'
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA 229chinking stones. The walls stood only from 1 foot (30.48 cm) to1 foot 6 inches (45.72 cm) in height at the time of excavation, butthere was sufficient fallen material to show that they formerly wereof single story height. The debris indicated that the masonry hadbeen consistent throughout. The general appearance of the remain-ing walls was comparable to that of the late construction in the sur-face building of the second unit, room 9 particularly, although in thisinstance the work was better done and exhibited an improvedtechnique (pi. 27, a).The floors were approximately on the old ground level. Theywere not depressed like those in the first units. There was notenough wall construction remaining to show the number or natureof the openings provided in them. There undoubtedly hadbeen doorways between 2 and 3, and 4 and 5, but they were highenough in the wall to have disappeared with the collapse of theupper courses. This also holds true for exterior openings for thechambers comprising the front row. In accord with the generalcustom of this stage the sills probably were from 1 foot 6 inches(45.72 cm) to 2 feet (60.96 cm) above the floor level, hence wouldhave been a course or two higher in the wall than the tops of theremaining portions. Two of the rooms had fire pits, three con-tained storage bins or basins, and the remaining chamber had nointerior features. The combination suggested a structure for twofamilies. Rooms 1 and 6 were unquestionably for storage purposes,while 3 and 4 may have combined the functions of dwelling andstorage places. This indicates a higher percentage of space perfamily than that suggested for the second unit. Such a relation,however, would be in keeping with later trends where three rooms,occasionally more, to a family unit was a common situation.Room 1 had little in the way of interior features. It seems tohave been for storage purposes. There was a small bin near onecorner. A low wall of horizontally laid stones formed the enclosure.The floor of the room was of hard-packed adobe plaster.Room 2 was larger than the first chamber, although it was asjacking in interior features. It was part of the original build-ing and also seems to have served for storage purposes. The sizewas such, though, that it could well have augmented the actualliving quarters as well as being used as a place in which to keep sur-plus food supplies and other objects. Close to the wall, separatingit from room 3, was an oval-shaped basin in the floor. There wasnothing in it to indicate its purpose. It had been dug into theground and was lined with plaster at the time when the floor waslaid. The latter continued unbroken down the sides and across thebottom of the basin.68764?39 16
230 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 121Room 3 was one of the dwelling chambers. It was of particularinterest because it had two floor levels. The upper one contained arectangular fire pit lined with slabs of stone, while the lower one hada simple plaster-lined circular basin for that purpose. The upperfloor also had a small ash pit adjacent to the fire pit. When theroom vras excavated this pit was filled with fine wood ashes. Com-pletely underlying both floors and the foundations of room 3, aswell as a portion of room 2, vvere the remains of a shelter that seem-ingly was similar to those previously described. It was not possi-ble to trace its entire outline, but the holes for the four posts thathad supported tlie shade or arbor were found, as well as a smallstorage pit, a shallow depression in the floor, and a small fire pit.All of these were below the layer of red clay that the house founda-tions rested on and that was discussed in connection with the pitstructure, No. 18. This occurrence suggests that the shelter was asso-ciated with the old pit dwelling. At all events it definitely ante-dated the surface building of which room 3 was a part (fig. 52).Room 4 was another of the chambers without a fire pit. Pre-sumably it also was for storage purposes; possibly was a combinationliving and storage chamber. Approximately midway of the roomalong the outside, western wall, was an oval-shaped basin in thefloor. The wall of the room formed a part of one side of the basin.Here, as in room 2, the lining of the sides and bottom was of plaster,a continuation of the layer that formed the floor of the chamber.When the room was being cleared of the debris from fallen wallmaterial the skeleton of a turkey was found on the floor. The birdeither had been confined and died there or its body tossed into thechamber after death. The bones were articulated, showing that theremains were those of a complete fowl and not kitchen scraps. Thebird had not been kept there long because there were no indicationsof droppings on the floor. Traces of a short occupancy could wellbe missed, but one of any duration would have been quite evident.The turkey may have been left there when the house was abandonedand succumbed from lack of food and water. Or its demise merelymay have coincided with the period of departure and the bodyplaced there as a final act on the part of one of the inhabitants. Thenumber of definite turkey burials in this district was such as to indi-cate that any haphazard disposal of dead fowl was not common pro-cedure. Hence it is quite possible that lack of time to bury it prop-erly led to the body's being deposited in the room when the groupwas moving from the site. Certainly it would not have been leftthere in a decomposing condition if the house had continued to beoccupied.Room 5 apparently was the other definite living chamber. Itcontained a rectangular, slab-lined fire pit near the center of the floor
ROBEETS] ARCHEOLOGICAL REMAINS IN ARIZONA 231area, an adjoining rectangular plaster-lined asli pit, and the remainsof a storage bin that had been placed between the fire pit and theoutside wall of the room. The combination of fire and ash j)its isinteresting because the position, nature, and general status is the sameas that found in room 3. The complex seems to have been a structuralfeature and not a fortuitous association. This no doubt representsprovision for saving ashes that had the same purpose behind thepractice as was suggested in the discussion of similar features in theother units. The bin between the fire pit and the eastern wall wasroughly crescentic in form. The Ioav partition that formed the en-closure was constructed of upright stone slabs and adobe plaster.Four of the stones were still in position, and although the others hadfallen, the outline of the bin was clearly apparent on the floor. Oneof the slabs used in its wall also served as part of the lining in thefire pit. The stone was embedded in the bottom and eastern face ofthe fire pit and rose to sufficient height to fill in that portion of thebin wall. The edge of the ash pit at its southeast corner was at thebase of the bin. This feature probably functioned in the same generalway as other storage places. A variety of things could have beenplaced there. The floor in room 5, like those in the other chambers,was of hard-packed, well-smoothed adobe plaster.Room 6 was a simple rectangular enclosure with no interior fea-tures. It undoubtedly served as a storage room. The remaining wallsdid not exhibit quite as good constructural work as most of those inthe structure, but they were of the same general type. The floor wasadobe plaster.The court at the southeast corner of the dwelling (fig, 60) wasenclosed by a low masonry wall that was erected at the same time asthose forming room 6. Judging from the amount of fallen materialand the remaining portions of the wall, the total height had neverbeen more than approximately 3 feet (91.44 cm). Due to the slopeof the original surface at this point, it was necessary for the buildersto do some terracing. The southeastern third of the enclosure wasbuilt up with dirt and rocks and the upper surface paved with stoneslabs. The wall at this end originally extended to the kiva and bothit and the paving seem to have been carried across the kiva roof tothe hatchway. This was indicated by the presence of wall materialand large slabs well up in the deposits in the pit for that structure.A low wall, connecting room 6 and the kiva, also served to terrace thatside of the court and hold back the earth placed there to level thatpart of the area. There was nothing to suggest that the wall had everrisen above the floor level along that side. The entire enclosed sur-face, including the paved section, was covered with hard-packed adobeflooring. There were no indications that fires had been lighted any-where in the court. Not far from the southwest corner a stone slab
232 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 121was set on end close to the enclosing wall. In all respects this stonewas similar to the steps described for the other units and may havebeen placed there to aid people in stepping over the wall and to obviatethe necessity for a person going all the way around the court to reachthe outside. A similar stone was found on the ground outside thewall opposite this one. It may represent a similar step at that point.As it was not in position and there were no clear traces of its place-ment, such a function can only be suggested. This court in its broaderaspects of location and general features is analogous to the area lyingbetween room 11, the wall extending from the granary, and the kivain the second unit. There is also a similarity between it and theshelter associated with pit structure 15 and its adjacent granaries.The present example probably illustrates the developed form of thefeature adumbrated by the others.The storage pit in the area between the kiva and the old pit struc-ture unquestionably belonged to the unit house. It was dug throughlayers that accumulated subsequent to the abandonment of the pitstructure and passed through the stratum of red earth previouslydescribed. The shape of the pit was not unlike that of a large pot-tery jug. The opening was small, and there was a short, necklikeshaft with vertical sides. From the bottom of these the walls slopedoutward to the point of maximum diameter and then curved in to-ward the bottom. The latter had a somewhat shorter diameter thanthe area of greatest circumference (fig. 52, bottom section). Theinside of the pit was plastered when it was in use. Only scatteredpatches remained on the walls when the debris was cleaned from theinterior, but these were sufficient to show that the entire surface wasoriginally so treated.The outdoor fire pit (fig. 50) was a rectangular box lined with stoneslabs. There were no traces of any shelter or arbor near it. It prob-ably was like those scattered over the refuse mound for the secondunit, merely a place where fires could be Irindled when no shade orprotection was needed. Some cooking was done there, however, be-cause the ashes and charcoal that filled the pit had some splinters andfragments of charcoal and bone scattered through them.The kiva, located in the angle formed by the surface building andthe court (pi. 27, 5), more nearly approximated the circular formthan any of the semisubterranean structures hitherto described. Boththe back wall of the bench and the main wall of the chamber weremore regular. The bench was partially dug and partially built up.When the pit for the main part of the chamber was dug, a narrowencirclmg shelf was left at the level for the bench top. Then whenthe walls of the pit were lined with stone the masonry was carried tothat point and the width increased in proportion to the thickness ofthe wall. The stonework was very good ; in fact it was the best wall
BUREAU OF AMERICAN ETHNOLOGY BULLETIN 121 PLATE 27
a. Surface structure in thiril unit.
b. ClfaniiL .U
BUREAU OF AMERICAN ETHNOLOGY BULLETIN 121 PLATE 28
n. Wall rnnstmotion in bench of third unit kiva.
b. Holes for wtiinscoting poles at back of kiva bench.
BUREAU OF AMERICAN ETHNOLOGY BULLETIN 121 PLATE 29'
a. Floor in kiva for the third unit.
6. Mound i-KM 1 in. 1 irge building in Great t'ueblo iiroup.
BUREAU OF AMERICAN ETHNOLOGY BULLETIN 121 PLATE 30
n. Portion of Great Pueblo ruin?, corner of buildincr below and to left of standinsr figure.
6. Arroyo cut througli refu.se mound.
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA 233construction found in the course of the excavations. Medium-sizedtabular blocks were used for the main courses, except for a few largerstones that occurred near the top of the wall, and many small frag-ments were employed as chinking (pi. 28, a). A thick coating ofadobe plaster gave a smooth-finished surface to the wall.One interesting structural feature in the kiva wall consisted offour timbers incorporated in the masonry at approximately the posi-tions that support posts for the truncated pyramidal type of super-structure occupied in some of the pit dwellings (fig. 51, a). The endsof these timbers were flush with the bench top. When the latter wascovered with plaster they did not show. They may have been placedthere solely to reinforce the wall. On the other hand they may havehad a certain ceremonial significance. It will be recalled that inhouse 15 three of the support posts were incorporated in the face ofthe bench. In some sections the use of actual posts for supportingkiva roofs continued into later horizons, as shown by reports of earlySpanish explorers '^ and by archeological work in various ruins.^? Apossible variant of the survival factor is that of the inclusion ofshort beams or posts in the masonry boxes that formed the pilastersin the kivas in the Chaco Canyon ruins, a feature discussed in con-nection with the plaster ridges on the bench in structure 12 at thissite. Had the builders intended the posts incorporated in the benchin this kiva to serve wholly as reinforcements for the stonework, itis odd that they restricted the number to four and that they wereplaced in the positions that actual supports would occupy. The wallproper did not have a vertical face. The diameter on the floor levelwas slightly greater than that at the bench level, so that the wallcurved inward to a degree. This curve started a short distance abovethe floor. That it was intentional may be questioned, because simi-lar construction in kivas is extremely rare. In this particular caseit seems more likely that it was fortuitous.The upper wall at the back of the bench had been finished with awainscoting of poles and plaster. Nothing remained of the con-struction except the butts of the poles, the holes where they were set(pi. 28, 5), and some traces of the plaster along the bench top. Thiswas sufficient, however, to show that there had been such a feature.The poles averaged 2 inches (5.08 cm) in diameter and were set verti-cally at more or less regular intervals around the periphery. Theymay have been placed there to serve as a support for the walls ofthe excavation. The nature of the earth at that location is suchthat it will hold a firm face and for that reason it does not seem thatany reinforcement would be necessary, particularly when the fill ofdirt over the roof timbers would cover that portion of the excavation.
'?Winship, 1896, p. 620.*? Cummlngs, 1915, p. 275. Jeancon, 1929, pp. 15-16.
234 I5UREAU OF AMERICAN ETHNOLOGY [Bull. 121
5 FEET
Figure 51.?Kiva for unit No. 3. a, posts incorporated in bench to reinforce wall ; h, holesfoi- poles forming wainscoting ; c, Katcina niche ; d, sipapu ; e, fire pit ; /, ash pit andladder box; g, deflector; h, shallow basin in floor; i, small circular basin with stonebottom ; /, stone embedded in floor ; k, ventilator opening ; I, ventilator shaft ; m, presentsurface ; n, old ground level.
