Aa hth yp UiaebokGan aa ay eal iy f Att ie ie ; HY aN is i. A aati staf Hig CY bs on, hy Ke is Lh at ; igi i ia 4 : te i it 4 ‘ al = pretty = p —- eine oe = Lean ane See Sees pein Tao nen Sass. SIS ie SSS ees =< a Se = SS eactes ae rts Sees Seer eee BAM) i iy nh if iw iin ‘ : iat Be afar Mae cae eee Setteres fa Ws E em piend 2 Sen is i i) H rt #4 a ) Aiea PORN fie Hid ¥ 1a dues an sen Hes pian ti Uy 7 Vat f ‘ i] Ne ns tie he ts io SMITHSONIAN INSTITUTION BUREAU OF AMERICAN ETHNOLOGY BULLETIN 126 ARCHEOLOGICAL REMAINS IN THE WHITEWATER DISTRICT EASTERN ARIZONA Part IJ. ARTIFACTS AND BURIALS By FRANK H. H. ROBERTS, Jr. WITH APPENDIX SKELETAL REMAINS FROM THE WHITEWATER DISTRICT, EASTERN ARIZONA By T. D. STEWART fa % (> ave 30 1940 4) Won a, UZ SMITHSONIAN INSTITUTION BUREAU OF AMERICAN ETHNOLOGY BULLETIN 126 ARCHEOLOGICAL REMAINS IN THE WHITEWATER DISTRICT EASTERN ARIZONA Part IT. ARTIFACTS AND BURIALS By FRANK H. H. ROBERTS, Jr. WITH APPENDIX SKELETAL REMAINS FROM THE WHITEWATER DISTRICT, EASTERN ARIZONA By T. D. STEWART UNITED STATES GOVERNMENT PRINTING OFFICE WASHINGTON : 1940 a ae ee A eee ee tl, For sale by the Superintendent of Documents, Washington, D.C. - - - - - - - Price 50 cents \ A i rane ey 4 ‘ : A eT 4 PM eh ay A ae ta Cae at ihe ie iden LETTER OF TRANSMITTAL SMITHSONIAN INsTITUTION, Bureau or AMERICAN ETHNOLOGY, Washington, D. C., June 10, 1939. Sir: I have the honor to submit the accompanying manuscript, entitled “Archeological Remains in the Whitewater District, Eastern Arizona, Part II. Artifacts and Burials,” by Frank H. H. Roberts, Jr., with an appendix entitled “Skeletal Remains From the White- water District, Eastern Arizona,” by T. D. Stewart, and to recom- mend that it be published as a bulletin of the Bureau of American Ethnology. Very respectfully yours, M. W. Stiruine, Chief. Dr. C. G. Axppor, Secretary of the Smithsonian Institution. HI MATVIMGZAZART FO ANTIMIT - Nia ‘ “ae Rely, ie ; NTOTTEAL Warnor wre | 1 ae oon waouius «0 saga 1k RRA OL sane 9 AOSQiitenW, —. dqiroan rece gaiyiaginospn, oft sintdya of srortod, out ia grates Joiteil tolevotidW oft nt aninmoll Inoigoldodot A F pirodoll, FE daar qd “eleisuf bas atontittA TE 44 aid W oly mov aninaroh Lateled@® boltitaa xtheaayp § > wto9 et o) baa diewote oT yd “yarosinA stolen , er an nscknoarh to weg oft Yo aitelind: Bek ae a . i baat haste omar W Mi CONTENTS PAGE LEROY WENA OD TO be ats 2a) ap bt a De re xI Introduction sess ee ners Try Hae Lee a Pee Sve 1 A RSVME WLP D ILS A. fal gla ce 2 LS LTE 0 TR OO 3 JECT Leia sk ELIE ayes Ee Ree She ad 2 2c SO ee ge ee ee eo ee ee nen 3 Designsion the paintedivessels. 22-2 22s Sie. US. See ue 2 52 Additions ObIEcssioiMmred Clay. oso oan ewe Seeds 109 Benes Ane epenewOrk. . = sees SUE Cer ee CR ee 112 LOOT TE LS] GVEA 0) 1s ate a ee Ce a nO DO em Ue NE pe cnet I ee Seep uly Objectsiolishell=s hal iawie By Seee Mee feu eee 130 Eisenia seem See Me eel Replies. egies Cael _ 2 dele 132 SUM sry ANdKchSGliGhiIOne NOLAN os SG os ee eerste sek ee 136 Hniterabureieited ames, . shay fel Eye AN. 2 Raat oe ey 141 Appendix A. Measurements and finding places of objects shown in illus- gfe 4 (a0 -1 Se ape gk pA a MB ARERR SCE Lok a oe Aa a Pe 145 Appendix B. Skeletal remains from the Whitewater District, Eastern TIZOn a byl ee Ste Warbn= ae it Se eis ee i Se 153 Methodsswe se. Sr. 2 teal. 5 ere B el tee sobs ate i Eye 2 Bea 154 ATP ss spa ea i pea a ity re a a ee aaa a DRI ap We 155 Phe rsmeleton Meee Meet Nettie Zhe RS ak nee ee eek eee es 159 RUBEN Shreya pe ame le enter iy e N Re 161 Whigerature cites set i wy Dhl Ses Pippi eh ahs at va Rae tay 166 ancl exsemee ares emia dar lange, mtn a 53 ae a Ne en 167 Vv 4 iad & - prt Se } i ee sn ee et ene en ene ee ees Pe ree me Pia can see ke ie he a ge ad 0d ev oon ws on i wi ie we ee ee ee ae > hipaa ad ih Fe te ae dialer th ten ee me NN aT ete a ge tm fee te ee ee ee ee ee er lecige eet Wt Te PM Lee Han rer er re . BS ae ce i ha ee ae ui 4 1 vel ERA ai PP BORO, RPE oe eRe NITIES -eseeeed nant al pid ay a tae ao ok ot bye hel OPE ape 2 ' te ye Ween ws Slain ‘aso sl tty Ay By het wn TPA dak Thm focal aF ony LAT Rad nade La TL las vo a pie Hp bergen well ya A ah ges Rang ete ie ih at Mp : ze aL Siete eg eee ORME BE Ue A DOR a, We diy aie a: el ba ht den le Ne wali ar awouls ainonie to. pact. che bev vhitoier mio) oh , don area eheealc ee ee en ont vada \gotuinidl xodaswatist ; eS oe eee od ak ae NF te fa eh ad ly pt a Resatoea aaron es ee Die LB | ip be ana te Bip ate ano ~ ~ ” - ae Cokes < i ig ee Aw bees vite ay' o-dalen | pe ke a tae pate ey ie key hig ent yl oe FR os RR aks CL ‘a 1 x 4 yeh i yey) ~ Bt, te OA Aree 4 Pad = % & Te) ee ee ee ee ee ec | . : ‘ 2 Pi ee . - wie se MOR Figciee teeth, Le RN, ok tortor dois coe Da ei bh At le i ee os Sail od A ae ae is Ll ieee bi (hae yon ny ap ae ee a ee er pi = 5b ean RT SL En wt oan re i ¥ es ie ILLUSTRATIONS PLATES . Typical potsherds from beginning Developmental Pueblo wares. Those in left-hand column are Chaco I; right-hand, Kana-a black on . Plain wares from Modified Basket Maker and beginning Develop- ANETTA LETS LO PMO TIZO TIS ess cee ee aaa ee a . Banded-neck pitchers from early Developmental Pueblo horizon _ _--- . Developmental Pueblo banded-neck pitchers_-_-__---_-.------------ : Developmental Pueblo pitchers]! 2 kee "" 22 $2 eer oe Le Sees . Corrugated pitchers from end of Developmental and beginning of Sie S42 red SA CEN Oey GY EN a 00 kee a i, ell pay WO ali yaaa ate cha . Late Developmental Pueblo and early Great Pueblo corrugated Nia sielelal See aed etd al aed pte ip elie Ir il ber gehen gene wy incl eglae ia Oy Eo NP Cari Li AN . Bowls made from bottoms of broken culinary jars_________________- . Miniature vessels from Modified Basket Maker and beginning De- VAG opamreran eee hil Ave oy Koy Voy VAD) AVS eee Be cel sa aN se a . Nonculinary pitchers. a and b are trilobed body type; c, d, and e, Variations Of Courd-suape Vessels. 252 2: 22 oo ees eee ae . Pitchers and canteen from the painted wares. These are examples of themlhittler@oloradouSeries ees oe eee agra a a . Various forms of black-on-white pitchers______---_________---_---- . Black-on-white ware pitchers. a and b, Middle phase of Develop- mental Pueblo. c and d, Closing stage of same period_-_-__-__-__-_ . Bird-shaped and ring-bottomed pitchers__.___..._.__.______--------- rblack-on-white: ware pitchers.2 2. ose bat ooh et et a nik . Late Developmental and early Great Pueblo black-on-white pitchers__ plaek-on-white ware Canteens. 20 22.23 t 22 eed St fee ks ESCOOD bYPES ss reas fi Eee A Pa a Oe ey oe Oe Peba Me wOrmst es Sau e) ot cee ee Be ta Se NE es Ley. . Deep bowls. a, Modified Basket Maker bowl; b, Great Pueblo bowl-_- . Three types of exterior surface treatment on black-on-white bowls-_-_-_ . Specimens from late Developmental Pueblo phase. a-—b, Handle from scoop? .¢, exterigniol bOWlet too oe ento et oF Be 8 eB ae + Decorations onpniteriors, of sbowis.25 +04 2h? 420 ee ee | ek SUESO WI COSI@ TINE oe eee eee OS oes Cre Mt Ri Ey t Black-on-white: bowl interiors2> =. 202 eee to eet ee », Band desiens on interior.of bowls ..32 2: 2 See 6 pe ee ee PT .) olack-on-white ware DOW ot th a hk aid ch es Se » Different styles’of bowl decorations. 8 42 ak . Bowls from end of Developmental and beginning of Great Pueblo PAGE 20 20 20 20 26 26 26 26 36 36 VIII ILLUSTRATIONS .» Clay and) stone pipés and ‘stone cylinders: ...._- 2-2 22 eee . Awls made with little modification of original bone________________- . Awls made from bones partly prepared for the purpose__-__________ Awis of the fortuitous splinter group. =/_.--.2. 252555 ee . owls and bodkins.._. 2+ 2 ys2o. ee 2 ae eee ep pee oe ee . Needles, whistle, bone tubes, and’gaming*pieces___________________- Sy DOne-scCraping (OOS. 62 2 2. ceterecreree| ie Ue ee MevViCtatesiand Manos: 222 2 2822S ee PESUOKELAKOd 2 02 2. Sek oo bk SR Se TPACKeS IN aU aN CNnOR 32 28 ee a akg ae . Ax, maul, and hand-hammer stones_ «+ ~~. --- phe So ga moconerpaletves-2- 2622.22. 2-225.5 3 eee eee USMOOLNING SEONCS or Se yin Re ta Io ee a ee + Knite blade ‘and. projectile points._._____--=-2 a ey) pe ee » Knives; drill, and, projectile, pointe. (4 ee eg small objectaiot atone... 22.2 aa eo te ee . Shell and stone ornaments, and clay pipe with an inset of turquoise___ PASUQUC DPAGS oes 9 Pa te ee ge ee oe . Collar style of necklace. Inset shows specimen in situ before removal_ . Lypical burials, a, Young ehild, 6, Adulto.) yes 525 0 2s op ee . Four views of deformed female skull, U.S.N.M. No. 367839_________ . Four views of deformed male skull, U.S.N.M. No. 367851_._________ . Four views of deformed male skull, L.A. No. 50-3__-...---________-_ . Four views of deformed female skull, L.A. No. 50-4.._....._______- . Four views of undeformed female skull, U.S.N.M. No. 367848___.___ . Four views of undeformed male skull, U.S.N.M. No. 368075________- TEXT FIGURES . Modified Basket Maker culinary vessel shapes___________________ ane . Developmental Pueblo banded-neck culinary vessel shapes__________ . Corrugated culinary jar shapes of late Developmental and early Great Pueblo phases: ===+2=-s22=22 2277280219 OOMOVIOCRIT DUG IYRIREG-N . Modified: Basket Maker jar shapes----=-.-2 222 SL 22 oe 22 eee . Developmental Pueblo jar ‘shapes_t Y690/ Vi180 DUR LAr POMQuio yon BF Great-Pueblo jar shapes: =2=-s222222222 2200009 918 SIO TO-a08 Design on large black-on-white -jar_._..__________- 2-22-22 . Design on late Developmental Pueblo jar____-_______________=_____ . Decoration on early Developmental Pueblo pitcher of the Little Colorado series 202 40 BO. TQ OW RST GIRS TOS 10 800 1) 90% . Design on Developmental Pueblo pitcher of Chaco type_____________ . Pitcher design from middle Developmental Pueblo phase____-________ . Decoration on pitcher of Little Colorado series from middle Develop- mental “Pueblo spa west --hn.0 = se erin ees ne by Desionrom bird=-shaned pitchers] 2 eae es Ee ee . Middle Developmental Pueblo pitcher decoration___________{_2_ _! . Decoration on ring-bottomed pitcher, side view__2_________________ -) Lopiview of rime=bottomed-piteher=* Seon es 22 es eee ee . Chaco series middle Developmental Pueblo pitcher design___________ . Decoration on Little Colorado pitcher from middle Developmental — Rueblovetacoss* 39" rke avs «Mop shite nee sos wee 6a eee . Heavy element type of design from latter part of the Developmental 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41, 42. 43. 44. ILLUSTRATIONS IX PAGE Early style of canteen design, Chaco type series____________________ 75 @haco style of desimurom-canteene 0500222 Ne ee 76 Kana-a style’ of decorationjon canteen... 2225-2 fk Lk 77 Little Colorado form of canteen decoration_--_____-__---------___- 78 Design on opposite side of canteen illustrated in figure 23___________ 79 Portion of band design on body of canteen__.__________-.---_---_- 81 Pattern on opposite side of canteen illustrated in figure 25__________ 82 Heavy element style of decoration on canteen from the end of the Developmental tucblo period. i222. se ee ee lk 83 Top view of canteen illustrated in figure 27.-__........-..----.-_.