ROBBETS] ARCHEOLOGICAL REMAINS IN ARIZONA 235Hence they may have been put there with some other purpose in view.The wall of the excavation may not have been as regular as the build-ers deemed necessary, and m order to obtain a smoother contour theyerected the jacal facing. On the other hand, a wainscoting of thisform possibly had some function other than a purely utilitarian one.Some of the kivas of the Great Pueblo period in other districts werecharacterized by a similar structural feature and in them there covddbe no question of a need for reinforcing the back walls because theywere built of stones and mortar and in no way likely to crumble orslump. This led to the suggestion that since the kivas were the out-growth of the old semisubterranean pit dwellings and many featuresfrom the latter survived in the ceremonial chambers, the wainscotingmay have represented the sloping side timbers of the original type ofroof. On the other hand, they could be a direct continuance of a pit-house practice of lining the upper wall for purely utilitarian pur-poses, although occurrences of the latter have not been noted withsufficient regularity to indicate that it was of any great significancein the evolution of the house type. Future work and more informa-tion on both house and kiva forms may solve this problem. Allthat can be done at present is to suggest possible explanations for thefeature. :The roof over the kiva was of the cribbed type, starting from thetop of the bench. The bottom tier of logs rested directly on thebench in the manner described for the kiva of the second unit. Therewas direct evidence of this in the charred remains of the timberslying along the bench. The superstructure was destroyed by fire,but enough of the charred wood remained to give an excellent ideaof the roof type. The charcoal fragments also supplied a good seriesof specimens for dendrochronological studies.Interior features consisted of a Katcina niche, a sipapu, fire pit,ash and ladder box, a deflector, a shallow bin formed by a plasterridge on the floor, and a slab of stone embedded in the floor at oneside of the chamber (pi. 29, a). The Katcina niche (fig. 51, c) was inthe wall directly opposite from the ventilator. It was a rectangularstone box incorporated in the masonry and quite comparable to theone in the kiva of the second unit. The sipapu was an oval-shapedhole (fig. 51, d) with vertical sides and a flat bottom. The insidewas well finished with a carefully smoothed coating of adobe plaster.The fire pit was rectangular in form (fig. 51, e) and the four sideswere faced with stone slabs, one to each side. The stones did notmeet at three of the corners and the interstices were filled with mudplaster that produced slightly rounded corners. The ash pit-ladderbox was a shallow basin faced with plaster (fig. 51, /). The onlystone in it was the deflector which formed one end. The ends of theladder rested against the deflector stone and the ashes were placed
236 BUREAU OF AMERICAN ETHNOLOGY [Bdll. 121at the opposite end near the fire pit. The deflector slab (fig. 51, g)was either broken when the roof collapsed or cracked by the heatfrom the burning timbers and the top knocked off by falling debris.The lower end was still embedded in the plaster when the pit wascleared, but the fragments from the upper portion were scattered overthe floor in its vicinity. These showed that the original height ofthe stone was half again as great as the standing remnant. In itsintact form it would have functioned properly in preventing thedraft coming through the vent opening from blowing directly onthe fire. The flat stone in the floor (fig, 51, j) occupied a positionsimilar to those described for the two floor levels in the kiva of thesecond unit. In the present case, however, the top was not flush withthe floor but projected above it. The top of the stone was smoothedfrom rubbing. There was nothing to indicate whether this had beendone before it was placed there or resulted from the use to whichit was put after installation.The plaster-rimmed enclosure (fig. 51, h) at the eastern side ofthe chamber was the only one of this type noted in the Allantowndistrict. From a broad aspect it was somewhat comparable to thestorage basins described for some of the pit structures, but in itsmore specific details it was unique. As the photograph of the floorof the chamber (pi. 29, a) and the plan of the floor (fig, 51) show,the feature was along the base of the wall and open at one end.The main part of the enclosure was rectangular in shape, while theclosed end was characterized by a circular basin (fig. 51, i). Thefloor of this part of the enclosure was slightly lower than that inthe rectangular portion and was paved with a stone slab. The outsideof the adobe ridge curved down to the floor of the room and theinside along the straight portion and part of the curved end waspractically vertical. The remaining section curved concavely tothe bottom, forming the basin at the end. Near the juncture of thestraight and curved ridges a small pole was set, rising verticallyabove the plaster. The upper part of this stick burned when thehouse succumbed to the flames and its original height could not belearned. The charred end, still in position, showed that it had ex-tended above the rim. The purpose of this stick is not known andthe junction of the enclosure proper can only be postulated. A metatewas present on the floor near the open end and it is possible thatit was a form of bin for use in grinding meal for ceremonial pur-poses. The circular portion at the end would have made a goodrest for either a basket or large pottery jar to contain the meal.Firewood could have been stacked there and a water jar, ratherthan one for other purposes, could have stood in the basin. Orthe feature may have had some connection with or functioned aspart of the equipment needed in a ceremony. In this connection
ROBERTS] ARCHBOLOGICAL REMAINS IN ARIZONA 237attention may be called to the fact that a large slab of stone, dressedaround the edges, was lying on the floor between the circular basinand the ventilator. The size and shape of this slab were such thatit would have neatly covered the enclosure and ridge if placed there.This stone is shown leaning against the wall in the photograph(pi. 29, a). Whether it served that purpose or was provided as amovable cover for the ventilator opening is not known. It wouldhave answered either purpose, yet may not have been so employed.Instead it may have been a cover for the hatchway and have falleninto the position found at the time of the destruction of the roof.The ventilator for the structure was of the tunneled type. The pas-sage apparently was cut through from the bottom of the shaft. Theinterior was lined with stone. From the sill of the opening to thebottom of the shaft the passage floor was practically level. Thebottom of the shaft sloped sharply upward from the passage. Theceiling of the latter had a pronounced upward slant and the openinginto the shaft was much higher than the aperture in the wall of thechamber. The opening into the room had a stone sill and stonelintel. The sides were part of the masonry of the main wall con-struction. Some of the latter had fallen from one lower corner ofthe aperture, giving it a slightly irregular shape, as shown by thepicture. Originally that side was vertical like the one opposite.The heat of the conflagration cracked the lintel across the middle andit had dropped out of position. Only a few of the stones in the wallabove were dislodged by the occurrence; the remainder were stillsolidly in place. Sometime after the completion of the ventilatorthe opening was reduced by putting narrow plaster walls at eachside. They were set back several inches from the edge of the orig-inal opening. The aperture which they formed was roughly ovalrather than rectangular in form and only about half as large as theformer one. The ventilator in all probability functioned too wellwhen first built and as a consequence the occupants found it neces-sary to reduce the size. Wlien the chamber was cleared of its ac-cumulated debris a stone covered with green paint and a small paintmortar stained with red pigment were found on the shelf formedby the vent sill and the inset plaster walls.The ventilator shaft was oval in contour and its sides approxi-mately vertical. The bottom sloped sharply from the outside edgeto the opening into the passage. The shaft was lined with hori-zontally laid masonry from top to bottom. There were no uprightslabs in the foundation as noted in the other structures. The open-ing into the passage was topped by a thick stone lintel that sup-ported the construction on that side of the shaft. The stoneworkwas not as well done as that in the main part of the structure. Largerblocks of stone were used and there was not as much chinking. Be-
238 BUREAU OF AMERICAN ETHNOLOGY [Bull. 121cause of the small space in which to work, the builders must have ex-perienced some difficulty in laying the wall. It served the purpose,however, as well as one with more finished construction.The fill in the kiva indicated that it had been occupied up to thetime that it was destroyed by fire. The charred timbers of thesuperstructure were lying directly on the floor and the other roofmaterial was on top of them. On the southwestern side of thechamber the roofing material was surmounted by large slabs and
15 FT.
court
Figure 52.?Sections in third unit. Heavy line passing through fill indicates layer ofred earth separating occupation level of pit structure and that of subsequent unit.building stones, debris from the paved and terraced end of the court.The remaining portion of the pit contained a mixture of sand, ashes,charcoal, and lumps of plaster?all drift material that was depositedin the pit by natural forces. The top 2 feet 2 inches (66.04 cm)above the kiva was clean, wind-blown sand. The character of thefill clearly indicated that the site was abandoned after the burningof the kiva. That catastrophe probably was an integral factor inthe movement. It could have been the cause of the departure or
RoBimTs] ARCHEOLOGICAL REMAINS IN ARIZONA 239may have been burned intentionally as the people were leaving inorder to prevent others from making use of it.Charcoal from the site gives an interesting series of dendro-chronological dates. Those from the house range between 994 and1004. A number of rings apparently are missing from the outsideof each and with the estimated bark date taken into considerationthe maximum in the range would give 1005 to 1014, a spread of nineyears. The kiva timbers give 1000 to 1011 and with the correctionfor estimated missing rings would be 1007 to lOlS.^^ On the basisof either set of figures there is good evidence that the house waserected first, or at least a part of it, and then the kiva was built.The spread of years in this unit is much less than that in some ofthe other dated structures and indicates less reused timbers andpossibly a shorter period of occupation. The latter can be estimatedin this instance, however, on the basis of group size and totalburials. The evidence from the latter is not as clear-cut as couldbe desired but with certain allowances for two factors, the olderpit structure and some previous digging in the refuse mound, isprobably fairly accurate.A total of 20 burials was found in and below the refuse mound.Of this number, four were definitely of the pit-house phase. This wasshown by their location in clean sand below the mound and thenature of the accompanying mortuary offerings. The previous dig-ging in the mound had uncovered four, probably five, burials thatwere in the upper levels and unquestionably were late, as their pitshad not penetrated to the bottom of the deposits. The bones hadbeen left in place but the funerary furniture removed. The ques-tion concerning the exact number was raised by the possibility thatone had been a double burial, that of an adult and an infant. Thelatter had reached such an advanced stage of disintegration thatonly a few fragments of bone were present and in the disturbedstate of the material might be questioned as acceptable evidence.Omitting the 4 pit-house burials and the doubtful later one fromconsideration and granting a removal of 4 from the subsequent inter-ments, leaves the number at 20. Of this total 8 were adults and 12were children. On the basis of 6 to a family and 2 families to theunit, the average population would be 12. Using the death ratefigures suggested by Dr. Hooton in his studies of the Pecos skeletalmaterial ^- of approximately 3 per 100 each year, the rate here wouldbe one death about every 28 years. With a total of 20 burials, thespan would be 56 years. This seems longer than is warranted bythe general aspects of the site and it may be the figures are too low.
81 Miller, 1934, p. 16.82 Hooton, 1930, p. 333.
240 BUREAU OP AMERICAN ETHNOLOGY [Bull. 121It is possible the average population was greater and the death ratehigher, and that a number of burials were missed, although the latterwould tend to increase the number of years. Probably elements of allthree were involved. The inhabitants of the unit may well have num-bered 15 over most of the period of occupancy and the death ratehave been on the basis of from 4 to 5 per 100 or one death in thepresent group approximately every year and a half. On this basisthe life span of the unit would be about 30 years, w^hich would cor-relate fairly well with other aspects. The answer nearest to thecorrect one is probably two generations.Room 1 was 8 feet 2 inches (2.489 m) long and 5 feet 8 inches (1.727m) wide. The bin at the northeast corner of the enclosure measured3 feet 2 inches (96.52 cm) by 1 foot 6 inches (45.72 cm) and was 41/2inches (11.43 cm) high. The western wall was 1 foot 1 inch (33.02cm) tliick, the northern 1 foot 2 inches (35.56 cm), the eastern variedfrom 6 inches (15.24 cm) at the southern end to 8 inches (20.32 cm)at the northern ; and the southern wall was 7 and 8 inches ( 17.78 and20.32 cm) tliick at the western and eastern ends.Room 2 was 10 feet 10 inches (3.302 m) and 10 feet 6 inches (3.200m) long on the west and east walls, respectively. The width was 4feet 10 inches (1.473 m) at the northern end and 4 feet 4 inches (1.320m) at the southern end. The wall at the northern end averaged 1foot 2 inches (35.56 cm) in thickness, the eastern 1 foot 4 inches (40.64cm), the southern ranged from 10 inches (25.4 cm) to 1 foot 2 inches(35.56 cm), and the western from 1 foot 2 inches (35.56 cm) to 1 foot4 inches (40.64 cm). The oval storage pit in the floor of the roomwas 2 feet (60.96 cm) long and had a maximum width of 1 foot 1inch (33.02 cm). It was 6 inches (15.24 cm) deep. The eastern endwas only 1 inch (2.54 cm) from the base of the wall.Room 3 was 12 feet (3.657 m) long on the eastern wall and 11 feet10 inches (3.606 m) on the western one. At the northern end it was6 feet 6 inches (1.981 m) wide and at the southern 6 feet 10 inches(2.082 m). The northern wall was 7 inches (17.78 cm) thick at thewestern end, 10 inches (25.4 cm) at the eastern. The wall at the eastside of the room was 1 foot (30.48 cm) thick at the northeastern cor-ner and 10 inches (25.4 cm) at the southeastern. The wall at thesouthern end of the room did not vary more than a fraction from 1foot 2 inches (35.56 cm) throughout its length. The wall at the westside was in common with room 2, and was 1 foot 4 inches (40.64 cm)average thickness. The slab-lined fire pit measured 2 feet (60.96cm) by 1 foot 11 inches (58.42 cm). Its depth was 6 inches (15.24cm). The small ash pit adjoining it on the north side measured 1foot 5 inches (43.18 cm) by 1 foot 2 inches (35.56 cm) and had a depthof 4 inches (10.16 cm). The original floor was 3 inches (7.62 cm)below the later one. The fire pit for the original floor was in practi-
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA 241cally the same location as the later example and the latter partiallycut into it. As far as could be determined the first pit measured 1foot (30.48 cm) by 1 foot 2 inches (35.56 cm) and was 4 inches (10.16cm) deep. The remains of the shelter were 2 feet (60.96 cm) belowthe floor of room 3 and were approximately 7 feet 9 inches (2.362 m)long by 5 feet 6 inches (1.676 m) wide.Room 4 was 10 feet 2 inches (3.098 m) long on the west side and10 feet 11 inches (3.327 m) on the east. The width at the north was4 feet 4 inches (1.320 m) and at the southern end 3 feet 10 inches(1.168 m). The northern wall was 1 foot 2 inches (35.56 cm) wideat the northwest corner and 10 inches (25.4 cm) at the northeast.The wall at the east side was 1 foot 2 inches (35.56 cm) thick atthe northeast corner and 1 foot 5 inches (43.18 cm) at the south-east. The southern wall was 1 foot 2 inches (35.56 cm) thick atboth the southeast and southwest corners. The wall along the westside of the room ranged from 1 foot 2 inches (35.56 cm) in thick-ness at the southwest comer to 1 foot 6 inches (45.72 cm) at thenorthwest. The oval-shaped depression in the floor along the baseof the west wall was 2 feet 5 inches (73.66 cm) long. Its greatestwidth was 1 foot (30.48 cm). The depth was only 2 inches(5.08 cm).Room 5 was 10 feet 4 inches (3.149 m) long on the west side and10 feet 9 inches (3.276 m) long on the east. The width at the northend of the room was 6 feet 9 inches (2.057 m) and at the south 6feet 5 inches (1.955 m). The wall at the west side was 1 foot 2inches (35.56 cm) wide at the northwest corner and 1 foot 5 inches(43.18 cm) at the southwest. The north wall was very regular; thewall was the same as that for the southern end of room 3, and was1 foot 2 inches (35.56 cm) thick at both the northwest and north-east corners. The wall at the east side of the room was 9 inches(22.86 cm) thick at the northeast corner and 1 foot (30.48 cm) atthe southeast. The southern wall was 1 foot 4 inches (40.64 cm)thick at the southeast corner and 1 foot 2 inches (35.56 cm) thickat the southw^est corner. The fire pit measured 2 feet 1 inch (63.50cm) by 1 foot 11 inches (58.42 cm) and had an average depth of 8inches (20.32 cm). The ash pit at the north side was 1 foot 10inches (55.88 cm) long, 1 foot 2 inches (35.56 cm) wide, and 4inches (10.16 cm) deep. The storage bin along the wall had amaximum length of 4 feet 6 inches (1.371 m) and a maximumbreadth of 1 foot 6 inches (45.72 cm). Indications were that thewall had had an average height of 1 foot 7 inches (48.26 cm) whenthe stone slabs were topped by the adobe plaster that supplementedthe stones.Room 6 measured 10 feet 5 inches (3.175 m) along its east walland 8 feet 11 inches (2.717 m) along the west. The north end of
242 BUREAU OF AMERICAN ETHNOLOGY [Boll. 121the room was 6 feet 2 inches (1.879 m) wide and the south 5 feet10 inches (1.778 m). The north wall was 1 foot 2 inches (35.56 cm)thick at the northwest corner and 1 foot 3 inches (38.1 cm) thick atthe northeast corner. The wall at the east side of the room meas-ured 1 foot 3 inches (38.1 cm) at the northeast corner and 1 foot 2inches (35.56 cm) at the southeast. The wall at the south was 1foot (30.48 cm) thick at the southeast corner and 10 inches (25.4 cm)thick at the southwest. The west wall was 1 foot 1 inch (33.02 cm)thick at the southwest corner and 1 foot 8 inches (50.8 cm) at thenorthwest corner.The court had an inside measurement of 17 feet 4 inches (5.283 m)along the southern wall. The remaining portion of the wall at theeast side was 9 feet 4 inches (2.844 m) long. The wall at the southside was 11 inches (27.94 cm) thick at the west end and 6 inches(15.24 cm) thick at the east end. The east wall measured 1 foot(30.48 cm) at its southern end and 11 inches (27.94 cm) at the north.The paved area was 4 feet 6 inches (1.371 m) wide at its southernend and 6 feet 5 inches (1.955 m) at the north side. Its length wasthe same as that of the court wall, 9 feet 4 inches (2.844 m). Thewall joining room 6 and the kiva and forming a terrace at that sideof the court was 10 feet 10 inches (3.302 m) long. At the end abut-ting room 6 it was 6 inches (15.24 cm) thick and at the kiva was 5inches (12.7 cm) thick. The upright stone, possibly a step, at thebase of the wall on the southern side of the court was 1 foot 1 inch(33.02 cm) long, 6 inches (15.24 cm) high, and % "^ch (1.90 cm)thick. It stood 2 inches (5.08 cm) from the wall.The storage pit was 13 feet 10 inches (4.216 m) from the wall ofthe surface dwelling and 6 feet 3 inches (1.905 m) from the kiva.The opening at the top was 1 foot 6 inches (45.72 cm) in diameter.The bottom was 2 feet (60.96 cm) in diameter and at a point 1 foot4 inches (40.64 cm) above the bottom had a diameter of 2 feet 4inches (71.12 cm). The bottom of the pit was 5 feet 4 inches (1.625m) below the present surface and 3 feet 2 inches (96.52 cm) belowthe old ground level.The outdoor fire pit was 18 feet 2 inches (5.537 m) from the north-east corner of the house and 26 feet 6 inches (8.077 m) from the kiva.It was 18 feet 6 inches (5.638 m) from the storage pit. The recti-linear fire box was 1 foot 5 inches (43.18 cm) by 1 foot 6 inches(45.72 cm) and 6 inches (15.24 cm) deep.The kiva had a diameter of 16 feet 6 inches (5.029 m) above thebench on the sipapu, fire pit, ventilator line. On this same levelfrom bench edge to bench edge the diameter was 12 feet 3 inches(3.733 m). On the floor level it was 12 feet 9 inches (3.886 m). Atright angles to this line the diameter above the bench was 16 feet 6inches (5.029 m), and from bench edge to bench edge was 12 feet 7
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA 243inches (3.835 m) . On the floor level it was 12 feet 10 inches (3.911m). At the ventilator side of the room the floor was 10 feet 8y2inches (3.263 m) below the present surface and 7 feet 7 inches (2.311m) below the old ground level. The bench was 3 feet 7 inches (1.092m) high at this point. At the opposite side of the chamber the floorwas 8 feet 9 inches (2.667 m) below the present surface and 6 feet 1inch (1.854 m) below the old level. The bench was 3 feet 10 inches(1.168 m) high. At the southwest side of the room the floor was 10feet 8 inches (3.251 m) below^ the present surface, 8 feet 11 inches(2.717 m) below the floor of the court, and 6 feet 6 inches (1.981 m)below the original ground level, that underlying the fill for the ter-race. The bench was 3 feet 9 inches (1.143 m) high at this point.At the opposite side of the room the floor was 8 feet 11 inches (2.717 m)below the present surface and 5 feet 1 inch (1.549 m) below tlie orig-inal ground level. The bench was 3 feet 8 inches (1.117 m) high.The timbers incorporated in the bench wall in the approximal posi-tion of the old roof supports had diameters of 3I/2 inches (8.89 cm),51/^ inches (13.97 cm), 6 inches (15,24 cm), and 6 inches (15.24 cm).The posts that formed the wainscoting around the back of the benchhad diameters ranging from 2 to 3 inches (5.08 to 7.62 cm).The Katcina niche (fig. 51, c) was 2 feet 5 inches (73.66 cm) abovethe floor and 1 foot 1 inch (33.02 cm) below the top of the bench.The niche was 7^/2 inches (19.05 cm) long at the bottom and 8^4inches (20.95 cm) long at the top. The opening in the wall measured5y2 inches (13.97 cm) across the lintel and 6 inches (15.24 cm) acrossthe sill. The north side was 4^ inches (11.43 cm) high and thesouth side 4 inches (10.16 cm).The sipapu was 3 feet 7i/^ inches (1.104 m) from the wall. Thehole had diameters of 4 inches (10.16 cm) and 5^/^ inches (13.97 cm)and was 9 inches (22.86 cm) deep. From the edge of the sipapu tothe edge of the fire pit was 3 feet 1 inch (93.98 cm). The fire pit(fig. 51, e) measured 1 foot 6 inches (45.72 cm) by 1 foot 4 inches(40,64 cm). The depth varied from 4 inches (10.16 cm) below thefloor level at the sipapu side to 7 inches (17.78 cm) at the ventilatorend. The slab at the shallow end projected 11/2 inches (3.81 cm)above the floor, making the actual depth 5i/^ inches (13.97 cm). Theash pit was 1 foot 71/2 inches (49.53 cm) long, 111/2 inches (29.21 cm)wide at the end next the fire pit, and 1 foot 21/2 inches (36.83 cm)wide at the base of the deflector stone. The depth was 4 inches(10.16 cm). The deflector stone (fig. 51, g) was 1 foot 7 inches(48.26 cm) long, 1 inch (2.54 cm) thick, and in its broken conditionstood 1 foot (30.48 cm) above the floor. Originally it had a heightof 1 foot 6 inches (45.72 cm). The stone embedded in the floor(fig, 51, j) measured 8 inches (20.32 cm), 9 inches (22.86 cm), 7inches (17.78 cm), and 71/2 inches (19.05 cm) on its four sides. The
244 BUREAU OF AMERICAN ETHNOLOGY [Bull. 121top of the stone was I14 inches (3.81 cm) above the floor. The stonewas 1 foot 4 inches (40.64 cm) from the ash pit.The enclosure formed by the adobe ridge (fig. 51, h) was 1 foot10 inches (55.88 cm) long and 1 foot 3 inches (38.1 cm) wide in itsrectangular portion. The circular part at one end had diameters of111/^ inches (29.21 cm) and 11 inches (27.94 cm). The stone in thefloor measured 6 inches (15.24 cm) by 51^ inches (13.97 cm). Theadobe ridge ranged from 5 inches (12.7 cm) to 6 inches (15.24 cm) inwidth along the straight portion. Wliere the small pole was erectedit was only 3 inches (7.62 cm) wide and the curved end did not varygreatly from a 3-inch (7.62-cm) average. The height varied onlya fraction at different points along the ridge from a 2%-inch(6.98-cm) average. The pole set in the ridge at the juncture of thecurved and straight portions had a diameter of 1 inch (2.54 cm).Its original height could not be determined. The large stone slabfound on the floor at one end of this enclosure was 1 foot 31/2 inches(39.37 cm) wide at one end and 1 foot 1% inches (34.92 cm) wideat the other. One side was 2 feet 5 inches (73.66 cm) in length,while the other was 2 feet 6I/2 inches (77.47 cm) long. The stoneaveraged 2% inches (6.98 cm) in thickness.The original ventilator opening was 1 foot 2 inches (35.56 cm)wide at the sill and 1 foot (30.48 cm) wide at the lintel. The eastside was 1 foot li/o inches (34.29 cm) high and the west 1 foot I/2inch (31.75 cm) high. The sill was 8 inches (20.32 cm) above thefloor of the room. After the two plaster walls w^ere placed in theoriginal aperture the size of the opening was reduced to 4I/2 inches(11.43 cm) at the bottom, 71^ inches (19.05 cm) across the center, and41/^ inches (11.43 cm) at the top. The passage was 3 feet (91.44 cm)long. Where it entered the shaft it was 1 foot (30.48 cm) wide and2 feet (60.96 cm) high. The roof sloped upward from the aperturein the wall of the chamber to the point where it entered the shaft.The inside of the shaft had diameters of 1 foot (30.48 cm) and 1foot 5 inches (43.18 cm). The bottom of the shaft was 9 feet 614inches (2.907 m) below the present ground level and 3 feet I14inches (95.25 cm) below the original surface.Great Pueblo EuinsOutstanding among the house remains in the Whitewater districtare those of two structures belonging to the Great Pueblo period.They are located in the central portion of the main site on top ofthe ridge south of Allantown (pi. 29, h). The largest concentrationof people in one community in this district probably occurred in thesebuildings. No excavations were attempted in these ruins, hence it isnot possible to give any extended description of them. A slight amount
ROBKBT8] ARCHEOLOGICAL REMAINS IN ARIZONA . 245of debris was moved to locate the walls at several corners and at afew places along the outside walls, but beyond this no digging wasdone (pi. 30, a). By locating the ends of walls and such corners aswere visible, a plan was drawn showing the general outlines of thestructures and those features apparent on the surface (fig. 53). Thelarger of the two main buildings conformed to a rectangular-blockplan. The smaller was made up of two units joining so as to forman obtuse angle and provide a court in which a large kiva or cere-monial chamber was located. Near one end of the smaller structurewas a great kiva that was connected to it by a wall extending fromthe corner to the curved wall of the circular chamber. This wallserved to accentuate, or perhaps better to complete, the court feature.Some distance from the two large mounds were the remains of twosmall rectangular structures. They stood in front of the main build-ings and seem to have been either single-roomed houses or the lowerparts of towers. This could not be determined without excavation.Their location and relation to the other remains is somewhat unusual.The ground plan of the group as a whole has a regularity and sym-metry that suggests the builders were following some preconceivedidea and not just erecting houses. The refuse mounds for the com-munity were located to the east of the buildings. One of them hadbeen cut by drainage water from the urea surrounding the ruins andas a result was traversed by two sizable arroyos. These exposed thestratification and showed the depth of the accumulation (pi. 30, &).The smaller of the two buildings, that at the northern end, wasinteresting because it gave indications of more than one stage ofgrowth. The original portion was a large rectangular block pre-sumably consisting of a double tier of rooms. As far as could bedetermined from surface evidence the rooms in this structure wereall rectangular and numbered 9 or 10 to the row. The second rowat the back of the building, the side away from the kiva, was atleast two stories in height. Parts of the second-floor walls were stillstanding. If the second floor had a similar number of rooms theoriginal unit probably contained 29 to 30, possibly even more, secularchambers. The presence of a small kiva in the block of the buildingor a variation in room sizes would alter the number, but from allappearances it seems that 30 would be a fair estimate. After thecompletion of this structure a smaller one, pentagonal in plan, wasadded at the southwestern end. The new wing contained a kiva andat least five rooms. There were other enclosures indicated. Theywere formed by the spaces between the curved walls of the kiva andthe surrounding straight ones. Judging by the nature of these odd-shaped cells, as shown by excavations elsewhere, it is not likely that68764?39 17
246 BUREAU OF AMERICAN ETHNOLOGY [BOLU 121they were of any significance here. This part of the building appar-ently was only one story in height.The original kiva probably was the one located in front of thebuilding. It was rather large compared with the general average
?__scAie rser -?.jji.-<^iir'<?fFiQDEE 53.?Plan of large ruins of Great Pueblo period.for such Structures. There is a possibility that it is a smaller exam-ple of the great kiva type, but it is not too large to fall within therange of the ordinary ceremonial chamber. Of course, excavationsmight reveal various internal features usually associated with greatkivas and then there would be no question. The diameter inside the
RoBEiHTs] AROHEOLOGICAL REMAINS IN ARIZONA 247upper wall, above the bench, was 38 feet 6 inches (ILTai m) andinside the bench 33 feet 6 inches (10.210 m). The kiva in the southwing had an upper diameter of 29 feet (8.839 m) and a lower oneof 25 feet (7.620 m). The main part of the rectangular structurewas 95 feet (28.956 m) long and 34 feet 6 inches (10.515 m) wide.The wing measured 41 feet 6 inches (12.649 m) along the northwestwall, 26 feet 6 inches (8.077 m) along the southwest wall, 54 feet 6inches (16.611 m) on the southeast wall, and 32 feet (9.754 m) onthe northeast wall.There was nothing to indicate the exact relationship between thegi-eat kiva at the northwestern corner of the structure and the build-ing proper. The wall connecting the two was erected aft?r the com-pletion of both structures. The circular chamber was of impressivesize even for great kivas. The diameter between the walls abovethe bench was 62 feet 10 inches (19.152 m), while inside the bench itwas 57 feet (17.374 m). The structure compares favorably withthe great kiva in Pueblo Bonito with its 60-foot (18.288-m) diameterand with the so-called Great Bowl in Chettro Kettle, also in theChaco Canyon, which has a reported diameter of 62 feet 6 inches(19.050 m).*^ It is not as large as the ^reat kiva in Casa Rinconadain the Chaco Canyon, located across the canyon from ChettroKettle and Bonito, with its 72 feet (21.946 m), nor the one at theVillage of the Great Kivas on the Zuni Reservation that measured 78feet (23.744 m).?* On the other hand, it surpasses the second ex-ample at the same location that has a diameter of 51 feet (15.545 m)above the bench and 47 feet 6 inches (14.478 m) below it.??Also one in the Aztec Ruin in northern New Mexico measuring 48 feet314 inches (14.719 m),^' and that at the Lowry Ruin in southwesternColorado with a 47-foot (14.326-m) diameter.*" How it comparedwith these remains in the matter of interior features can be deter-mined only by excavation. There were slight suggestions, however,of the presence of subfloor vaults in the central portion of the floorwhere there was only a small amount of fallen wall material andother debris.The purpose of these superceremonial chambers is still to belearned. They unquestionably are a heritage from Modified BasketMaker times. An example was found in a village belonging to thatperiod in the Chaco Canyon,?? and the La Plata district farther northin southern Colorado has several dating from that horizon. In thediscussion of the one in the Chaco Canyon it was suggested that thelarffe communal rituals, those in which the whole village was con-WHewett, 1922, p. 122.?? Roberts, 1932, p. 96.?? Roberts, 1932, p. 95.?? Morris, 1921, p. 115.?T Martin, 1936, p. 46.w Roberts. 1929, pp. 73-81.