- 84 Solid and hachured type of decoration on canteen_________________-_ 85 Hachured meander motif on top of vessel illustrated in figure 29_____ 86 DECOTRTIONNOMUSCER gt ee fe ee Re ba Be ae chk ub 87 Westpnioniseedeiar eters. ose a Se bel a ae eee he 88 Wecoranon OnmMitenor, Of SCOOPS ==" 2s. S525 seo oo eee see he 89 Pesign’on interior of cap-ty pe Scoop. 5. == 5. Se ee 90 Decoration on half-gourd style of scoop_.-.......-.-------------_- 91 Design on interior of late Modified Basket Maker bowl____________- 92 Pattern on early Developmental bow] of the Chaco-type series______- 93 Quartered design in early Developmental Pueblo bow]________--___- 94 Concentric figures on interior of Developmental Pueblo bowl-_-_-_--__- 95 Triangular units in decoration on bowl of the Chaco series_--____-___- 97 Quartered style of decoration on early Developmental bowl_---_-__-_- 99 Band pattern on Kana-a black-on-white bowl_____________________- 101 Kana-a style of bowl decoration from early Developmental Pueblo jOVET LO leg ech eet Ne II ei RAS ERY Se 0 aS AR OO CAN Hs Bethy Le 102 Band design on bowl of Little Colorado series____________________- 104 Penne ts. arte Or. canada ed 5 ey. Paes Pl peed eee: To) SOO pitt Dane 4 ANOS SEU Ce Oh NOE grey fees uy SOR AR St eT I Om EMI CaaS EA SCRA 3 el ay a Liban ci ia Lhe a mn rm Wed aoa ay Abs Lhe | vith, FOREWORD The nature, location, and general cultural status of the archeological remains in the Whitewater District in eastern Arizona, with specific information on the house and village types, was presented in Part I (Bur. Amer. Ethnol. Bull. 121) of this report. Further considera- tion of those features is not essential to the discussion of the lesser phases of the material culture appearing in the following pages and will not be repeated here. Readers desiring additional data on char- acteristics of the stages of the Anasazi or Basket Maker-Pueblo pat- tern indicated by the collections described in this volume may refer to the proper sections in Part I. Discussion of the ceramic types and variations in the several forms of pottery is based for the most part on an archeological approach to the subject rather than on any of the methods of laboratory examina- tion and analysis that have been developed in recent years. Most of the work was done before those systems of testing attained their present state of elaboration and intricacy. While the results unques- tionably fail to furnish much of the minutiae deemed indispensable by the ceramic technicians, they nevertheless shed some light on stylis- tic trends and changes in the local industry and give evidence of an interplay of influences spreading from several centers of development. Names for some of the types and subtypes occurring in the White- water District had not yet been formulated when this report was pre- pared and as a consequence do not appear in the text. Where it is possible to correlate some of the Whitewater forms with those de- scribed and named in some of the recent publications, the relationship is indicated in footnotes. There undoubtedly are others that do belong to sundry categories now identified with a specific designation, but the lack of accompanying illustrations in the descriptions of numerous forms makes recognition uncertain and for that reason men- tion of the names is omitted. The collection of specimens was distributed after the completion of the work. Some went to J. A. Grubbs of Houck, Ariz., some to the Laboratory of Anthropology at Santa Fe, N. Mex., and the remainder to the United States National Museum at Washington. The dispo- sition of the articles illustrated in the plates and text figures is indi- cated in the tables in the appendix. x1 qaowEnon oa ¥ a ne rtm it me op nome Deena oe Ro abate Lerudisr lei: bai Neal a cat a ftir ona irtevenn ai tointei sama 9 ail ges ean inl nomic tipsy vil iti animoqqe rapa babewiarn wilt to / ‘tatty fo sieh Isnonibbs guiviesh ebssl .ered botseqor od « dag oldoyT-godeM jodent to ixeaumA ali to vogete orld toe ps sakes Wears agate eid? mf beditozob atoltsalios els yd baiaa ar \ Faw iti antoiings 14 ae atttot Taras of} of anolisivey bas esqvt ofan od To mo ie DAL NOS ‘nd is ee fanloeiiots a fo ds sas ald x08 huand ei ; bolt 2UH8Y, Redes at rman ck sa puke jc ainylnne baw: tied bokeite oiritea) to amateye sxodkt soled omob- an irs q Soupni: cilreot odd oid Wo -qonsiniol bis sotterodsle Yo otate th Aldexnagethat boaweb ovitunion odt to domes deiaust ot [tet y >) seilyde ao detail arcoe bode eeslodtavon yardd ectoisindoad ofenm199 # » th lovsenabtro evig has qiesbat Lwool alt mi eogrtedo Des ebnoss - eenqgplevab te esti Inaavas enortt gaibnorqa eansrontiat to valor tA aris at gnirmoso eoqzidue bis esqyt add to anos tot 29 ) -otg enw dogs aidd vothw boislsartot aeed toy tom bed jotta ei tt ored WT .tzod oft ai ae9qqe tom ob gomatipoetos © en its b ve bb aeodt div sartet toievoitd W odd to emoe otelorios ote ) qislesotialor oft nivoitesifdisey teooor od) to onmoe at boran bas ob fal) eredio oi yiboidvebay stoff edtonioct ai. hates “ froilntrgiaah sitionga « ditv belitaebi wou asitogeine yibaaoh ¢ to anptiqitesb of} wi ancitettianlii yatyasquwoor to Aouk | J leah a is “ot hice wishtoons noisimgooss codeat enrtot ail BRE he SS ballin +i coment off ¥ potbletes ott ratte’ basndinteib-e0w atommisoqa to aofsellas orlt of antox .xitA lonolt to addireD .A Lot dow omto® olsow: yabaiscrotodt bie olf 7 9 rinse in yeoioqordtaé to V0! soqeth off LaoiynidenW ts amon Innotte 2oint2 bodialD “ies at soa tot ne eonley adi oi hotwrisu!lt 2aloitie oft Yo sibooquys adtat eoldnt sdt af § ARCHEOLOGICAL REMAINS IN THE WHITE- WATER DISTRICT, EASTERN ARIZONA Part II. Artiracts AND Buriats By Franx H. H. Roserts, JR. INTRODUCTION The artifacts recovered during the digging in the several sites investigated in the Whitewater District consist of pottery, additional objects of fired clay, bone and stone implements, minor articles fashioned from those materials, and shell worked into objects for personal adornment. These specimens are all of durable materials and represent but one phase of the arts and industries of the people. That part of the material culture composed of things made from per- ishable substances is practically missing. A few potsherds bearing basketry impressions and several pieces of charred cord found in one of the structures that had been destroyed by fire are all that survived to indicate the existence of this group. The imprints made in the damp clay of vessels and fixed by the firing process show that the basketry was of the characteristically coiled Anasazi form. The cord is of two types. One was made from cotton, the other from apocynum fiber. Because of the charring and small size of the fragments, identification of species is not possible. Neither can their use be determined. From materials obtained in the excavation of dry caves, however, it is known that both apocynum and cotton were employed in the making of various articles. ‘There is no evidence on the kinds of clothing, textiles, or wooden implements that unquestionably played a prominent part in the daily life of the occupants of the district. Hence the collection does not accurately portray the handi- work of the period. It gives only a one-sided picture and in consid- ering the degree of development indicated by the specimens this fact should be borne in mind. Actually what remains is only a minor proportion of the total and there is much, unfortunately, that can never be known about the arts and industries. The specimens in the collection were found in the remains of houses, in the various refuse mounds, as offerings accompanying burials, and, 1 2 BUREAU OF AMERICAN ETHNOLOGY [BULL, 126 in a few cases, were lying on the old surface of occupation near dwell- ings and outside fire pits. There is no discernible difference between those found in graves and those from other locations. Consequently it is apparent that no articles were made especially for funerary purposes. Those interred with the deceased were either personal be- longings or part of the ordinary household equipment contributed as a final offering. Although most of the artifacts represent one stage, the Develop- mental Pueblo, there are a few from two others in the Anasazi pat- tern, Modified Basket Maker and Great Pueblo.t. In order to make clear the period differences and facilitate discussion, the various ob- jects comprised in each of the several groups will be considered together. Thus all of the pottery will be included in a single section, such characteristics and variations as have a bearing on the stage represented being indicated in the descriptions of the forms. The other specimens will be treated in a similar manner, although there are not as many period differences to be noted as in the case of the ceramic group. Inhumation was the prevailing method for disposing of the dead. No traces of cremation appeared in any of the digging. That prac- tice apparently had not been introduced into this district at that time, even though it was in use in regions farther south and west. Most of the burials were in the refuse mounds accompanying the house clusters. A few had been made in rooms in some of the dwellings, some were in or beneath the pit portions of granaries, and others were in shallow pits scraped out of the natural surface of the earth. Several body positions were noted in the series as a whole, but most of them were some variant of the general contracted form found throughout most of the Anasazi province. An interesting and puz- zling feature was the small number of graves attributable to the Great Pueblo stage of occupancy. While no excavations were carried on in the ruins of that period, sufficient work was done in the refuse mounds and area adjacent to the buildings to reveal any series of burials possibly present there, yet less than a half dozen were located. Furthermore, run-off waters, following heavy rains and from melting snows, cut channels through one of the dump heaps associated with the main ruins and exposed large areas of the deposits without reveal- ing evidences of interments. Hence it seems that the dead were either transported some distance from the village before being buried or else were disposed of by some other method. This lack of human remains for the Great Pueblo horizon corresponds to a similar condition in the Chaco Canyon in northwestern New Mexico and at other sites 1See Part I (Bur. Amer. Ethnol. Bull. 121), p. 253. ROBERTS ] ARCHEOLOGICAL REMAINS IN ARIZONA 3 where the cultural pattern shows definite relationship to the Chaco center. As previously mentioned, in discussing the possible affini- ties of the Whitewater group, there was a strong Chaco influence in this district and the absence of Great Pueblo period graves is likely due to that factor. Most of the bones were in a poor state of preservation and as a consequence only a small series of remains is available for study. They indicate a number of interesting things, however, and Dr. T. D. Stewart, assistant curator of physical anthropology, United States National Museum, points out their significance in his notes on the skeletal material. THE ARTIFACTS Porrery True pottery first appeared in the Anasazi province in the Modi- fied Basket Maker stage of development, in fact it is one of the traits that help to differentiate the period from the preceding Basket Maker.