248 BUREAU OF AMERICAN ETHNOLOGY ^^vvu. 121cerned, were performed in these great kivas while ordinary cere-monies restricted to special groups took place in the smaller circularchambers. Morris in discussing the structures in his report on theexample at Aztec postulated that they were sanctuaries for the entirepueblo where the most sacred rites known to the people of that periodwere performed by members of the various priesthoods.^^ On theother hand, it is possible that dances of the kind now held in enclosedcourts or patios in the modern villages were originally restricted tothe great kivas. The significance in the presence of one of thesestructures at this site lies in the fact that they generally occur inassociation with some form of the Chaco cultural pattern, as wellas at places where the latter had a definite influence on develop-ments. At the present time there is a question concerning the originand growth of the superkiva. Some think that its elaboration tookplace in the Chaco center and that it spread outward from there.Others believe that it attained its greatest development in the south-ern Colorado region and spread from there to the Chaco and thenceon toward the south and the west. Present evidence tends to supportthe Chaco adherents, although it is too early to attempt definite con-clusions. There is no question, however, but what it is a northernfeature and that it is one of the traits usually found in the Chacopattern complex. As suggested in the discussion of the dance courtin earlier pages in this report, there likely was some similarity inpurpose between it and the great kivas. It is possible that the dancecourt here was merely a temporary arrangement to answer ceremonialneeds while the great kiva was under construction.The second building, the larger of the two, consisting of a rec-tangular block with a series of kivas along the front, although in-corporated in the structure, gave outward indications of having beenbuilt from a definite plan. That it was erected as a unit could notbe established without excavation and while all exterior walls tracedalong the surface appeared to be continuous it seems more likely thatit actually was constructed in several sections along preconceivedlines. The group of rooms forming a rectangular block comprisingthree rows at the back of the building is like the original block inthe smaller building to the north and may represent the nucleus ofthis structure. To this was added the row of wider rooms and thenthe kiva group. The series of small chambers along the south endof the building probably represents the last of the construction work.However, it is not wise to be too emphatic in the matter, as excava-tions frequently reveal curious and unexpected features in largebuildings. The last two rows of rooms, those farthest removed fromthe kivas, at one time were at least thi-ee stories m height and it is
"Morris, 1921, p. 133.
RoBEBTs] ARCHEOLOGICAL REMAINS IN ARIZONA 249not improbable that the outside row was four. There was suflScientfallen material along the outside wall to have formed a fourth floor.Portions of the bases of third-story walls were still standing in thesecond row from the back. The third row gave indications ofonly two stories and the next, one. It is doubtful that the kivas andthe rooms adjacent to them stood a full story above ground, judgingfrom the debris. At all events the building had been a characteristicterraced dwelling of Great Pueblo type.The series of rooms along the south end and the east side, the frontof the building, possibly were for storage purposes. Their size andshape suggest that function rather than dwelling chambers, althoughsome of them could have served in the latter capacity. The largerrooms in the main block back of the kivas unquestionably were themain domiciliary enclosures. It is doubtful that all Avere used asliving quarters. The lower and back chambers in buildings of thistype generally were dark and poorly ventilated, particularly those onthe ground level where no openings were left in the walls, and wereused as storage places. The roofs of the kivas and the rooms sur-rounding them probably served as a dooryard for the first row ofrooms, the roofs of the latter bearing the same relation to the second-floor chambers of the next row and so on to the top and final row.Under such an arrangement, approximately 40 rooms would opento the outside light and air, and these no doubt were the main livingquarters. On the basis of conditions in more recent villages thiswould indicate a probable maximum occupation of 40 families. Itis not likely that there ever was a complete contemporaneous use ofall these rooms, so that an estimate of 30 families is more in keepingwith the conditions indicated in most communities of this type. Onthis basis the population may be postulated as ranging between 150and 200. The three kivas suggest three clans with a possible 8 to10 men for each ceremonial chamber. These figures are, of course,purely speculative, but the general indications augmented by knowl-edge of conditions elsewhere warrant the opinion that they are notfar from a reasonable approximation of the possible size of thegroup. This building had a length of 121 feet 6 inches (37.033 m)and a width of 87 feet (26.518 m). The kivas were approximately20 feet (6.096 m) in diameter.Whether the second structure was built by the people living in tliefirst and smaller one, that appeared to have been constructed first,or by groups withdrawing from various outlying unit villages andjoining tlie community at the old site is not known. The first build-ing may have continued in occupation after the completion of thesecond or possibly was partially abandoned in favor of new accom-modations in the larger structure. The smaller one may have beendeserted gradually until only portions of it were inhabited at the
250 BUREAU OF AMERICAN ETHNOLOGY [Boll. 121time when the entire community withdrew from the district. Themaximum population for the original dwelling was probably between50 and 60 people. A gro\ip of this size would not require a structureas large as the second and there is little question but what they werejoined by others from the immediate neighborhood. If both werecompletely filled at the same time the village would have approxi-mated a population of from 200 to 250 people. Judging from pot-sherds scattered about the ruins of the buildings and in the refusemounds, the site was abandoned before the peak of the Great Puebloperiod was reached. Reasons for the withdrawal were not indicatedby any of the surface features.The walls in both structures, where visible, were, built of care-fully selected stones sliaped for the purpose and dressed along theedges. They formed the outside surfaces while the inner portion, thecore of the wall, consisted of earth and small stones?a rubble fill.Very little mortar was used between the courses. There was somechinking in the form of small stones and a sporadic course of thinnerblocks ap])earing here and there in the walls suggested some of thebanding so characteristic of Chaco Canyon masonry. It was not as])ronounced nor as extensive here, nevertheless it was reminiscentof that construction. The masonry was also suggestive of that inthe ruin on the Zuni Reservation that was accompanied by two ofthe superkivas,^" and hnd some similarities to walls in the surfaceruin at the Ijong H Ranch,^^ Both of these sites exhibited definiteafTmities to the general Chaco cultural pattern, and they, as well asthe present ruins, probably represent some of the more importantcenters in the southern periphery of the Chaco range.The two small rectangular structures in front of the main housemounds are somewhat puzzling because their purpose is not clear.It is not common to find remains of this type in association withlarger ruins in this part of the Anasazi province. Whether theywere single-roomed domiciles, granaries, or the bases of towers isnot known. There was sufficient fallen material around them to havecarried the walls to a height of approximately a story and a half.It was this factor that suggested the possibility of their havingbeen towers, although low examples of that form of structure. TheSan Juan region in the north, particularly the southern Coloradoarea on the north side of that river, has many structures of thattype in association with regular conununal dwellings.?- If the ruinshere are the remains of that form of building they indicate a consid-erable spread for the feature. The function of towers has not beendetermined as yet. Some have suggested that they were for lookout
?^Roberts, 1932.?> IJoberts. 1931.
*?- I^ewkes, 1919.