* At about the end of the latter horizon, possibly in the tran- sition from it to the Modified Basket Maker, vessels were occasionally made from unfired clay that had been tempered with shredded cedar bark or grass. These were the prototypes for the later wares that came to play so prominent a part in the industry of the people. A combination of methods was used in the manufacture of the mud vessels. In some cases the bottoms were molded in baskets, the walls or rims rising above the mold being formed of fillets of clay that were looped around the upper edges. Each strip made a single cir- cuit of the perimeter and was welded or pressed to the one below while the clay was still moist. Others indicate that they were fashioned entirely by hand without the aid of a basal support, the bands of clay being employed throughout in this early form of the coiling process. Principal shapes were large, shallow trays and deep bowls. Additional articles of unfired clay also appeared at about the same time. They consisted of crude figurines of human females, models of bifurcated baskets of the type made by the Basket Makers, and small nipple-shaped objects that may be representations of carry- ing baskets to be attached to the backs of figurines but whose actual purpose is unknown.* These also were prototypes for more elaborate forms in subsequent stages and with the development of true ceramics they too were fired. The complex of unfired clay objects may be attributed to influences reaching the province from regions to the South. Indications are that 2Part I, pp. 251-252. 3 Part I, pp. 8-9. Guernsey, 1931, pp. 84-88; Guernsey and Kidder, 1921, p. 98: Morris, 1927, pp. 138-160 ; and Nusbaum, 1922, pp. 138-144. 4 BUREAU OF AMERICAN ETHNOLOGY [BULL, 126 the manufacture of these articles resulted from an impulse of ideas rather than froma copying of actual objects carried into the area, There is good evidence to support the belief that the methods used in making the objects were mainly indigenous and that the entire subsequent development of the industry grew out of local experi- mentation and effort. The Basket Makers were adept in the working of clay because they made extensive use of it in all forms of archi- tectural construction, particularly in their granaries, and it would be a relatively simple matter to transfer the technique to the fashioning of clay vessels. This is quite evident when it is remembered that the juglike necks at the top of many of the granaries were built up of large rings of clay strengthened with grass and cedar bark, each successive ring having a shorter diameter than the one below so that the walls were drawn in toward the opening. Furthermore, the small shelflike handles on the outside of the bowls are miniature reproductions of the mud steps placed on granary walls and the slop- ing stone floors in caves.° The walls of the granaries were smoothed and in many cases the evidence for the rings almost obliterated. The same was true for the vessels. Hence it seems that the close similarity between the various steps in the making of granary tops and mud bowls is attributable to more than mere coincidence. When the firing of vessels developed, it became necessary to replace the bark binder in the clay with some noncombustible substance in order to avoid holes and cavities in the walls where the vegetal matter burned out. Thus the custom of mixing sand with the clay was introduced and true pottery, in which the constituent paste had been tempered with sand, appeared. As the industry progressed there was a marked reduction in the amount of sand used, the potters apparently discovering that a smaller proportion and a better knead- ing of the clay produced a harder and more compact texture. Sand was gradually replaced by ground or powdered rock, the kind vary- ing in accordance with the available material in the district where the potter lived, but generally it was either of the light-colored quartz or quartzite group or a dark igneous stone. Wares of this type were standard for the Modified Basket Maker period. Modified Basket Maker pottery occurs in a variety of shapes. There are full-bodied jars with restricted necks and small openings, full-bodied jars of an elongated spherical shape with very short neck and large orifice, globular or spherical pots with slightly de- pressed top and wide orifice, globular vessels with a small opening, bowls, pitchers, scoops and ladles, and small pots with lateral spouts. There are three main groups of vessels. The largest consists of ® Morris, 1927, pp. 159-160. ROBERTS | ARCHEOLOGICAL REMAINS IN ARIZONA 5 those with a light gray to fairly good white color and includes all of the different shapes. A much smaller series is that of bowls with a dull black interior and a gray exterior. The third class is an extremely minor one comprising pitchers and bowis of a reddish cast that seems to have been produced by overfiring. Whether this was intentional or the result of secondary firing in the burning of houses is not known, but occasional specimens exhibiting the fea- ture occur at most sites. The vessels were made from the same materials and in the same manner as the light gray to white examples, no wash that would produce a red color under proper firing conditions being applied to the surface. The finish on most vessels was somewhat rough because of protruding particles of the sand or crushed rock used in the tempering process. For this reason the pottery has a characteristic, irregularly stippled appearance. Decoration, for the most part, was confined to the interior of bowls, although there are sporadic examples with a design on the outside of a bowl, in a ladle or on a pitcher. Two kinds of pigment, one a carbon and the other an iron, were used in applying the ornamenta- tion. The carbon type prevailed in northeastern Arizona, while the iron form, in later stages an iron-carbon, was widely distributed over the remainder of the area. The designs are generally ribbon- like panels embellished with dots, zigzag and stepped-line elements, occasionally with wide-spaced squiggled hachure lines, and less fre- quently with zoomorphic or anthropomorphic figures. Most of the decorations were carried over from basketry to pottery and the earliest designs are close copies of those which already were familiar to the people. Many vessels during this period were treated, after firing, with a wash of red pigment. This is impermanent and is known as “fugitive red.” The opening of the Developmental Pueblo period witnessed some changes in the pottery. As the period progressed new features were introduced and there was a marked expansion in the industry. The various potters continued to use white sand or pulverized rock in tempering their clay, but in addition they began to add ground potsherds to the mixture. As more of this material became available through the continued breaking of dishes in the regular course of usage, it formed a larger percentage of the mixture and there was a noticeable drop in the amount of sand and rock particles. Surfaces received more careful smoothing and the application. of a slip was introduced. This was a thin coating of “liquid” clay rubbed over the smoothed vessel to give it an even better surface. One group of vessels, those intended for culinary purposes, took on a feature that is diagnostic of the period. The outer surface around the necks. was not rubbed and the bands of clay from which they were 154468—40 7 6 BUREAU OF AMERICAN ETHNOLOGY [BULD, 126 fashioned were left as a decorative element. At first these bands were broad and each one made only a single circuit of the vessel wall. As the stage advanced the fillets were rolled smaller and were long enough to circle the perimeter several times and true coiling resulted. This was followed by indentation of the coils by pinching with the finger, by the working in of simple designs through in- dentation, and the incising of figures with a tool or finger nail. Sometimes the bands or coils were rubbed and then reemphasized by the use of a tool. In all of these forms the bottom, from the shoulder down, was carefully smoothed, Developmental Pueblo pottery has a great diversity of forms. There are large full-bodied jars with long tapering necks and re- stricted orifices; full-bodied jars with short, squat necks and con- stricted orifices; globular-bodied jars with short necks and wide ori- fices; full-bodied vessels with an elongated spherical or ovoid shape with squat neck and wide orifice; spherical or globular pots with a small, round opening at the top and similar-shaped vessels with a short-cylindrical neck and small orifice, the type called canteens; globular pots with a flattened top and medium-sized circular orifice, the so-called seed-jar shape; pitchers of many varieties, including the bird-shaped forms; cylindrical jars; double-lobed jars; bowls; ladles and scoops; eccentric forms; and effigy vessels. The colors are plain gray, black on white, black on red, and there are bowls with a brownish exterior and slightly burnished interior. As the period advanced the latter became a good red with a highly burnished black. All types of vessels were decorated. In the earlier stages the main design elements were zigzag, parallel and parallel- stepped lines, squiggled lines; solid triangles and dotted triangles; volutes and ticked volutes; interlocking frets; running frets; concen- tric rectilinear and curvilinear figures; and checkerboard patterns. In addition to the decorations taken from baskets are those copied from head bands, sandals, and textiles. Later in the period the pot- ters began to drop the series of bordering lines from the designs and used more broad, heavy elements. The black-on-white pottery, that is, light-colored vessels orna- mented with painted designs in black, of the Developmental period was characterized by two main groups. This was more pronounced in the earlier part of the horizon than in the later stages, but forms derived from both persisted throughout all phases and even continued into the subsequent Great Pueblo. There were many local, minor variations, of course, yet all seem to have been related to one or the other of the two main groups. One of these occurred throughout the eastern part of the Anasazi province and the other predominated in the western precincts. The eastern form centered about the Chaco ROBERTS ] ARCHEOLOGICAL REMAINS IN ARIZONA 7 Canyon area and apparently diffused from that locality. The west- ern developed and spread from the Kayenta district in northeastern Arizona. The eastern type had a wider distribution, covering the area from the San Juan basin in southern Colorado to the Upper Gila region in southern New Mexico and from the Rio Grande on the east to approximately the Arizona-New Mexico boundary on the west. Its southern fringes extended into Arizona along the Puerco River and as far south as the juncture of the Zuni and Little Colo- rado Rivers. The western ranged from northeastern Arizona to the Little Colorado in the eastern part of the state, crossed that stream somewhat farther west and faded out in southeastern Nevada. The eastern boundary is not sharply defined and there is a strip along the Arizona-New Mexico line where the two overlap. Both forms represent what is generally known as Pueblo I pottery. The western was recognized first and for a long time was considered typical for the period. It has been called by a number of names. Slab-house was the first,° and this was followed by Pre-Pueblo,’ Kayenta Pre-Pueblo,? Western Pueblo I,° First Tusayan,! and the present designation of Kana-a black on white.11 The eastern was not identified as a Pueblo I ware until much later? It has been known by a number of different names. “Quien Sabe” was com- monly used until its status was determined, then Chaco Canyon Pre- Pueblo, Chaco I,* Eastern Pueblo I,1* and Kiatuthlanna black on white.