Roberts] ARCHEOLOGICAL REMAINS IN ARIZONA 251purposes, others that they were for defense, or were observatoriesand aided in the astronomical studies by which ceremonial perform-ances were governed. They have been considered in the light ofplaces where special rites connected with the worship of astralphenomena were held, or simply as granaries. At the present sitetliey would not have been particularly advantageous lookout posi-tions because people on tlie housetops could have seen farther. Themain buildings undoubtedly had ample storage space in the lower,back rooms, so the granary function hardly seems logical. Theycould have served as defensive towers, in that men on their roofswould have been in advantageous positions to harass the rear orflanks of an attacking party that had reached the front walls of thedwellings, but this indicates a better understanding of military tacticsand strategy tlian is generally accorded the peoples of that group andperiod. They certainly would have had little defensive value beyondthat of the large buildings from the standpoint of enemies approach-ing the site from the front and sides. Hence that explanation d?jesnot seem very plausible. There were conditions under which theycould have been of value in making observations of the position of therising sun to aid the priests in setting the time for winter or summersolstice ceremonies or for those pertaining to the spring and autumnalequinoxes. There is no way of establishing such fimctions and theycan only be postulated. Perhaps they should not be thought of asanything but the remains of small secular structures. Their location,both with respect to each other and to the main buildings, seems tooexact and definite to be mere coincidence and for that reason thevarious suggested explanations have been presented. The structureat the south end measured 18 feet (5.486 m) and 17 feet 3 inches(5.257 m) on the sides and 15 feet 9 inches (4.800 m) and 14 feet 10inches (4.521 m) on the ends. The remains were 48 feet 6 inches(14.782 m) from the larger dwelling. The second small structure was68 feet (19.202 m) north of the fir.^t. It was 17 feet 6 inches (5.334 m)and 17 feet (5.181 m) on the sides and 13 feet 10 inches (4.216 m) and14 feet (4.267 m) on the ends. It was 27 feet 6 inches (8.382 m) fromthe gre,at kiva and 105 feet (32.004 m) from the smaller of the twodwellings.The two arroyos that had cut their ways through the northernrefuse mound exposed most of the layers in the deposits. This madeit evident that the waste material from the village had been thrownon the slope of the ridge dropping away from the village toward theedge of the low bluffs to the east of the site. The refuse attained amaximum depth of 7 feet (2.133 m) just below the upper edge of theslope and fanned out gradually down the hillside. Portions of ithad been washed some distance from the original termini of thedump. The main factor of importance revealed by the banks of the
252 BUREAU OF AMERICAN ETHNOLOGY [Bull. 121arroyos was the total absence of all human bones. Generally whena mound of such size is disturbed to the extent this one was, a numberof burials are exposed or washed out. It is not possible to reachconclusions on the problem without complete investigation of themound, but it appears that very few of the inhabitants could havebeen interred there and in this respect there is another similarity toChaco remains. Burials are only rarely found in the refuse heapsassociated with the large house ruins of that cultural pattern. Theyare often present in the dumps of small-house remains but not thelarge ones. The reason for this has not yet been determined.The mound at the south was thoroughly trenched because it wasthought that it might cover some pit dwellings. That work revealedfour burials, three of which were below the bottom of the mound andwere accompanied by late Developmental style pottery and hencecould not be attributed to the dwellers in the large structures. Thefourth did have mortuary offerings characteristic of the Great Puebloperiod, but that does not vitiate the comparison to Chaco traitsbecause a sporadic example is sometimes found even in the maincenters of that pattern. This second mound was large in extent butrather shallow. Its maximum depth was 3 feet (91.44 cm). Themound cut by the arroyos was the first, judging from potsherd ma-terial, and probably represents, for the most part, the sweepings andwaste matter from the original structure. The northwestern side, inthe vicinity of the great kiva. incorporated most of a refuse moundthat contained Developmental type potsherds. As a matter of fact,part of the great kiva pit had been dug into the old dump. Thismaterial may have come from structures 12 and 14 but it is morelikely that part of the original large dwelling was erected over theremains of some pit structures. The second refuse mound undoubt-edly correlates with the larger structure.The potsherds and other material scattered through and over thesurfaces of these two mounds indicate that community was roughlycontemporaneous with the Great Pueblo village on the Zuni Reserva-tion, where the two examples of great kivas, previously mentioned,are located. In some respects the indications suggest a slightlyearlier beginning for the structures near Allantown and an earliertennination of occupancy. This may be relative rather than actual,however, and the result of a cultural lag. Dendrochronological datesfor the Village of the Great Kivas place its beginning at 1015 to1030. That date would agree rather well with the indications here,because there is a definite overlap in the nature of the potsherdsfrom the north mound and those from the third unit in the valleybottom where the dates were 1005 to 1014 for the house and 1007 to1014 for the kiva. On the basis of this evidence, meager to be sure,when it is considered that there has been no excavation in the large
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA 253ruins, it may be suggested that the first section of the Great Pueblocenter was built at about thei same time as the third unit and thattJie abandonment of the latter possibly correlates with the beginningof the second and larger dwelling. There is no question but whatthe large ruins fall within the eleventh century, because there wereno traces of pottery types prevalent in this general area that havebeen dated to the early twelfth century and the beginning of thevillage must be placed after 1000 A. D. Complete excavation andcareful study of these structures w^ould no doubt throw interestinglight on a nimiber of problems connected with the diffusion of cul-tural traits from various districts and relationships between differentpatterns. These ruins are on the very periphery of the westwardexpansion of the Chaco pattern and at the eastern edge of influencefrom Little Colorado centers located farther west. As a matter offact, it is necessary to go only a few miles west until, in the vicinityof Houck, differences in the pattern become quite apparent. This istrue for the earlier horizons as well as those of the more highlydeveloped stages.SUMMARY AND DISCUSSIONIn the Whitewater district in the vicinity of Allantown, in east-ern Arizona, are numerous remains belonging to several stages inthe growth of the Anasazi cultural pattern. Three season's work inthe ruins concentrated in one small section of the region producedconsiderable information on various elements in the complex andtlii'ew interesting light on some phases in the growth of character-istic features. Although there are traces of Modified Basket Makerand some large ruins representing the Great Pueblo period, most ofthe digging was done in Developmental Pueblo remains. This re-sulted from the fact that the area under excavation produced onlyscattering indications of Modified Basket Maker, more w^ere expectedwhen the project was started, and from the necessity of bringing theinvestigations to a close before the digging planned for the GreatPueblo structures could be begun. Such Modified Basket Makerfeatures as had been present on the site were so disturbed and mixedby the activities of the subsequent occupants that little could belearned about them beyond the fact that they had been there. Afairly complete picture of the nature and trends of the Develop-mental Pueblo period is presented by the information obtained from20 pit structures and their accompanying granaries and surface shel-ters, and from three unit-type ruins with associated ceremonialchambers. Tlie remains clearly show the growth from small single-roomed pit dwellings with brush and plaster covered truncated-pyramidal superstructures supported by four upright posts set in
254 BUREAU OF AMERICAN ETHNOLOGY [Bull. 121the floor of the chamber, to above-ground houses with several con-tiguous rooms, masonry walls, and flat roofs. Correlated with thisis the shift in function of the semisubterranean domicile from secularto ceremonial purposes. These changes were accompanied by com-parable ones in other features, particularly in the lesser objects of thematerial culture, such as pottery, stone and bone implements, that areconsidered in a second j^art of this report to be published at a laterdate.The pit structures on the whole are comparable to those in othersections of the Anasazi province. There are various individual andlocal diflferences of a minor nature that probably have no significanceas far as the structural type is concerned. They no doubt representpurely personal whims on the part of the builders and had no bear-ing on the subsequent trend in construction. On the other hand,there are features that adumbrate later developments. In only threeexamples (four if Kiva A of the first unit is included) was the pitportion of a dwelling encircled by a definite bench. This correspondsclosely with the situation at a Modified Basket Maker site in theChaco Canyon where 2 out of 20 structures had benches and theremainder did not. At the Long H Ranch, southwest from theWhitewater ruins, in a series of 16 pit remains, 9 were without abench and 7 had the feature. This tends to strengthen the sug-gestion made in a previous discussion of the Long H group to theeffect that the structures witli encircling bench probably representa later development of the pit type of house. The adoption of thisstructural element apparently did not occur until shortly beforethe transition that culminated in the circular, subterranean ceremo-nial chamber called kiva. The lack of a bench, however, does notnecessarily mean that a structure is older than one that has such afeature, because the older forms persisted to some extent after newones appeared. The proportion of one style to another in a seriesof structures or in a village probably does have significance, althoughhow much emphasis should be placed on it is still questionable.Two of the pit structures, 1 and 3, contained examples of whatmay be considered a prototype of the pilaster, as well as an illus-tration of the manner in which that architectural feature developedin the Southwest. In the series of houses as a whole was evidence ofa shift in function in the floor basin adjacent to the fire pit; as amatter of fact it is possible to trace the origin, growth, and changeof the feature in this one set of house remains. A nmnber of thepits that probably represent the early stage of occupancy have anabraded depression in the floor adjacent to the fire pit that seemsto indicate the resting place of the base end of a ladder. Thewearing of a basin apparently suggested the construction of an actualbox to hold the end of the ladder. In a number of the structures
ROBBBTS) ARCHEOLOGICAL REMAINS IN ARIZONA 255this takes the form of a simple concavity, with carefully plasteredsides, ill the floor. Then the use of stones to line the pit was intro-duced and the feature became a definite box. In the case of pole-and-rung ladders so much of the space would be unoccupied by thqladder ends that it offered a place where ashes cleaned from thefire pit could be deposited temporarily until they could convenientlybe removed from the structure. When the advantage of this becamefully apparent it may have led to the idea of removing the ladderfrom the pit and embedding the ends in the floor, to prevent slip-ping, so that the entire box could be employed as an ash basin.Whether the evolution of the ladder box-ash pit followed exactlythe steps as outlined or not there is no question of the change, asthat was clearly shown in the remains of the pit structures. In somecases the houses may have been equipped with a notched-log ladderrather than the pole-and-rung style, but that would have little effecton the question of a placement for the lower end. There was noevidence for the single-log type in any of the remains, althoughthere were definite indications of the other form.A unique feature in the pit structures at this locality was that ofcombining several into one larger building, as illustrated by 6 and7, 9, 10, and 11. An interesting factor in the joining together ofthese dwellings was that the units that went into the make-up of thelarger building were individually complete. An interpretation ofthe situation is that although the people were ready to combine theirhouses they were not yet willing to give up house characteristics thathad long been present in the single dwellings. Reference was made,in this connection, to the old theory that the rectangular-roomedcommunal buildings of the Pueblo peoples were an outgrowth of thepractice of combining a number of circular houses into one largestructure. The evidence offered by other examples, the structures16 and 16 groups and the first unit, indicates a more logical stej) inhouse development, however, and for that reason the combined pitstructures, one containing two and the other three houses, are con-sidered as a peculiar local development that did not have a directbearing on the growth of the communal type building in general.The superstructure for No. 12 was the only one of its kind in thisdistrict. It may represent a purely local method for covering a pitthat was above the average in size, or it may be characteristic of thetype of construction employed for larger structures. The latterprobably is the pro])er answer rather than that it was a local devel-opment. A certain similarity between the framework here, especiallythe outer ring of supports, and some of the Plains earth lodges wasmentioned. Attention should also be called to the fact that a compa-
256 BUREAU OF AMERICAN ETHNOLOGY i^vvx^ 121rable secondary outer framework was used in some of the Hohokamand Mogollon houses "^ and also was present in a large circular struc-ture in a Modified Basket Maker village in the Chaco Canyon.^*Since these other structures are consistently larger than the ordinarypit dwellings of the Anasazi province it is probable that the style ofconstruction was that used when a greater area was roofed over.Due to its apparent widespread distribution and the fact that it ispresent in various stages in different cultural patterns it probablyrepresents a common trait in the general pit house-earth lodge com-plex rather than evidence for later borrowing or influence from somespecific pattern. Future work, however, will no doubt throw addi-tional light on the subject and may give some indications of the linesalong which it diffused.The evolution of the unit-type dwelling, which has long been con-sidered the basic component in the development of the great com-munal buildings that were characteristic of the Anasazi province, iswell illustrated by several groups at this site. First in the series isstructure 16 with its associated granaries. In structure 16 entrancewas by means of a ladder placed in a small shaft that was connectedto the main chamber by a covered trench or passage. The granarypits, located west of the structure, were roughly oval in form andtwo of them gave evidence of having been used as shelters or a make-shift habitation. The second in the series is the structure 15 group.There the side shaft and passage was too small to serve as an actualentrance and had become a ventilator. Access to the chamber wasby means of a ladder through the hatchway in the roof. The gran-aries, located north and west of the structure, had become rectangularin form, were built in a row, and some of them gave evidence ofliaving been lived in. Outdoor fire pits and brush shelters were asso-ciated with the granaries and a rude form of portico w^as erected infront of one of them. The third group is that of the first unit inthe Developmental village. The pit structure retains many of thefeatures noted in the others and probably was an actual dwellingwhen first completed. The gi'anaries had become rooms, however,built in a contiguous row and some of them contained marked evi-dence of use as a definite habitation. The portico was retained at oneend of the building and a small court with open fire pit provided atthe other. The pit structure probably was used less and less fordomiciliary purposes and more and more for ceremonial purposes,so that it represents a transition in function from house to kiva.Fourth in the series is the second unit in the Developmental village.Here the pit structure had become a kiva with typical characteristicsand the granaries were definitely a habitation, although they occu-
rs Haury, 1932, fig. 14 ; 1936 b, fig. 26.? Roberts, 1929, pp. 78-79.
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA 257pied the same position relative to the pit remains. The last in theseries is the fully developed form illustrated by the third unit withits surface dwelling, court, and subterranean ceremonial chamber.The sequence was determined by stratigraphical evidence based onmaterial from the various refuse mounds and test sections, supple-mented by that from the pillars left from the fill in various struc-tures, and by dendrochronological dates. While there is a certainoverlap in the case of structure 15 and the first unit, it is not of sucha nature as to vitiate the illustration of the growth in type.The changes taking place in the various structures can be sum-marized briefly by saying that the typical small house and its asso-ciated subterranean ceremonial chamber grew from a semisubter-ranean pit structure and its accompanying granaries. The evolutionin this district includes shifts in nature and function for both thegranaries and the pit structures. The former passed from roughlyoval or circular forms to those wath a rectangular outline; theirsuperstructures shifted from a pyramidal to a truncated type; thewalls from a sloping pole and plaster variety to vertical ones madeof stone. Associated with this was the growth in masonry types.From a series of disconnected cells the granaries became a row ofcontiguous rooms. At each change they became shallower until theirfloors ultimately attained the ground level. From storage places theybecame habitations. In the circular structures there was progressionfrom a side-entrance passage to the smoke-hole-hatchway mode ofaccess. The old entrance was retained in reduced and slightly modi-fied form as a ventilator. The truncated-pyramidal superstructuresupported by upright posts was replaced by the cribbed style of roofresting entirely on the top of the bench encircling the pit, evidenceof such replacement being found in Kiva B of the second unit. Thebench apparently was a development out of the narrow shelf formedat the top of the pit wall when the sloping side poles of the super-structure w^ere set back from the edge of the excavation to preventtheir slipping into the chamber or damaging the sides of the roomIn contrast to the rising floor level in the granaries, the pit of theoriginal domicile became deeper and deeper in the shift from dwell-ing to ceremonial chamber until from a semisubterranean structureit became one that was wholly below the level of the ground. Afterthe adoption of the large communal houses, erected by combiningnumbers of units together, the ceremonial chamber emerged fromthe ground and was incorporated in the block of the building, itsformer subsurface characteristic being simulated by a fill of earthbetween the curved wall of the kiva proper and the enclosing rec-tangular room that made its inclusion in a building possible. Thekivas along the front of one of the Great Pueblo ruins probably
258 BUREAU OF AMERICAN ETHNOLOGY [Bull. 121represent this last stage, although no definite statement can be madedue to lack of investigation in the ruin.Another unique feature at the main site consisted of a large cir-cular adobe pavement bordered by a series of small upright stoneslabs. Adjoining the circle at its northern side were traces of athree-sided structure that enclosed a form of dais opening towardthe floored area. For want of a better name these remains werecalled a dance court. That the place actually was used for such apurpose is not known. There was no evidence for a superstriictur?or covering of any nature over the enclosed area, but because thegeneral plan is suggestive of a great kiva or superceremonial cham-ber an explanation is proffered to the effect that it possibly func-tioned in a similar capacity. During Developmental Pueblo times atthis location tlie rites and rituals that later were performed in thegreat kiva present at the north end of the nearby Great Pueblo luinsmay have been held in the dance court. It is possible, of course,that the dance court was merely a temporary provision for the ob-servance of communal ceremonies while the great kiva was underconstruction. The court is unusual, nevertheless, and thus far noother example has been described for the Anasazi province. Hough'sgreat dance pit at I^una may be analogous in purpose, but the formsdiffer to a considerable extent.?^ Jeancon's dance plaza on the PiedraRiver in southern Colorado more nearly approximates the presentexample, yet it is different.^" There does not seem to be any validbasis for the assumption, made by several visitors to the site whilework was under way, that the dance court represents a variation ofthe so-called ball courts of regions farther to the south.The Great Pueblo ruins in their unexcavated state suggest twolarge buildings and a great kiva placed so as to form a partiallyenclosed court in which two rectangular towers were located. Suchof the masoniy as is visible and the potsherd and other materialsscattered over and through the refuse mounds indicate that the com-mujiity had certain affinities with the Chaco Canyon pattern and thatit was roughly contemporaneous with a Great Pueblo village on theZuiii Reservation. The smaller of the two buildings seems to havebeen erected first, and it is possible that its original nucleus datesfrom about the same time as the third unit in the floor of WhitewaterValley. The larger building may have been constructed by gTou^Jswithdrawing from the various scattered units in the district andjoining the center on the ridge that was started when the first ter-raced house was erected. All indications ai-e that these ruins fallwithin the eleventh century and that while their main features point
? Hough. 1919, pp. 414-415.