** No one designation has been adopted as the preferred form, although Chaco I is probably more widely used than all the others combined. The difference between the forms is twofold, the surface appearance and the type of pigment employed in the decorations. A carbon paint was applied to the western or Kana-a vessels, while an iron mixture was used on the eastern or Chaco-style wares. The decoration on the Kana-a, black on white seems to penetrate into the slip, an appearance like that made by ink lines on unsurfaced paper, while the Chaco designs stand out from their background (pl. 1). Although such is not actually the case, the Kana-a vessels look as though the surface polish was applied over the ornamentation. The difference in pigment was a continuation of the practice noted for Modified Basket Maker wares. ® Kidder and Guernsey, 1919, pp. 152-153. 7 Kidder, 1924, pp. 74-76; and Morris, 1919, p. 204. 8 Roberts, 1931 a, p. 10. ® Roberts, 1931 b, p. 165. 1° Gladwin, 1930 a, p. 181. 1 Hargrave, 1932, p. 15. ? Roberts, 1931 a, p. 10; Roberts, 1931 b, p. 165. 143 Mera, 1935, p. 3, pl. 1. 14 Roberts, 1931 b, p. 165. % Gladwin, 1934, fig. 3, fig. 4. 8 BUREAU OF AMERICAN ETHNOLOGY [Bue. 126 In combining the Pueblo I and Pueblo II periods of the Pecos Classification to make the Developmental Pueblo of the system used in this report,!* a number of pottery forms hitherto considered as dis- tinct from the Pueblo I wares are grouped with them in a sequential series. There is a noticeable difference between early and late Devel- opmental vessels as will become apparent in the detailed discussion of many of the specimens from the Whitewater District. but in most cases the derivation of the later from the earlier types is so evident that there is no need for confusion. The stages in the growth of the cultural pattern as outlined by the Pecos Classification were identified in large degree by the pottery types and there is sufficient variation between early and late Developmental wares to warrant a distinction. There are forms intermediate between the two, however, that could not be assigned to either and the problem of where to place them was troublesome. They were usually referred to as transitional and some confusion was caused by the fact that one field worker would con- sider them Pueblo I while another would list them as Pueblo IT. Under the new classification they take their proper position as middle Developmental forms. The pottery of the Great Pueblo period retained many of the fea- tures of the closing stages of the Developmental. There were the gray wares of the culinary group, the black on white, black on red, burnished black interior with red exterior, and in addition new types consisting of polychrome, black-on-yellow, and black-on-orange ves- sels. The texture for the most part was finer than that for the pre- vious stage and there was an even greater use of powdered potsherds for tempering the clay. A characteristic feature in the gray wares, the culinary vessels, is an allover indented corrugation of the coils. There was not as great a variety of vessel shapes as in the Develop- mental group, but the potters continued to make wide-mouthed, glob- ular-bodied jars; globular-bodied jars with slightly depressed tops and short cylindrical necks with small orifices; small globular-bodied vessels with short necks and small openings, the so-called canteens; rounded-bottom jars with flat top and medium-sized orifice, the seed- jar form; pitchers; bowls; and ladles. Decorations were character- ized by elaborate detail and careful execution and heavy, solid ele- ments were widely used. This was an era of marked specialization and the crystalization of previous trends into patterns typical of various centers so that it is possible to tell at a glance whether the maker of a vessel was influenced by Chaco Canyon, Mesa Verde, Mimbres, Kayenta, or Little Colorado styles.17 16 See Part I, pp. 6-7. 11For more details on ceramics of this period see: Cosgrove, 1932; Hargrave, 1932; Kidder, 1924; Pepper, 1920; and Roberts, 1932, for additional references, ROBERTS | ARCHEOLOGICAL REMAINS IN ARIZONA 9 During the Regressive and Renaissance periods there was a marked inerease in the manufacture of various kinds of colored wares and a dropping off and final abandonment of the black on white. This characteristic prevailed in the modern period as well. Since none of the types made subsequent to the beginning stages of the Great Pueblo horizon are present in the material from the Whitewater District further consideration of their features is not essential to this report and will be omitted. The collection from the Whitewater sites contains a few examples of the Modified Basket Maker wares, a large number of Develop- mental Pueblo pieces, and some from the Great Pueblo horizon. In- teresting data on the sequential relationship of various types was obtained from the stratigraphic tests made in some of the refuse mounds and from the pillars in the pit houses. As mentioned in the discussion of the excavations, a 3-foot square pillar was left near the center of the different house pits when the accumulated debris was removed from them, and these pillars were taken down, layer by layer, the potsherd material being segregated for study purposes. The evidence from 12 such pillars gives a clear picture of the order of appearance of different forms and the steps in the development of various styles of decoration. The proportions were not the same for all tests, as some of the houses were later than others and the pillars did not contain identical strata in all cases. The bottom levels in some were not present in others and the upper strata in the later pits yielded examples completely missing from the fill in the earlier ones. Furthermore, the bottom layers in some of the refuse mounds represented older materials than any from the fill in the houses. By correlating similar levels, however, a continuous sequence is obtained. Wherever a series of identical strata occurred in the pillars and refuse heap tests, the order of the forms was consistent. No attempt was made to work out a detailed statistical study of the potsherds because simple percentages are sufficient to show the main trends and progress in the ceramic industry. The results from different parts of the district check so closely and are in such accord with those from other portions of the Anasazi province where similar studies have been made that there can be little question of their correctness. The wares from the Whitewater District consist of the plain gray, the so-called culinary or utility group, black on white, burnished black, and black on red. All of the larger vessels and most of the small ones were made by the coiling process; a few of the miniature forms seem to have been molded from lumps of clay rather than built up by the use of fillets rolled out from that material. None of them, however, can be considered as examples of the paddle and 10 BUREAU OF AMERICAN ETHNOLOGY [BULL, 126 anvil method of finishing vessels nor did any of the larger forms suggest that technique, one of the characteristics of the Hohokam.* Wherever coils were obliterated, the typical Anasazi system of scrap- ing with a tool and smoothing the surface, adding to the finish with a polishing stone on later forms, seems to have been employed. Percentages for the different wares, based on all the potsherd material from the test sections and from the debris in the houses, show 64.1 for the gray or utility group, 32.7 for the black on white, 9.5 for the black interior with the dull and burnished forms com- bined, 0.2 for the black on red, and 0.5 for the undecorated nonutility class that was treated with the “fugitive red.” The utility group breaks down into three subdivisions with 3.4 for the indented corru- gated, 9.2 for the banded necks, and 51.5 for the plain gray. The latter comprises fragments from the sides and bottom of vessels that had been smoothed. While a majority of these sherds no doubt repre- sent the banded-neck group, some unquestionably must have come from vessels that were completely smoothed and others from those with indented-corrugated necks and smoothed bodies. The gen- eral character of the paste, the mixture of clay and the material with which it was tempered, and surface finish are such that it is only possible to tell in rare cases which of the three forms the plain pieces represent. Hence the necessity for grouping them under the single heading of plain gray. In all probability a good 70 percent of the plain fragments are from the banded-neck class, with the other 30 percent divided about equally between the indented corru- gated group and that composed of the vessels on which all traces of the coils had been obliterated. The preponderance of the banded- neck form in the present series is attributable to the fact that a large part of the excavation was done in structures and refuse mounds belonging to the period when that was the prevailing style. Further investigations would, no doubt, considerably increase the number of indented corrugated pieces, because that type, in which the entire surface of the vessels is so treated, was characteristic of the Great Pueblo stage and most of the work remaining to be done in the district is in ruins of that horizon. This probably would cause a proportionate drop in the percentages for the banded-neck and completely smooth groups. From a broad point of view the significant factor is the proportion of the utility or culinary group as a whole in relation to all of the wares. That the percentages for the various forms in the Whitewater dis- trict are not far from a general pattern is shown by comparison with those from Pueblo Bonito, in the Chaco Canyon. The total for the 18 See Part I, pp. 14-15. ROBERTS ] ARCHEOLOGICAL REMAINS IN ARIZONA 11 culinary group at the latter place is 65.6 and is not greatly divergent from the 64.1 of the former. The black-on-white wares at Bonito were 30.3 percent of all the pottery, a fairly good approximation to the 32.7 percent in the Whitewater series. The burnished black were not as numerous in the Chaco and constituted only 1.4 percent of the total, in contrast to the 2.5 percent of the Arizona material. The black-on-red wares were just the reverse with 2.7 percent in the Bonito group and 0.2 percent in the Whitewater. There are plausible explanations for some of these differences. The burnished black is a form that had its greatest development in southern centers. There is some disagreement about its place of origin, with indications that it may have been a Little Colorado type or that it came from the Mogollon district, and for that reason would have a greater in- fluence on communities near at hand than on those farther removed. Because of its proximity to regions where the ware was popular, the Whitewater District would tend to be affected more than the Chaco. The case of the black-on-red wares is somewhat different. Although they were made from early times they did not attain much prom- inence in the Anasazi province until later stages and the White- water center was abandoned just as they were beginning to become more numerous. Pueblo Bonito and the Chaco Canyon, on the other hand, flourished for some years thereafter and not only had a fair representation of the black on red but even some of the earlier forms of the polychrome that was to become widely used in the following Pueblo stages. The same situation prevails in respect to the black- on-red vessels in the Whitewater District as that noted for the utility wares; namely, the lack of excavation in the Great period ruins probably lessened the number of specimens of that type. However, sherds from such vessels are noticeably rare in the refuse mounds belonging to that stage. This factor is particularly noteworthy be- cause the Whitewater remains are in a region where both the black on red and the polychromes predominate in the fragments scattered about most of the sites. Age differences are unquestionably the rea- son for the variation in the ceramic forms. The larger the propor- tion of colored wares the later the site, seems to be a definite archeo- logical tenet in this particular area. Some of the more significant features emphasized by the material from the stratigraphic tests are those pertaining to the Modified Basket Maker wares, progress in the development of the culinary or gray wares, the sequence in which some of the early Developmental period black-on-white types made their appearance, and changes in the burnished-black group. The oldest deposits of those investigated 19 See Part I, pp. 15-16. 12 BUREAU OF AMERICAN ETHNOLOGY [BULL, 126 were the bottom layers of the refuse mound located southeast from structures 1, 2, and 3, in Group 1. The lowest strata in tests No 1 and No 2 (see Part I, fig 1, p 22, for plan of group and mound and location of stratigraphic sections) were the only ones in all of the sections where Modified Basket Maker sherds occurred unassociated with other forms. Here they were at the very bottom of the strata on the contact line between the dump material and the undisturbed sub- stratum. They either were lying on the surface when the first refuse was dumped there or were a part of it That the latter may have been the case is suggested by the fact that the upper portion of the bottom layer in both No 1 and No 2 contained a mixture of Modified Basket Maker and early Developmental Pueblo fragments. This would indi- cate that the refuse was deposited there during the transition from Modified Basket Maker to Developmental Pueblo. The fact that the Modified Basket Maker pieces found alone are representative of the highest development of that pottery and that the Developmental Pueb- lo pieces are examples of its simplest forms substantiates that con- clusion. Other features associated with this portion of the site, par- ticularly some of those mentioned in the discussion of the structural remains, also suggest a changing pattern. Later levels in these and other stratigraphic sections contained sporadic sherds from Modi- fied Basket Maker vessels, but there is no significance in the occur- rences because they merely illustrate the survival of objects in subse- quent periods and their tendency to reappear under any and all circumstances. The important evidence is the isolated series underlying Developmental Pueblo types in undisturbed deposits. It demon- strates that here as elsewhere the Modified Basket Makers actually preceded the Pueblos, although they do not seem to have been more than well established before the latter arrived. The gray or utility wares had completely smoothed exterior sur- faces in the earliest form represented in the sequential series. This was followed by pots with banded necks, the bands being rather broad and flat. Eventually the bands became narrower and more rounded and on occasional vessels the groove between them was accented through the use of a tool. Next came a slight overlapping of bands accompanied by a pinching of the upper border that caused the lower edges to project slightly and produced a pronounced corrugation. Shortly after the appearance of this form the custom of pinching the coil to make an indentation started. The indentations were widely spaced and the contrast between them and later examples is so marked that it led to the name “exuberant” indented corrugation. Synchronous with the introduction of this form was the development of simple designs pinched into the neck coils or incised with a finger nail or an implement or merely rubbed into the surface with a finger ROBERTS ] ARCHEOLOGICAL REMAINS IN ARIZONA 13 while the clay was still moist. At the same time some of the potters were manipulating the coils to produce a waved effect and the necks of some of the culinary pots had alternating series of plain and waved coils. Closely allied to this form and not long subsequent to it was one on which there were alternating bands of plain and in- dented coils. This entire group, in common with the flat-band forms, was characterized by smooth bodies. It was not long, however, until these varieties of manipulated coils appeared as allover ornamenta- tion. The narrow-coil indented-corrugated, with the corrugations covering the entire outer surface of the vessel, was the last style developed. None of the earlier forms were immediately replaced by new ones after their introduction, but in all cases seem to have per- sisted for a time and then gradually faded from the ceramic pattern. The group with the various manipulations of the neck coils is repre- sented only in the surface layers of tests 1 and 2. Its occurrence is of little significance because fragments from every type of pottery made at a site are likely to be found in surface debris. No examples were present in the pillar in structure 1 below the surface layer, the material there checking closely with that from tests 1 and 2. Midway in the structure-2 pillar, however, these forms began to appear and were the predominant type in the upper strata. In discussing the fill in structure 2, mention was made of the fact that the pit was used as a dumping place for a short time after the house was abandoned. Then for a while no refuse was deposited there. This was followed by an interval when considerable rubbish was thrown into the de- pression *° and included in the material were the potsherds from the vessels with altered neck coils. In structure 12 they were present from top to bottom. Coupled with the evidence from the house types this is a rather clear indication that this group of utility wares made its appearance at about the middle of the Developmental period, per- haps slightly before it, and predominated throughout its latter half. Three forms of early black on white are present in the stratigraphic material. One is closely related to that from the Chaco Canyon, the second suggests that it may be derived from the first but is usually considered one of the Little Colorado wares, and the third is char- acteristically Tusayan or Kayenta. The earliest black on white is Modified Basket Maker of the type prevalent in the Chaco region.” Not a single decorated sherd of the northeastern Arizona variation with carbon paint was noted. A few pieces from vessels of that type are in the series from the Chaco, but they represent a small pro- portion. The iron-paint class is the predominant one there as well 2'Part I, p, 43: 21 Roberts, 1929, pp. 118-123. 14 BUREAU OF AMERICAN ETHNOLOGY [BULL, 126 as in the Whitewater District. Following this ware, and associated with the banded-neck culinary vessels, was the Chaco I, eastern Pueblo I, or Kiatuthlanna black on white. There is no doubt that it was a direct outgrowth from the Chaco type of Modified Basket Maker. The Little Colorado variation represents a group that is found extending west from the Whitewater District to the vicinity of Holbrook, Ariz., and south to St. Johns, Ariz., occurring for the most part in the triangular area bounded by the Puerco and Little Colorado Rivers. Sherds from vessels of this type are abundant on small sites near Houck, Adamana, and Navajo farther west, and in the Milky Hollow and Long H Ranch districts farther south. The base color is more of a gray than that on the Chaco vessels and the lines of the decorations tend to be somewhat heavier. The elements in general are the same, but there are some slight differences in the designs as a whole. The pigment is an iron, in some cases possibly an iron-carbon, mixture. Consideration of the various traits of this pottery suggests that it was a local development strongly influenced by the Chaco pattern. It did not appear in the Whitewater District until the Chaco I form had become well established. Although it is second to the Chaco series, it never constituted a large percentage of the ceramic group as a whole. The Kayenta Pueblo I, or Kana-a black on white, was the last of this early Developmental Pueblo group to reach the Whitewater and it also forms a small percentage, even less than that of the Little Colorado series. During subsequent phases, trends in the black-on-white wares were mainly the result of develop- ments growing out of these three types and a certain amount of blend- ing between them. Toward the end of the Developmental period influences from the San Francisco or Tularosa region in central western New Mexico and from some of the Little Colorado centers farther west were felt in the Whitewater District and the effects are shown by some of the pottery. The burnished-black ware sherds from the stratigraphic tests record an interesting progression. The earliest examples of the form consisted of bowls with a dull black interior and a gray exterior. Their first appearance in the series was in association with the black on white of the Chaco I type. There is no evidence for the form in connection with the Modified Basket Maker, which is interesting because the black-interior gray-exterior bowls were contemporaneous with late Modified Basket Maker in the Chaco area.??. As the Devel- opmental period continued, black-interior bowls with a gray-brown exterior were introduced, and by the time the manipulated-coil culi- nary vessels were coming into vogue the black interior was being 2 Roberts, 1929, pp. 117-118. ROBERTS ] ARCHEOLOGICAL REMAINS IN ARIZONA 15 polished and the exterior had