?" Roberts, 1930, p. 33.
RoBKKTs] ARCHEOLOGICAL REMAINS IN ARIZONA 259to a nortliern and eastern influence, there are some traces of elementsreaching them from the Little Colorado region to the west and othei^from the south and southeast.Dates obtained from charred beams found in different structuresprovide some interesting data on several aspects of the growth of thePueblo peoples. Approximately 2 centuries are represented by theremains excavated and described in the preceding pages. This omits,of course, the Great Pueblo ruins for which no dates are available.The earliest building date is that for structure 2 and the latest thatfor the third unit. Timber from structure 2 gave dates 814 and 815,the latter being considered as the year for the erection of the struc-ture because it is the latest in the remains. Orie timber from the kivain the third unit recorded the cutting date as 1011?2 or a maximumof 1013. A fragment from the house in the third unit is recordedas 994?:20, or a maximum of 1014. Using the minimum of 814 andmaximum of 1014 gives exactly 200 years. The third unit was no<loubt occupied for a number of years after it was completed, butfrom the standpoint of the evolution of the house type that is notimportant. The significant thing is that the unit type in the fullsense of the word was fully establislied by 1014 in this district. Dur-ing the span of 2 centuries the domicile changed completely from asimple pit dwelling and associated granaries to an above-groundhouse and associated ceremonial chamber. When regarded more spe-cifically, however, the transition covered a shorter interval. Struc-ture 3, it will be recalled, was considered as a kiva for a unit groupand its building date was 867 or soon thereafter. Kiva A in the firstunit in the Developmental Village contained the single date 845.There is some question about this date, but inasmuch as the latterstructure gave every indication of being first a house, then a cere-monial chamber, and as an accompaniment of the change associatedgranaries were replaced by actual dwelling rooms, the year may notbe far off if considered as representative of the original construction.This is given some support by the fact that one beam from structure3 showed a cutting date of 842 and that a number of those in struc-ture 12 read 844. Although the latter were no doubt reused timbersin structure 12, they show that considerable activity was taking placeat about 845. On the other hand, structure 15 with its associatedgranaries, a group that illustrates an elementary stage in the evolu-tion, yielded the building date of 888, although this relates to a re-construction following a fire and not the original building of thehouse. In this connection it was pointed out that structure 15 pos-sibly was the dwelling of an immigrant group coming in from thenorthwest and for that reason was somewhat of a laggard in develop-ment. The occupants might have been a conservative element inthe communitv and slow to adopt new ideas. Still, on the basis of
2QQ BUREAU OF AMERICAN ETHNOLOGY [Boll. 121these several remains it can be said that the trend toward the unit-type dwelling had its inception shoi-tly after the middle of the ninthcentury and in at least one example had attained a definite patternby 867. The second unit in the Developmental Village produced nodates. On the basis of potsherd comparisons between it and struc-ture 12, with its readings of 918, there is little doubt but that it closelyapproximates that year. The second unit was not a fully developedexample of the perfected type, yet it exhibits sufficient characteristicsto warrant its being considered an illustration of a crystallizing planof construction. Thus it seems logical to conclude that the majordevelopments took place in a span of from 50 to 60 years. The ap-proximately 80 years elapsing between the erection of the second unitand the third unit were marked by refinements in the plan, not byany major changes. Whether the development and transition wasas rapid in other sections or not still has to be determined. At pres-ent there is not enough evidence on this stage for other districts towarrant conclusions. Because of indications that the Whitewaterregion was influenced by the larger centers to the northeast and thatit was marked in some degree by a peripheral lag, it seems that thedevelopment may have been more rapid here than would have beenthe case had it resulted from independent efforts. Nevertheless itis a good illustration of the extent of change that can take place ina short space of time, even among people considered to be gropingtheir way toward a higher cultural level.The dendrochronological dates from the individual structures sug-gest a number of interestiiig items and also raise several troublesomequestions. Perhaps the most important single factor in this connec-tion concerns the stress that should or should not be placed on thesignificance of a date for a ruin based on only one or two timbers.Three of the structures in the Whitewater series contained beams thatcovered a span of from 25 to 74 years for a single building. Thosefrom 3 ranged over a 25-year interval, those from 15 included 31years, and those from 12 incorporated the 74-year group. With theinformation available from the work at the site it is possible to con-clude that this condition is due to the use of old timbers and beamssalvaged from abandoned houses, but under other circumstances anentirely erroneous conception could be given by a limited set of dates.For example, if only 3 or 4 fragments of charcoal had been savedfrom structure 12 and, as could easily happen, they were all fromthe early end of the series, the implications would indeed be mislead-ing. From such evidence the remains would naturally be placed ap-proximately 74 years earlier than their actual date and in the presentinstance that length of time woidd more ihan cover the interval dur-ing which there were marked changes in the cultural pattern. Wlien
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA 261the rapidity of change is considered even the quarter of a centuryrepresented by the wood in structure 3 could make a tremendous dif-ference in interpretations of evidence and the problem of correlations.For that reason it seems that dendrochronological dates, accurate andhelpful as they are, must be used with a great degree of discretionand when only a few are a^'ailable from a site they should be regardedmore as an approximate indication of age rather than a definitedemonstration of the precise interval of time encompassed by theremains. Even with such reservations, the years recorded will be ofconsiderable assistance in solving numerous problems. In the presentgroup of ruins little emphasis could be placed on the 845 from KivaB were it not for the complementary evidence from other structuresand also from various archeological manifestations that augment thedendrochronology. As it is, the correlation between the date and thestructure is not above question, as was pointed out in the discussionof the second unit.The question of how to interpret the dates from a single structureis not easily answered. From the standpoint of accepted archeologicalpractice the latest material is the date determinant and for generalpurposes the mere statement that the last year recorded was suchand so would suflBce for the age of the house. Yet there may be some-thing of significance in a series covering an interval of some lengthand proper interpretation of it may have much more value than theactual terminal date. The timbers from structure 2 show 814 and 815and probably represent the span of a single year. The explanationthat seems most logical is that the house was built in 815 from timberscut in 814 and 815. In this case the problem is relatively simple.The house in the third unit yielded timbers with dates from 994 to1004. The erection of the house would, according to rule, be placedat 1004, but when the series is studied it is noted that all but the1004 beam have dates either 994 or 995. Hence it seems that theactual building date was probably 995 and that the single timbercut 9 years later represents a replacement or bit of repair worknecessitated by the wear and tear of occupancy. Wliile it is notimportant in this particular instance, it is not inconceivable thatsuch a 9-year difference might have a pronounced effect on thevalidity oi the conclusions made with respect to specific featuresat a site.The 74-year span in structure 12 furnishes another example of aproblem in explanation and the choice of procedure in the use ofthe evidence. The terminal date for the structure is 918, the earliest844. In the discussion of the structure it was suggested that timbersfrom an older house, probably a smaller one occupying the same68764?39 18
2Q2 BUREAU OF AMERKJAN ETHNOLOGY [??"- 121spot, were used in the construction work of 918, or shortly there-after. This is a good explanation and likely a correct one. Examin-ing the series of beam years, however, the student is struck by thefact that there is more to be considered than the building of astructure in 918 making use of a number of logs cut in 844. Thereis one cut in 845, one cut in 853, several that show 862 as the yearwhen they were felled, another with the date 870. Now the questionis, how are these to be regarded ? Was the original structure builtin 845 from timbers cut for the most part in 844? That seems alogical conclusion. Then, was the 853 specimen a timber replacementrepairing a damaged section of the roof? Such an explanation isentirely plausible. But the 862 group is more difficult to interpret.Were extensive repairs called for at that time in the original struc-ture, or were timbers salvaged from still another house to completethe material needed in the 918 construction? Either would be asatisfactory answer, yet there is nothing to indicate which is morelikely the correct one. Again it might be asked, can not the 862 seriesrepresent the actual time when the structure was built and the 870and 918 dates be those t)f repaii-s? Offhand the reply would seemto be yes. Further investigation, however, shows that the 870 tim-ber was one of the main support posts for the supei-structure andit probably was not a replacement or bit of repair work. Hence theerection of the structure in 862 could be ruled out and as it pre-sumably was after 870 and 918 is the next date [)resent that appearsto be the logical one for the remains. The other features can notbe wholly ignored, though, and there is still the problem of howthey should be interpreted and what emphasis should be placed onthem. Perhaps mention of the various possibilities is enough. Thewriter frankly is at a loss to know just how to use such informationto the best advantage. The situation is further complicated by thefact that it frequently is necessary for the dendrochronologist towrite his dates with a plus or minus sign qualification. This meansthat the year is as given but that several more should be added,just how many not being known, up to the limit of the qualifier.When the date is used should the qualifier be dropped consistentlyor should the maximum be added in every case and that made thenumber? A majority of the archeologists seem to be in a similarquandary and there is no standard followed in all cases. From aconservative point of view the safest plan is probably that of habit-ually adding the qualifier to the date.From a comparative standpoint the dates from the Whitewaterruins indicate certain interesting factors. The earliest, 815, from
:roberts] ARCHEOLOGICAL REMAINS IN ARIZONA 263Structure 2 is from a house showing much less development than a pitstructure in the Chaco Canyon that produced the date 777 ?10, or amaximum of 787." The structure in the local series that was morenearly comparable to that in the Chaco was No. 15 with its 888?15,or a maximum of 903. Then structure 12 with 918?3, or a maximumof 921 rather closely establishes the date for the second unit. In the?Chaco Canyon both architecture and ceramics were mucli fartheralong by 921 than they were in this district. Pueblo Bonito hadbeen started and types of pottery had appeared that were still un-known in the vicinity of Allantown. The maximum date for thethird unit was 1014, 994 ?20. By 1014 Pueblo Bonito was definitelya Great Pueblo center and various features characteristic of thatstage were well established in the Chaco. but many of them ap-parently had not reached this location. The ruins at the Village ofthe Great Kivas on the Zuni Reservation gave a date of 1015?15, ora 1030 maxinuun. The general pattern at that location was Chacoan,and while not as advanced as that in the Chaco at 1030, did havesome things not present in the 1014 unit on the Whitewater. How-ever, the beginnings of the Great Pueblo stage at the main part ofthe site may have antedated slightly the beginnings at the Zuiii com-munity, as was pointed out in the discussion of those structures. Onthe basis of this data it seems quite clear that the ruins in the White-water district south of Allantown represent a peripheral lag in theChaco pattern and, despite many recent expressions of opinion to thecontrary, the flow of influence was from the Chaco toward Allan-town and not the reveree. The evidence indicates that the move-ment was slower in earlier than in later times and that when theGreat Pueblo stage was reached the lag was not very pronounced.Actual dates from the large ruins here might change the picturesomewhat, but the information at hand suggests the trend was asdescribed.Stratigraphic tests made from pillars left from the fill in variousstructure pits and in different parts of several refuse mounds aremainly important for a study of the ceramic sequence and for changesin artifact types. The fill in a number of the houses indicated cer-tain intervals when the area immediately adjacent to them was un-occupied. This probably means that the people had moved else-where on the site o?- in the surrounding district rather than that theregion was completely deserted. During the earlier stages of theDevelopmental period there was more of a tendency to shift aboutthan was the case after the larger, more permanent type of dwellingbecame established. For this reason the ruins in a district suggesta larger population than actually inhabited it. In the present in-stance it is not improbable that the people, for one reason or another,
^ Judd, 1924, house No. 2 ; Douglass, 1935, p. 61.
2g4 BUREAU OF AMERICAN ETHNOLOGY [Bull. 121moved from the top of the ridge to the valley bottom, then back (othe ridge again, or to another valley location before returning tohigher ground.Most of the burials for the various groups and units were madein the refuse mounds. In some cases the interment was in a roomor beneath an occupied area, as mentioned in the discussion of thevarious structures. The flexed type of burial predominated and ina majority of them funerary offerings of pottery vessels and otherobjects accompanied the remains. Most of the skeletons were in apoor state of preservation and out of a total of 150 only 15 aresuitable for anthropometric studies. The material is in the hands ofDr. T. D. Stewart, assistant curator of the Division of PhysicalAnthropology, United States National Museum, and a complete re-port on it is to be included as an appendix to the second part of thereport on the Wliitewater district. Burials of another type, thoseof turkeys and dogs, are common in the refuse mounds of the dis-trict. In a majority of cases the interment appears to have beenintentional, not an accidental covering over of a dead fowl or animal,and the body frequently was accompanied by offerings of one kindor another. Potsherds rubbed down to form shalloAv dishes, com-plete miniature vessels in some graves, and corn or animal bones,according to whether it was a turkey or a dog, were the customaryobjects for that purpose. The location of all burials is sliown onthe maps of the various groups and units.The food supply for the district consisted of corn, beans, seeds froma species of wild grass,^^ pinyon nuts, rabbits, deer, and antelope.It is questionable that the turkey was used for food. Most of theturkey bones found were parts of articulated skeletons in definiteburials. Other vegetal products were no doubt utilized, but thosenamed are the only ones that were found in the houses or refusedeposits. Every example recovered was charred, which probably ac-counts for the preservation. On the basis of what is indicated thereis no question but what the group here had the same economic pat-tern as the rest of the Pueblo peoples. They were mainly agricul-turalists, supplementing their diet by such hunting as the countryafforded.Estimates of population for the district are not possible becauseonly a portion of the ruins was excavated and it is not known howmany were contemporaneous or what the total number for any onestage may have been. The most that can be said is that the unit-type structures probably housed from 10 to 20, the pit dwellings t>to 10, the smaller of the large structures from 50 to 60, the larger
?? Eriocoma cuspidata Nutt. (Oryzopsis hymenoidea (Rocm and Schult.) Ricker), Identi-fied by Dr. Melvin R. Gilmore, Ethnobotanical Laboratory, Museum of Anthropology,University of Micliigan.
ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA 265from 150 to 200. As suggested in the discussion of the Great Puebloruins, the maximum for that community was probably about 250people. If the Whitewater Valley proper is considered, the area with-in a radius of 2 miles (3.219 k) of the large ruins probably nevercontained more than 300 people at any one time. The average wasnot likely that large. Suggestions concerning the district as a whole,however, must wait until further work has produced more evidencethan is available at the present time.

LITERATURE CITED
BUSHNEXL, D. I., Jr.1922. Villages of the Algonquian, Siouan, and Caddoan Tribes West of thfrMississippi. Bulletin 77, Bureau of American Ethnology, Wash-ington.Ck)UES, EJlliott.1897. New Light on the Early History of the Greater Northwest. TheManuscript Journals of Alexander Henry and David Thoujpson,1799-1814, vols, i-in. New York.CuMMiNos, Byron.1915. Kivas of the San Juan Drainage. American Anthropologist, n. s..vol. 17, no. 2, pp. 272-282, Lancaster.Gushing, F, H.1896. Outline of Zufii Creation Myths. Thirteenth Annual Report Bureauof Ethnology, pp. 321-447, Washington.Douglass, A. E.1932. Tree Rings and their Relation to Solar Variations and Chronology.Annual Report Smithsonian Institution for 1931, pp. 304-313,Washington.1935. Dating Pueblo Bonito and Other Ruins of the Southwest. PuebloBonito series, National Geographic Society, Contributed TechnicalPapers, no. 1, Washington.Fenneman, N. M.1928. Physiographic Divisions of the United States. Third edition, revisedand enlarged. Annals of the Association of American Geogia-phers, vol. xvni, no. 4, Albany N. Y.Fewkes, J. W.1904. Two Summers' Work in Pueblo Ruins. Twenty-second .\nnualReport, Bureau of American Ethnology, pt. 1, pp. 3-195, Wash-ington.1909. Antiquities of the Mesa Verde National Park: Spruce-tree House.Bulletin 41, Bureau of American Ethnology, Washington.1911. Preliminary Report on a Visit to the Navajo National Monument^Arizona. Bulletin 50, Bureau of American Ethnology, Wash-ington.1919. Prehistoric Villages, Castles, and Towers of Southwestern Colorado.Bulletin 70, Bureau of American Ethnology, Washington.Gladwin, Winifred and H. S.Undated. The Red-on-BufE Culture of the Papagueria. Medallion Papers,rv. Globe, Ariz.1929. The Red-on-Bufif Culture of the Gila Basin. Medallion Papers,.in. Globe, Ariz.1980. An Archeological Survey of Verde Valley. Medallion Papers, vi^Globe, Ariz.1935. The Eastern Range of the Red-on-Buff Culture. Medallion Papers,.XVI. Globe, Ariz. 267
268 BUREAU OF AMERICAN ETHNOLOGY f^ui'i- 121Gbkgoby, H. E.1016. The Navajo Country. A Geographic and Hydrographie Recon-naissance of Parts of Arizona, New Mexico, and Utah. UnitedStates Geological Survey, Water-supply Paper 380, Washington,GUKRNSEY, S. J.1931. Explorations in Northeastern Arizona. Report on the ArcheologicalFieldwork of 1920-1923. Papers of the Peabody Museum ofAmerican Archeology and Ethnology, Harvard University, vol.xn, no. 1, Cambridge.<JuEBN8EY, S. J., and Kidder, A. V.1921. Basket-Maker Caves of Northeastern Arizona. Papers of the Pea-body Museum of American Archeology and Ethnology, HarvardUniversity, vol. viii, no. 2, Cambridge.Harorave, L. L.1930. Preliistoric Earth Lodges of the San Francisco Mountains. MuseumNotes, Museum of Northern Arizona, vol. 3, no. 5, Flagstaff.1933. Pueblo II Houses of the San Francisco Mountains, Arizona. Bul-letin 4, Museum of Northern Arizona, pp. 15-73, Flagstaff.1935. Report on Archeological Reconnaissance in the Rainbow PlateauArea of Northern Arizona and Southern Utah. Based upon Field-work by the Rainbow Bridge-Monument Valley Expedition of1933. University of California Press, Berkeley.Harrington. M. R.1927. A Primitive Pueblo City in Nevada. American Anthropologist, n. s.vol. 29, no. 3, pp. 2G2-277, Meuasha, Wis.Haury, E. W.1932. Roosevelt 9:6, A Hohokam Site of the Colonial Period. MedallionPapers, xi. Globe, Ariz.193G a. Some Southwestern Pottery Types, Series IV. Medallion Papers,XIX, Globe, Ariz.1936 b. The Mogollon Culture of Southwestern New Mexico. MedallionPapers, xx. Globe, Ariz.Hewett. E. L.1022. The Chaco Canyon in 1921. Art and Archeology, vol. xiv, no. 3,pp. 115-131, Washington.Hodge, F. W.1923. Circular Kivas Near Hawikuh, New Mexico. Contributions fromthe Museum of the American Indian, Heye Foimdation, vol. vn,no. 1, Hendricks-Hodge Expedition, New York.Hcx)TON, E. A.1930. The Indians of Pecos Pueblo, a Study of their Skeletal Remains.Papers of the Southwestern Expedition, no. 4, Department ofArcheology, Phillips Academy, Andover. Yale Press, New Haven.Hough, W.1903. Archeological B^eld Work in Northeastern Arizona. The Museum-Gates Expedition of 1901. Annual Report of the Smithsonian In-stitution for 19'^1, Washington.1919. Exploration of a Pit House Village at Luna, New Mexico. Proceed-ings of the United States National Museum, vol. 55, no. 2280, pp.409-43.1, Washington.Irving, J. T., Jr.1835. Indian Sketches, taken during an Expedition to the Pawnee Tribes.2 vols., Philadelphia.
KOBEHTS] ARCHP:0L0GICAL remains in ARIZONA 269JlE.\NCON, J. A.1922. Archeological Research in the Northeastern San Juan Basin ofColorado during the Summer of 1921. State Historical and NaturalHistory Society and the University of Denver, Denver.1929. Archeological Investigations in the Taos Valley, New Mexico, during1920. Smithsonian Miscellaneous Collections, vol.' 81, no. 12, pub.no. 3015, Washington.JlTDD, N. M.1924. Two Chaco Canyon Pit Houses. Smithsonian Report for 1922, pp.399-413, Washington.192.5. Archeological Investigations at Pueblo Bonito, New Mexico. Ex-ploration and Field-work of the Smithsonian Institution in 1924.Smithsonian Miscellaneous Collections, vol. 77, no. 2, pp. 83-91,Washington.1!)26. Archeological Observations North of the Rio Colorado. Bulletin 82,Bureau of American Ethnology, Washington.1930. Pueblo Bonito and its Architectural Development. Proceedings ofthe 23d International Congress of Americanists, pp. 70-73, NewYork.Kidder, A. V.1924. An Introduction to the Study of Southwestern Archeology, with aPreliminary Account of the Excavations at Pecos. Departmentof Archeology, Phillips Academy, Andover. Yale Press, NewHaven.Martin, P. S.1930. The 1929 Archeological Expedition of the State Historical Society ofColorado in Cooperation with the Smithsonian Institution. TheColorado Magarfue, vol. vii, no. 1, pp. 1-40, Denver.19.36. Lowry Ruin in Southwestern Colorado. Anthropological Series,Field Museum of Natural History, vol. 23, no. 1, pub. no. 3.56,Chicago.Maximilien (Alex. P.), Le Prince de Wied-Neuwied.1843. Voyage dans L'int^rieur de L'Amerique du Nord. (vol. 1 published1840, vol. 2, 1841, vol. 3, 1843), Paris.MiLLEE, C. F.1934. Report on Dates on the Allantown, Arizona, Ruins. Tree Ring Bulle-tin, vol. I, no. 2, pp. 15-16, Flagstaff.1935. Additional Dates from Allantown. Tree Ring Bulletin, vol. i, no. 4,p. 31, Flagstaff.MtNDELEFF, V.1891. A study of Pueblo Architecture: Tusayan and Cibola. Eighth An-nual Report, Bureau of Ethnology, Washington.Morris, E. H.1019. Preliminary Account of the Antiquities of the Region between theMancos and La Plata Rivers in Southwestern Colorado. Thirty-third Annual Report, Bureau of American Ethnology, pp. 155-206,Washington.1921. The House of the Great Kiva at the Aztec Ruin. AnthropologicalPapers American Museum of Natural History, vol. xxvi, pt. 2,New York.1924. Burials in the Aztec Ruin, the Aztec Ruin Annex. AnthropologicalPapers American Museum of Natural History, vol. xxvi, pts. 3 and4, New York.
270 BUREAU OF AiMKRICAN ETHNOl-OGY [Bvhu. 121Morris, E. H.1925. Exploring in the Canyon of Death. National Geographic Magazine,vol. XLviii, no. 3, pp. 263-300, Washington.Pepper, G. H.1920. Pueblo Bonito. Anthropological Papers American Museum of Nat-ural History, vol. xxvri, New York.
-Pbudden, T. M.1914, The Circular Kivas of the Small Ruins in the San Juan Watershed.American Anthropologist, n. s. vol. 18, no. 1, pp. 33-58, Lancaster,Pa,Roberts, F. H. H., Jb.1929. Shabili'eschee Village: A Late Basket Maker Site in the Chaeo Can-yon, New Mexico. Bulletin 92, Bureau of American Ethnology,Washington.1930. Early Ptieblo Ruins in the Pledra District, Southwestern Colorado.Bulletin 96, Bureau of American Ethnology, Washington.1931. The Ruins at Kiatuthlanna, Eastern Arizona. Bulletin 100, Bureauof American Ethnology, Washington,1932. The Village of the Great Kivas on the Zuiii Reservation, NewMexico. Bulletin 111, Bureau of American Ethnology, Wash-ington.1935. A Survey of Southwestern Archeology. American Anthropologist,n. s. vol. 37, no. 1, pp. 1-35, Menasha, Wis.1937. Archeology in the Southwest. American Antiquity, vol. ni, no. 1,pp. 3-33, Menasha. Wis.Simpson, J. H.1850, Report of an Expedition into the Navajo Country in 1849. In Re-ports of the Secretary of War, Senate Executive document no. 64,31st Congress, 1st se8.sion, pp. 55-64, Washington.Stevenson, Matilda C.1904. The Zuiii Indians: Their Mythology, Esoteric Fraternities, and Cere-monies. Twenty-third Annual Report, Bureau of American Eth-nology, Washington.Whipple, A. W.1856. Reports of Explorations and Surveys, to Ascertain the Most Prac-tical and Economical Route for a Railroad from the MississippiRiver to the Pacific Ocean. Vol. iii, Senate Executive Documentno. 78, 33d Congress, 2d session, Washington,WiNSHIP, G, p.1896. The Coronado Exi>edition, 1540-1542. Fourteenth Annual Report,Bureau of Ethnology, Washington.