assumed a brownish-red hue. Ulti- mately this form achieved a highly polished interior and a good red exterior. The steps in the growth of this ware in the Whitewater District are particularly interesting because they parallel those noted in the Chaco Canyon and also those observed at the Long H Ranch,” south and a little west from the Whitewater. They differ, however, in the matter of correlation with other forms and horizons in the cultural pattern. The Whitewater examples indicate a degree of lag and appear to be somewhat later, relatively, than those in the other districts. Furthermore, there were no pitchers or jars in the group. The black-interior gray-exterior forms did not disappear from the ceramic scene with the development of the subsequent gray-brown, brown-red, and red varieties, but persisted in a minor capacity and evolved through a number of stages in which the gray became lighter and lighter until it attained an exterior comparable to that of the black-on-white wares, the interior being a burnished black. This seems to have been a wholly local development as burnished-black interior white-exterior bowls have not been reported from other sites in the Anasazi province. There may be some relation between this type and one, apparently later in date, that occurs sporadically in small numbers at scattered sites in New Mexico and Arizona but which thus far has not been localized. It is characterized by a black interior, white to cream exterior with heavy, broad-line decoration in a buff to brown slip paint. There are no traces of designs on the Whitewater specimens, though, and the suggested affinity with the other type may be only a superficial resemblance rather than a real relationship. No whole vessels were recovered and the form is rep- resented in the collection only by sherds. There are a number of vessel shapes in the culinary or utility wares, and modifications and variations appeared from time to time through- out the course of the ceramic industry illustrated by the Whitewater specimens. Some of the larger sizes are not represented by whole pots, but the portions found are sufficient to indicate general shapes and approximate measurements. The smooth-surfaced group had elongated spherical forms with slightly depressed tops and a wide orifice (fig. 1, a) that ranged from 10 to 12 inches (25.4 cm. to 30.48 cm.) in height, 9 to 10 inches (22.86 to 25.4 cm.) in diameter, an opening 5 to 6 inches (12.7 to 15.24 cm.) in diameter, and a wall thickness of from 1% to 34 of an inch (6 to 9 mm.). Another com- mon shape consisted of a fairly globular body with a short, squat neck and wide orifice (fig. 1, b). The larger sizes in this group range from 9 to 934 inches (22.86 to 24.76 cm.) in height, 834 to 10 °3 Roberts, 1931 b, pp. 117-118. 16 BUREAU OF AMERICAN ETHNOLOGY [RULD, 126 inches (22.22 to 25.4 cm.) in diameter, with a neck height of from 2 to 214 inches (5.08 to 5.71 cm.), and an orifice diameter of from 5 to 514 inches (12.7 to 13.97 cm.). Wall thickness varied from 14 to 4, of an inch (6 to 7 mm.). A third shape was that of a globular- bodied pot with a slightly depressed top and wide orifice (fig. 1, ¢). This group was not as numerous as the other two and apparently not Gi d Ficurb i.—Modified Basket Maker culinary vessel suupes. as commonly used in this locality. The vessels averaged from 8 to 9 inches (20.32 to 22.86 cm.) in height, 914 to 1014 inches (28.49 to 26.03 cm.) in diameter, and the orifice from 5 to 6 inches (12.7 to 15.24 cm.) in diameter. The thickness of the walls ranged from 14 to % of an inch (6 to 9 mm.). A few bowls were also made and used for culinary purposes in this particular group. They were of the deep variety with rounded sides and slightly incurved walls at the rim (fig. 1, d). The size range was from 314 to 414 inches ROBERTS] - ARCHEOLOGICAL REMAINS IN ARIZONA 17 (8.89 to 10.79 em.) for height, 514 to 614 inches (13.97 to 15.87 cm.) in diameter, with a wall thickness of 1°44 to 14 of an inch (5 to 6 mm.). Shapes similar to some of those just described were made in smaller sizes. They differ from the larger ones, however, in that they have handles. In some cases a simple, shelflike lug was applied to the outer surface at the shoulder, the place where the neck and the incurving body wall meet, but on most examples the handle is a horizontal, single-loop type. An illustration of the globular body with short, squat neck and wide orifice class with handle is c, plate 2. The vessels in this group average about half the size of those with the same shape but lacking handles. The smaller series ranges be- tween 334 and 414 inches (9.52 and 11.48 cm.) in height, 444 and 514 inches (10.79 and 13.383 cm.) in diameter, with an orifice diameter of from 214 to 31% inches (6.35 to 8.89 cm.). Globular-bodied pots with slightly depressed tops and wide orifice in the group with handles (pl. 2, a) fall within limits of 414 to 514 inches (10.79 to 13.33 cm.) in height, 514 to 614 inches (13.33 to 15.55 cm.) in diameter, and an orifice diameter of from 21% to 314 inches (6.85 to 8.25 cm.). Dif- ferences in wall thicknesses between the larger and smaller forms in this group are so minute that they are negligible. Handled bowls of the culinary wares were of the form illustrated by d, plate 2. They were slightly smaller than those without handles and varied in height between 214 and 314 inches (6.35 and 8.25 cm.), and from 414 to 514 inches (10.79 to 18.97 cm.) in diameter. The handles in all cases were fastened to the side of the vessel by the riveting process. That is, while the vessel was still green or moist holes were poked through the side, the ends of the handle loop were inserted in the holes, extending through the wall to the interior, and smoothed into the inner surface of the vessel. The outer surface of the wall and the handle were welded by smoothing the edges of the perforations and the adjoining portions of the handle until they fused into a unified mass of clay. Other handled forms comprise the type of vessel generally called pitcher. In this group, as in those described above, the body shapes are quite similar to the vessels without handles. The chief difference is in the fact that vertical loop handles were attached to some and not to others. There are globular-bodied pots with short, squat necks (pl. 2, 7) (the handle is missing from this particular vessel but its former presence is indicated by the places where it was attached), and globular-bodied forms with an indication of a shoulder and longer, tapering necks (pl. 2, ¢). These are the prin- cipal shapes in the utility group. The sizes more closely approximate the forms without handles than was the case in the preceding handled 18 BUREAU OF AMERICAN ETHNOLOGY [BULL, 126 examples. Those with the short, squat necks range between 714 and 81% inches (18.41 and 21.59 cm.) in height, 714 and 814 inches (18.41 and 20.95 cm.) in diameter, with orifices from 3% to 5 inches (9.84 to 12.7 cm.) in diameter. The series with longer and tapering necks is from 714 to 81% inches (19.05 to 21.59 cm.) in height, 634 to 8 inches (17.14 to 20.32 cm.) in diameter, and has orifice diameters from 314 to 414 inches (8.25to 11.48 cm.). The range in wall thickness was from 34, to 14 of an inch (4.5 to 6 mm.). The handles are all of the single-loop variety, although in a few cases a double loop was indi- cated by incising a median line down the center on the exterior. This same trait was noted for a few examples in the series of hori- zontal handles. Pitcher handles in this group characteristically start flush with the lip of the rim or slightly above it, as shown by e, plate 2. The entire series just described, regardless of shape, size, and the presence or absence of handles, has the same quality of paste, surface finish, and general treatment. The clay was tempered with consid- erable white sand or light-colored quartz that was ground for the purpose. Surfaces were scraped and rubbed but owing to the quantity and size of the material used in the tempering process tend to be rough and pebbled. The texture of the vessel walls is rather coarse, quite so, in fact, when compared with other Whitewater types but when considered with respect to some of the Mississippi Valley and other eastern pottery it would be called fine. In breaking, the edges tend to crumble and do not have a sharp, straight fracture. This is prob- ably attributable to the large sand or ground-quartz content in the clay. The inner portion of the vessel walls, the core, has a medium- to light-gray streak. The outer surfaces range from a medium dark gray to a very light gray, in some cases are almost a yellowish white. Most of the examples were stained with smoke or coated with soot when found. Traces of “fugitive red” were present on a large per- centage of the specimens, although not very distinct beneath the stain and soot. In all respects this group is typical of the Standard Com- plex of the Modified Basket Maker period.** Similar to the vessels of the preceding group in shape and gen- eral appearance, but differing in nature of the paste and surface texture, are culinary pots that were made in the early stages of the Developmental period. They consist of globular-bodied vessels with short, squat necks and wide orifice (pl. 2, 6) and globular-bodied forms with somewhat longer and more pronounced necks and wide orifice. No complete vessels of the latter shape were found, but there were numerous fragments of sufficient size to establish the 2% Morris, 1927, pp. 161-167. ROBERTS J ARCHEOLOGICAL REMAINS IN ARIZONA 19 form and a little more than half of several different pots is present in a number of cases. This form was well represented in the col- lection of pottery from the Long H Ranch, the Kiatuthlanna of the Zui, in a comparable horizon *> and seems to be a definite part of the complex in this general area. Numerous sherds from such ves- sels occur in association with Chaco I and the Little Colorado form of black on white related to it at sites near Houck and in the region just west of the Whitewater district. In some places it is the only culinary form present with the early Developmental black-on-white wares which suggests that the completely smoothed style of finish persisted longer in some sections than in others and that the neck- banding probably came in from the east as a part of the Chaco in- fluence. Vessels with the globular body and short neck range from 614 to 714 inches (15.87 to 19.05 cm.) in height, 714 to 83% inches (18.09 to 21.27 cm.) in diameter, and have an orifice diameter of from 414 to 5 inches (10.79 to 12.7 cm.). Exact measurements for the other form cannot be obtained because of the lack of complete specimens, but the heights closely approximated 914 to 10 inches (24.13 to 25.4 cm.) and the diameters 934 to 1014 inches (24.76 to 26.03 cm.). Wall thicknesses ranged from 34, to 4, of an inch (4.7 to 7.9 mm.). Vessels in this group tend to a dark-gray color. The clay was tempered with ground quartz and some powdered potsherds, and the texture of the paste is somewhat finer than that of the Modified Basket Maker group. The surfaces were