INDEX
PsKeAiXANTowN, Akiz., ruius near 1Anasazi :burial custom of 14dwellings of 14use of the name 4-5Anasazi bemains, location of 5Animals of region of ruins 2Aecheology, southwestern, sum-mary of status of 4-5Ash pits, evolution of 255Ashes :pits for saving of 230, 231use of 200Aztec, N. Mex., reference to ruinat 218Aztec buins, mention of kiva in__ 247Basket Maker cultubb, develop-ment of 7-9Beads, discovery of mass of 202Bench:development of 257discussion of significance of_ 254of Kiva B 209ridges of plaster on 107-108Bins:purpose of 25stone slab 25Blocks, stone, purpose of, an-known 38Brown, Ralph D., assistancerendered by xiSee also Shelti.:b remains;SHEXTBaiS.Brush shelters :description of 56-57probable use of 57Burials :in and under refuse mound- 239in room 12 204in structure 16 164Mogollon 15
?of children 205-206of dogs 186,204of parts of bodies, occur-rence of 185-186of turkeys 230. 264offerings Interred with 264summary of 264under granaries 158unusual position in 204See also Skeleton.Bubhnell, D. I., Jr., cited 25,104, 122
PageBt tchart, Ruth R., work of___ xiiC.vcHB found in fill 123Cache ptt, jug-shaped, descriivtion of 178Carving, shell, mention of 14Casa Rinconada, mention ofkiva in 247Ceremonial chamber :evolution of 48,257See also Kiva.Chaco Canyon :pit struclture, date of 263reference to 218,233,247reference to masonry of 2.50Chettro Kettle, mention ofGreat Bowl in 247Clans, number of, suggested bykivas 249Climate, description of 4Compartment, stone-walled 89-90CooLEY, Harold E., assistancerendered by xnCorners, reinforcement of 199-200Court:description of 175-176Unit 3, description of 231-232CovEB STONES, usc of, for vcntlla-tors 19Cremation, practice of 14Cultural stages, determinationof sequence of 12CuMMiNGS, Byron, cited 233CusHiNO, F. H., cited 97Dance court:description of 126-128discussion of 258measurements of 129-130pos.sible use of 128remains covered by 124suggestive of great kiva 128Dates :covered by remains 259-263cultural lag shown by 262-2a3Deflector :example of 24incorporated in fire pit 75materials used for 41use of 19use of, In kivas 215Dendbooheonology :as an aid to archeologlsts? xndates obtained by 259-260discussion of dating by? 260-263271
272 INDEXDepbicssions in floobs : Pagecorresponding to kiva vaults. 107problematical 23, 64, KM-lOOunusual construction of 122Developmental Pueblo stage :discussion of 9-10remains belonging to 110ruins showing trend of 253-254De\^elopmental village, investi-gation of 17Dogs:burials of 186, 264use of, for food 176Douglass, A. E., scientific workof 13Douglass meh^hod of dating fromtree rings 30DWEUJNGS :construction of 14Mogollon, dating of 15used as turkey pen 118See also Structures; Sur-face houses.Enclosure, plaster-rimmed, dis-cussion of 236-237Entbanch! :descriptioui of 151measurements of 156ventilator replaced by 151Fenneman, N. M., cited 2Fewkes, J. Walter:cited 218i, 250perforated stones found by? 69Fill:cache found in 123conclusions drawn from 112-113, 118, 163in kiva, study of 238-239of Kiva A, examination of_ 184-185of Kiva B, significance of 220of Kiva B, study of 218-220significance of__ 30, 42-43, 50, 71-72significant features of 154stratigraphic study of 27-29, 93-96, 119-112use made of tests of 263FINKELSTE3N, JoE, assistance ren-dered by xiTFire pits :association of, with dwell-ings 172combined with ladder pit 49description of 37, 74^75, 87outside, use of 195perforated stones covering 71plastered rim of 49,69, 74, 87, 180secondary, discussion of 49
Fire pits?Continued. Pag?slab-lined 65surface, investigation of 176Floor :of Kiva B 21ftplaster ridges on 180-182See also Depressions; Pav-ing.Food supply, remains indicating- 264GiLMORE, Melvin R., reference to_ 264Gladwin, W. and H. S., cited? ISGranaries :Basket Maker 8construction of 100-101contiguous rooms suggestedby 147description of 134?135, 157-160, 163-164description of group of? 145-149development of 257measurements of 135,147-149, 159-160, 16ftor corner closets, descriptionof 203-204remains of 100-101significance in group of 159use of, as dwellings 158, 159See also Storage.Great kiva :dance court suggestive of? 128discussion of purpose of? 247-248measurements of 246-247Great Pueblo period:development of 10-11ruins representing 253Great Pueblo ruins :conclusions drawn from? 258-259description of 244-251Great Pueblo structures :dates of 252-253measurements of 247, 249, 251Gregory, H. E., cited 1Grubes, J. A.
:
acknowledgment to xiiinformation furnished by? xiHargravf; L. L., cited 110Harrington, M. K., cited 20, 57Harvard University, mention of_ xi, xnHatiry, E. W., cited 15,255Hawikuh, reference to archeo-logical work at : 71Hematite, use of, in kiva floor? 216Hewett, E. L., cited 247HoDGB, F. W., reference to workof 71, 214HOHOKAM
:
cremation practiced by 14cultural stages in 14
INDEX 273HoHOKAM?Continued. Paeodwellings of 14use of the name 5Holes and depressions:problematical 104-106See also Depressions.Hominy, use of ashes in prepara-tion of 200HooTON, E. A., cited 239Hopi:cited 57, 258reference to kivas of 21SHough, Walter, reference to in-A'estigations of 70Irrigation, progress in 14Irving, J. T., cited 25Jeancon, J. A., cited 108,233,258Jones, David, work of xiiJuDD, N. M., cited 57,108,182,218Katcina Kihu:explanation of 212niche compared with 68Katcina niches, discussion of 212Kidder, A. V.. cited 218Kimball, Solon T., assistancerendered by xiKing, Dale S., assistance ren-dered by nKitchen, outdoor 205KivA A:description of 180-188discussion of date of 187-188discussion of status of 186-187masonry in 180measurements of 191-194superstructure of 180KivA B
:
description of 207-218measurements of 223-226KivA, Unit 3:description of 232-238measurements of 242-244KivA VAULTS, depressions corre-sponding to 107KivAS
:
Chaco, feature suggestive of- 107number of clans suggestedby 249timbers in wall of 233See also Ceremonial cham-ber.Kneipp, Robert, work of xiiLaboratory of Anthropology,Santa Fe, work of xiLadder box ;change in function of 21combined with fire pit 49development of 255
Ladders : Pagekiva, placement of 215pits for 21,49two-pole-and-rung type 21types of, in use 255LowKY RUIN, mention of kiva at? 247Luna, mention of dance pit at 258Martin, P. S., cited 20, 247Masonry :of Great Pueblo structures,. 250of Kiva B 209superior grade of 232-233use of, in pit houses 33-34See also Walls.Mesa Verde, reference to 218Meh'ate and manos, found in po-sition 139MiLua;, Gael F. :assistance rendered by xicited 110, 142, 239work of XII, 30Mindeleff, v., cited 55, 122, 198Modified Basket Maker houses :feature of 25horizontal masonry in 34Modified Basket Maicer period:discussion of 8-9traces indicating 253Mogollon :correlation of, with Anasazi. 15-16correlation of, with Hoho-kam 15-16cultural stages of 15use of the name 5Mogollon culture, character-istics of 15MoRiiis, Earl H. :cited 108, 218, 247, 248reference to work of 128NioHE
:
as a feature of kivas 217-218between rooms 202kiva, discussion of opening to_ 218suggestive of Katcina Kihu- 122See also Katcina niches.Nusbaum, Derio, assistance ren-dered by xnOven. See Pit oven.Peoos Classification, adoptionof 6Paving, remains of 122Perfor.\ted stone slabs, discus-sion of 69-71Periods, culture, classification of_ 6-7Piedba River, mention of danceplaza on 258Pilaster :construction of 50
274 INDEXPilaster?Continued. Paaeevolution of 35-36examples of developmentof 254-255use of 35Pit dwfxlings, structures com-parable to 19Pit ovein, description of 55Pit remains:D-shaped 46,58,102description of 21,58investigation of 20kiva suggested by 108significance of till in 30,42Pit structures :changes made in 257conclusions drawn from 168-169, 254connected, discussion of 80-81connected, group of, de-scribed 67-80elaborate interior featuresof 137-138partially superimposed 62theory concerning combina-tions of 255three forming one 81three-iu-one, discussion of 96-98transition in function of? 256-257See also Pit dwklijngs ; PitREMAINS ; Structure No. 1_ 18POPUUITION :estimated, of Great Pueblo 249,250estimated, of Unit 3 239-240estimates concerning 264-265Portico, description of 174Posts, bracing of 35,47Pot rest:description of 163in Kiva A 183Potsherd material, stratigraphictest of 27-28, 4lHi2Pottery:cultural stage determined by- 12method of making 14-15Mogollon 15Sec also Potsherds.Prudden, T. M., cited 218Pueblo Bonito:date of 263mention of kiva in 247Pueblo communal buildings, de-velopment of 97Pueblo culture, development of_ 7, 9-11Pueblo Grande, Nbivada, refer-erenceto 20
PUKBLO-TYPE STRUCTURE: Pagedescription of 248-249See also Pueblo communalbuildings.Primco River, discussion ofstream bed of 2-3Reed, Erik K., assistance ren-dered by XIIRefuse mounds :conclusions drawn from 252examination of 251-252Great Pueblo, discu.ssed 251-252location of? 172, 194-195, 227, 245purpose of trenching of 20RPGBESSivE PERIOD, causes of 11Remains :southwestern, divisions of__ 4-5See also Ruins.Renaissance pebiod, developmentattained in 11-12Ridges, adobe:compartments outlined by 3J)on bench 107-108ItoBEitTs, Frank H. H., Jr., cited- 6,12, 19, 20, 25, 56, 57, 67, 69, 107.127, 182. 204, 207, 214. 218. 247.248, 250, 255, 256, 258.RoiiKRTS. Linda B., work of xnRoof, cribbed:explanation for 49of kiva 235presence of 46significance of 48Roofs:construction of 48kiva, cribbed type 235Kiva B, description of 209-211use made of 137See also Roof, cribbed.Rooms:in tiers, description of 248-249of surface structure 197-206,229-231rectangular block of 248-249steps in 197,198,201,203Ruins :location of 1periods represented in 15time covered by 259unnamed xnSelenite, use of, for windows 174Shelter remains:conclusions drawn from 133description of 130-132,147, 176-177, 206measurements of 132-133.149, 222-223
INDEX 275^Shelter remains?Continued. paueproblematical feature of 131question concerning 207Simpson, J. H., deductions drawnfrom diary of 3Sjpapu :jug-shaped 116occurrence of 40, 49rectangular example of 75symbolism of 23Skelkton, parts of, in Kiva A- 184-185Slabs, stone:imbedded in floor 38, 39possible use of 38use of, in pit houses 33-34Soil, quality of, around ruins 2Stewart, T. D., material assignedto 264Stone, perforated, use of, toframe aperture 69Storage:wall pocket for 116-117See also Granabiks ; Storagebins; Storage cist; Stor-age PITS ; Storage boom.Storage bins :description of 200-201of Kiva B 217Storage oist, description of 100-101Storage pits :discussion of 178-179jug-shaped 232Storage room :description of 201See also Granaries.Storage structure:beehive-shaped 140See also Granaries.Stratigraphy of fill, conclusionsdrawn from 96Structure No. 1
:
description of 21-30measurements of 30-33Stuuctuise No. 2dates of 43description of 83-46measurements of 43-46two occupations of 34-35Structure No. 3:dates of 48,52description of 46-53measurements of 52-53Structure No. 4description of 58-62measurements of 60-64Structure No. 5a
:
description of 62-65measurements of 65-66
Structure No. 5b
:
Page-description of 26-65measurements of 66-67Structure No. 6
:
description of 67-72measurements of 72-73Structure No. 7:description of 73-75measurements of 75-76Structure No. 8description of 77-78measurements of 78-79^STRUcruRB No. 9
:
description of 81-84measurements of 84-85Structure No. 10:description of 85-87measurements of 88-8&STRUCTtTRE No. 11description of 89^1measurements of 91-93-Structure No. 12:dates of 108description of 102-110kiva suggested by 1 108measurements of 113-115STRUCTtTOE No. 13a
:
description of 116-120measurements of 118-120Structure No. 13b
:
description of 120-123-measurements of 123-124Structure No. 14description of 124-12&measurements of 128-129Structlire No. 15
:
date of 141-142description of 135-141measurements of 142-145Structure No. 16description of 149-156measurements of 155-156relative age of 155unusual entrance to 151Strtjcfure No. 17
:
description of 160-164measui'oments of 164-166Structure No. 18description of 166-169measurements of 169-170Structures :connection of 19-20construction of 18general characteristics of? 18-19interior features of 19opening in wall between 75problematical 250-251Pueblo-type 97, 248-249
276 INDEXStructures?Continnod. p?geunidentified, remains of 177use of ttie term 17See also Ceremonial cham-ber ; KivA ; Pit struc-tures ; Shexters ; Super-structures ; Surfacehouses ; Surface struc-tures.Superstructures :change made in 257construction of 104, 137description of 27, 102-1(?4discussion of 255-25()of Kiva B 209-211of three-iu-one pits 97-98restoration of 27Surface houses :description of 53-56,98-100measurements of___ 57-58,101-102Surface structure. Unit 2 :construction of 196-1{>7description of 19(5-206measurements of 220-222rooms of 196Surface stkucture, Unit 3:description of 227-231measurements of 240-242Thomas, Sidnetst J., assistancerendered by xilTopography, of region of ruins__ 1-2Towers, discussion of 250-251Turkey pen, remains of 118TuTtKEnrs
:
burials of 230, 264domestication of 14Unit No. 1:description of 171-194surface building, measure-ments of 188-191Unft No. 2:description of 194-226number of families in 196Unit No. 3
:
dates of 239discussion of 227-240measurements of 240-244Unit-type dwei.ung :evolution of 256-2.^8time involved in evolution of_ 260Unit-type structures :discussion of 170-171general description of 171University of Arizona, mentionof XIUniversity of Denvek, mentionof XI
PageIlMvERSiTY OF MINNESOTA, men-tion of XIII'NivFJisiTY' OF Oklahoma, men-tion of XIIUniversity of Texas, mention of_ xiiVentit^tor :construction of 50description of 19,7.5,83-84,108development of 257ut kiva, description of 183, 237origin of 25remodeling of 38-39subfloor type of 212-214imusual construction of 126unusual features of_ 26, 71, 139-141Village of the Great Kivas,dates of 252, 2a3ViLL.'i.GE REMAINS, Change re-corded by 171Wainscoting of walls, discus-
.sion of 233-235WAix niche:possible explanation of 68purpose of 83See also Wall pockejts.Wall pockets :for storage 116-117,162in Kiva, use of 180Walls:construction of___ 173, 196, 19&-199horizontal masonry in 196rubble finish on 196timbers incorporated in 199, 233wainscoting of 233-235Watercoup^e :buried 172, 195diversion of water from 172Water supply, discussion of 3-4W^EAPONs, mention of 15Whipple, A. W., cited 8Whipple Railroad Survey, refer-ence to 3Windows, discussion of 173-174WiNSHip, G. P., cited 2.33ZUNI
:
culinary practices at 200granary found at 204reference to Great Kivas of- 218reference to storage bins of- 198ZuNi KIVA, perforated slab usedin 69ZuNi Reservation :baking pit found on 56mention of kivas on 247reference to masonry of 250o







nil
i\Vliiiil
SMITHSONIAN INSTITUTION LIBRARIES
3 9088 01421 8655
11.m iliiiiii
Ml