scraped and rubbed, but are rough because of protruding particles of the material used in tempering. They are not as rough, however, as the preceding forms. When a pot was broken, the edges of the fragments tended to crum- ble, although the fracture was somewhat sharper than in the case of the Modified Basket Maker. Another feature of this group is that the surfaces occasionally exhibit a sloughing quality and are pitted where small flakes have fallen away. Fragments from ves- sels of this class are indistinguishable from those from the banded-neck type smoothed sides and bottoms, but can be differentiated from the Modified Basket Maker with little difficulty. The vessels with banded necks in the larger culinary jars were made in two general shapes. One had a roughly globular body with a vertical-sided neck (fig. 2, @) and the other a slightly ovoid body with a neck whose sides had a slight outward curve (fig. 2,6). The bands around the neck were broad and flat and as a rule were lightly rubbed. They were never entirely obliterated. In all cases they were definite bands, each making but a single circuit of the vessel 25 Roberts, 1931 b, pl. 11, a. 20 BUREAU OF AMERICAN ETHNOLOGY [BULD, 126 wall, and not spiral coils. This feature is in accord with that noted for similar vessels from a comparable stage in ceramic development at other Anasazi sites. Complete jars in the large sizes are lacking in the Whitewater collection. Numerous sherds and sizable por- tions are available for study and these are sufficient to give a reliable indication of shape, the nature of the bands and general surface treatment, but not actual measurements. It can be stated with rea- sonable assurance, however, that they approximated a height range of from 12 to 14 inches (30.48 to 35.56 em.), and a diameter varia- tion of from 11 to 121% inches (27.94 to 31.75 cm.). The neck bands range from 1% to 34 of an inch (1.27 to 1.9 cm.) in width and the wall thickness from 14 to 34 of an inch (6 to 9 mm.). a b Fieurp 2.—Developmental Pueblo banded-neck culinary vessel shapes, Pitchers are more numerous in the banded-neck culinary group and they are well represented in the collection. There are a number of shapes. One group is characterized by globular bodies and ver- tical-sided necks (pl. 3, a), the body wall curving inward to form a definite shoulder at the base of the neck. A somewhat similar shape has a globular body, but there is only a suggestion of a shoulder and the neck rises more directly from the body wail (pl. 4, 6). Globular- bodied pitchers with tapering necks probably constitute the largest group, the neck bands starting just above the area of greatest diam- eter (pl. 3, 6; pl. 4, @). In some cases the neck tapers inward for about half its height and then rises vertically or curves slightly outward as in plate 5, ¢ and d. As the period progressed some of the necks became more curved and the ultimate development. was the form illustrated by the example e, plate 5. As shown by the photographs, the number of neck bands varied from 2, 3, or 4 on BUREAU OF AMERICAN ETHNOLOGY BULLETIN 126 PLATE especataes Tee cAI ay TYPICAL POTSHERDS FROM BEGINNING DEVELOPMENTAL PUEBLO WARES. Those in left-hand column are Chaco I; right-hand, Kana-a black on white. BUREAU OF AMERICAN ETHNOLOGY BUEEERING126> (PEATE S2 PLAIN WARES FROM MODIFIED BASKET MAKER AND BEGINNING DEVELOPMENTAL PUEBLO HORIZONS. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 126 PLATE 3 BANDED-NECK PITCHERS FROM EARLY DEVELOPMENTAL PUEBLO HORIZON. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 126 PLATE 4 DEVELOPMENTAL PUEBLO BANDED-NECK PITCHERS. ROBERTS ] ARCHEOLOGICAL REMAINS IN ARIZONA 21 smaller vessels to 7, 10, or 11 on the larger examples. Fragments from the necks of broken pitchers indicate that in some cases there was only one band and it formed the rim for the orifice. The bands on the neck of the pitcher a, plate 4, are interesting because they illustrate the practice of emphasizing bands that have been lightly rubbed by using a tool to deepen or accent the grooves. In this instance the treatment was only applied to the top seven bands and to part of the eighth band. The tool was not used on the remaining lower ones, the edges being sufficiently pronounced without additional emphasis. It is possible that a slip on the part of the potter making the pitcher blurred the upper bands and the tool was brought into use to repair the damage and restore the desired appearance. The tooled effect on this particular specimen is not as pronounced as in cases where there was a careful smoothing of the bands followed by the application of an incising implement. There is considerable variation in size in the pitcher group. Those with definite shoulders and cylindrical necks with vertical sides range from 5 to 734 inches (12.7 to 19.68 cm.) in height, 514 to 65% inches (13.33 to 16.82 cm.) in body diameter, and have orifices with with 314 to 314 inches (8.25 to 8.89 cm.) diameters. The group with tapering necks varies from 334 to 734 inches (9.52 to 19.68 cm.) in height. Diameters at the maximum body circumference are from 47% to 65% inches (12.38 to 16.82 em.). Orifice diameters are from 314% to 3% inches (7.93 to 9.52 cm.). Pitchers in the series with necks that taper inward and then rise vertically have heights of 514 to 7 inches (13.97 to 17.78 cm.), diameters from 55 to 614 inches (14.28 to 15.55 cm.), and orifice-opening diameters from 314 to 4 inches (8.89 to 10.16 cm.). The concave-neck forms range from 514 to 64% inches (13.01 to 15.87 cm.) in height, from 47% to 51% inches (12.88 to 13.97 cm.) in maximum body diameter, and from 21% to 314 inches (6.35 to 8.89 cm.) in orifice diameter. The minimum diameter for the neck will be smaller in all cases than the orifice diameter because of constriction at the midway portion. Wall thick- nesses in all the groups vary from 34¢ to %» of an inch (4.7 to 7.1 mm.). Handles on the pitchers are of two forms, the loop and the flat band. The loop examples occur in the single, the single with median incision to simulate double loops, and double parallel loops. Fragments from broken handles indicate that double loops were sometimes twisted to make a spiral loop handle and in some cases three small fillets were braided in shaping a single piece for attach- ment to the vessel. The flat-band handles consisted of a single loop that was pressed to make a broad flat ribbon of clay. In the case of these vessels, the handles were also attached by the riveting method and extended from the shoulder or the area of greatest body diam- 154468403 22 BUREAU OF AMERICAN ETHNOLOGY [BULL. 126 eter to or just above the rim. In this they resembled the earlier group. For some reason no whole specimens of the smooth-bodied culinary vessels with manipulated-neck coils were found. There were numer- ous sherds and portions of vessels in the fill in house pits and scattered about the refuse mounds, but no more than approximately half of any one example was se ee Judging from the potsherds most of these jars were in the larger sizes, the group that rarely was used for mortuary offerings and the one in which the percentage of breakage seems to have been highest. This may explain the lack of such jars in the collection. Another factor, however, that probably has a definite bearing on the problem, concerns the remains investigated. Mention of a possible gap between the second and third unit-type structures in the discussion of the house types also called attention to the likelihood of evidence for intervening steps in a number of small house mounds that were not excavated.*° These represent the stage when vessels of that type would have been in the greatest vogue, as the potsherds scattered over their surfaces indicate. Furthermore, this group belongs to that phase in the Whitewater District when changes were taking place with great rapidity and the style no doubt was of short duration, the progression to the allover corrugation fol- lowing close on the development of coil manipulation. The portions of jars in the collection show that the shapes were much the same as those for the large-sized banded-neck group, globular bodies with ver- tical necks and ovoid bodies with slightly outcurving necks. Indica- tions are that the height for these vessels was approximately 10 to 12 inches (25.4 to 30.48 cm.), and the diameter 914 to 11 inches (24.13 to 27.94 cm.). These figures are based on estimates from the portions of vessels and are not to be regarded as other than reasonably close to the actual measurements. The designs pinched into the coils were simple chevrons, pyramidal and rectangular figures. The decorations cut into the surface were of the chevron type, diamond-shaped, rectangular frets, or series of oblique parallel lines. This style of ornamentation was fairly wide- spread. It was not used on a large percentage of the culinary wares in most of the places where it is found but was not uncommon. Evi- dence throughout the Anasazi province is that it came into promi- nence late in the Developmental period and, except for one district, Jasted only a short time. There is a good showing of the type at some of the small house sites in the Chaco Canyon, particularly those in the vicinity of the Fajada, a butte some 3 miles (4.828 k.) east of the cluster of large communal houses, and some elaborate examples came Party.) GL, ROBERTS ] ARCHEOLOGICAL REMAINS IN ARIZONA yas) from Pueblo Bonito.27 One of the most extensive series thus far noted is in the Tularosa, or San Francisco-Upper Gila, district in central- western New Mexico.”®> There is some indication that the style started in that section and spread from it to other parts of the area. There is no doubt that it persisted there longer than elsewhere and continued for a time in the early part of the Great Pueblo period. Some attribute the development of this method of ornamentation to the peoples occupying the region just south of the Tularosa in the Mogollon district. Considerable use was made of it there, but the style of designs is somewhat different.”® Also, if the latter is as much older than the former as some think it to be, the influence of the one on the other, jumping a gap of several centuries, is hard to explain. On the other hand, if the Mogollon dates are later, and certain fea- tures tend to indicate that may be the case, the derivation of the Tularosa form of incised decoration from that district is more plausible.*? The two are so close geographically and culturally, how- ever, that the developments may well have stemmed from a common trait and placing them in separate categories may be drawing too sharp and misleading a distinction. Vessels in the group with manipulated-neck coils range from a medium dark to light gray in color. The clay was tempered with some powdered rock and ground potsherds. The texture of the paste is medium to coarse. The exterior surfaces were scraped from the bottom neck coil over the remainder of the body. The smoothed areas were not thoroughly polished, but they were rubbed after scraping and there are small areas suggestive of a polish. The surface is rough, although not as rough as in the case of preceding forms. Occasional pieces of material used in tempering project from the surface. Because less of the powdered rock and more pulverized potsherds were used in the paste the stippled appearance common to earlier forms is not as pronounced. Breakage tends to be along sharper lines and there is not as marked a tendency for the edges to crumble as‘in some of the other types. Firing seems to have been better as there is not as definite and extensive a dark core in the paste of sherds from these vessels and smoke smudges are not as common. The group composed of vessels with allover corrugation and the manipulation of coils, the extension of features described in the pre- 27 Pepper, 1920, fig. 121. 2 Hough, 1914, pls. 6, 7. 2 Haury, 1936, a, b. #9 Results of recent work in the Mimbres area in southwestern New Mexico, particularly in some Mogollon sites, add weight to the belief that the Mogollon pattern may be later than previously postulated. See Nesbitt, 1938. 24 BUREAU OF AMERICAN ETHNOLOGY [BULD, 126 ceding series to the entire outer surface, exhibits a variety of sizes and shapes. Many of the latter are a continuation of those made in older types of culinary jars, but there are some that show modifica- tion. The large sizes, represented only by potsherds and portions of vessels up to approximately half the complete container, were made in two shapes. One consisted of a full body with rounded bottom and walls tapering to a wide orifice with outcurved rim, the area of greatest diameter being at approximately the middle line of the jar (fig. 3, a). The other had a rounded bottom, sides taper- ing to a wide orifice with slightly outcurved rim, the line of great- est diameter occurring about two-thirds of the distance down the side from the top (fig. 3, 6). Those in the first group had a height range between 1014 and 12 inches (26.67 and 30.48 cm.), diameters from 1214 to 1334 inches (81.11 to 34.92 em.), and orifice diameters a 6 Figure 3.—Corrugated culinary jar shapes of late Developmental and early Great Pueblo phases. from 8 to 91% inches (20.32 to 24.13 cm.). The second group aver- aged somewhat larger with a height range of 12 to 14 inches (30.48 to 35.56 cm.), diameters from 123¢ to 1414 inches (31.48 to 36.19 cm.), and orifice diameters from 81% to 95@ inches (20.63 to 24.44 cm.). Wall thicknesses range from 34. to 144. of an inch (4.7 to 8.7 mm.). Most of these examples had alternate bands of plain coils and “ex- uberant” indentations. One combination consisted of 3 plain coils and 9 indented coils, repeated several times. Another had alternat- ing bands of 4 indented coils and 4 plain coils on the upper part of the jar while the lower, from the line of greatest diameter, was. all indented coils. A still different combination had 6 indented | coils and 7 plain coils recurring from top to bottom. Apparently | there was no particular significance in the different numbers as | those used and the variation in association did not follow a definite | pattern from vessel to vessel. ) ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA 25 Shapes in the smaller jars in this group were a little more diverse than in the case of the larger ones. The forms described for the latter are noted, but there are others in addition. The majority have rounded bottoms, although a few were provided with a flattened area on which to rest. Some have rather globular bodies with a relatively short neck (pl. 7, f) ; others a bowllike bottom with sides tapering to the orifice from the line of greatest diameter (pl. 7, c, ¢), and an outcurved rim; and some a globular body with a longer neck, the sides tapering for about half the distance and then rising ver- tically to a direct rim (pl. 7, @). A minor form was that with a globular body with a wide orifice and outcurved rim with no neck (pl. 7, 6). This shape was more prominent in the series with finer coils and smaller indentations that was subsequent to the present group, judging from potsherd evidence. A good example of the flat-bottomed form is 0, plate 6. This vessel also exhibits another feature occasionally present on some in this group, namely, the small lug handle attached to the rim. The latter occurs on various forms from the earliest to the latest and does not have any particular sig- nificance with respect to horizon or type. On all of these specimens the coils are relatively large and the indentations rather widely spaced. Most of the indentations were obtained by pinching the moist coils of clay with the fingers, but occasionally the indentation was emphasized through the use of some tool, as on the bottom of c, plate 7. This vessel also illustrates the style of ornamentation in which the bands were wiped with a finger to produce a design. These are not as effective as those produced by incision with a tool, but some of the potters occasionally resorted to the method for mak- ing a decoration. A rather crude form of alternating indented and plain coils is shown by e, plate 7. The pitcher d, plate 7, is a better example of that style, however. The size range in this group was not great. The heights ranged from 41% to 7 inches (11.43 to 17.78 cm.), and the diameters from 4 to 6 inches (10.16 to 15.24 em.). The orifice diameters were from 34, to 414 inches (7.77 to 10.79 cm.). Wall thicknesses were from 34, to 5%, of an inch (4.7 to 7.9 mm.). Pitchers in this series had the same general body shapes and char- acteristics as the other vessels. The commonest was that illustrated by d, plate 7. The chief distinction in most cases was the addition of a handle. The latter as a rule was the single-loop variety, although a few had the simulation of double, parallel loops through the incision of a median line and some actually had double, parallel loops welded together. Most of the handles started at the edge of the rim or slightly below it, not above, as noted for examples in preceding series, and were fastened to the side by the riveting method already described. The pitcher or mug a, plate 6, is unique 26 BUREAU OF AMERICAN ETHNOLOGY [BULL, 126 in the series. It is flat-bottomed, almost cylindrical in form and has a slightly outcurved rim. Some of the indented-corrugated pot- sherds in the collection may be from similar vessels but none are of sufficient size to warrant the conclusion that they are. Those big enough to indicate the type vessel are obviously from other forms. Hence it seems that very few of this shape were manufactured by the local potters. Why this should be is not known because the form may well be derived from that illustrated by }, plate 6, and it was made in sufficient quantities to warrant acceptance of the later shape with handle. Pitcher heights in general range from 51% to 7 inches (13.97 to 17.78 em.), and maximum diameters from 4% to 6 inches (11.11 to 15.24 em.). Orifice diameters are from 314 to 414 inches (8.89 to 11.43 cm.). Wall thicknesses are %¢ to %¢ of an inch (4.7 to 79 mim.). One curious practice in this district was the making of shallow dishes or bowls from the bottoms of broken culinary vessels. When the upper parts of jars were damaged so the container was no longer serviceable, the bottoms were removed, the edges smoothed, and they were continued in use as saucers or plates. Two examples are illus- trated in plate 8. One was from a vessel with a plain-band bottom, (pl. 8, a, d), and the other had been part of a jar with “exuberant” indented corrugation (6, c). The sides of large jars, both culinary and nonculinary, were treated in the same way and occasionally a por- tion of a bowl was shaped into a small saucer. Since these dishes were obviously efficient and useful it seems strange that similar ones were not made as original pieces rather than as adaptations from broken vessels. The paste texture, material used in tempering, and general char- acteristics of this group have no perceptible differences from similar traits in the group of manipulated-neck coil smooth-bodied wares. The main variation, if any, is in the likelihood of a larger percentage of pulverized potsherds in the tempering material mixed with the clay than was the case with the older forms. This can only be deter- mined by study with the microscope and polarized light; it is not apparent to the naked eye or even with the aid of a hand glass. Throughout the course of southwestern pottery the culinary vessels have had a coarser texture and a rougher surface finish and there is no doubt that these characteristics are correlated with the matter of function. The reaction to continued effects of heating was probably more favorable in vessels of this type, not only from the action of fire on the container itself but from its quality as a conductor as well. The roughened surfaces, particularly in the case of the allover in- dented-corrugated vessels, would make them easier to lift about and with less danger of slipping from the hands when the latter were wet BUREAU OF AMERICAN ETHNOLOGY BULLETIN 126 PLATE d e DEVELOPMENTAL PUEBLO PITCHERS. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 126 PLATE 6 CORRUGATED PITCHERS FROM END OF DEVELOPMENTAL AND BEGINNING OF GREAT PUEBLO PERIOD. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 126 PLATE 7 LATE DEVELOPMENTAL PUEBLO AND EARLY GREAT PUEBLO CORRUGATED VESSELS. BUREAU OF AMERICAN ETHNOLOGY BULCEERING 1263 PE Aes BOWLS MADE FROM BOTTOMS OF BROKEN CULINARY JARS. ROBERTS | ARCHEOLOGICAL REMAINS IN ARIZONA 27 and greasy. The indented corrugations would have a tendency to increase the amount of surface exposed to the heat, but whether or not this was sufficient to make a perceptible increase in the cooking qual- ities is a problem still to be solved. Laboratory experiments with proper equipment would show how much difference, if any, occurred between the smooth and corrugated body forms in the transmission of heat from the fire to the contents of the vessel. The nonculinary wares exhibit a greater variety of forms and sizes than the culinary group. The potters made numerous miniature vessels as well as those large enough for actual use. The group as a whole includes jars, pitchers, canteens, seed jars, ladles, scoops, and bowls. In each of these classes there are a number of different styles and shapes. The bowls are the most consistent, but even they show some variation. All things considered, however, the nonculinary group in the Whitewater District is not characterized by as many dif- ferent shapes and forms as in some other sections where there are remains of comparable horizons. The main reason for this is prob- ably to be found in the peripheral nature of the Whitewater District and its definite lag in various features. Centers that influenced the ceramic arts in this locality had no doubt passed the stage in which there was considerable experimentation and diversity in shapes and had settled on the more conventional forms, and it was from the latter that the Whitewater potters received their stimulus. The effigy types and eccentric vessels common to the beginning stage of the Developmental period in the nuclear portions of the Anasazi province are missing here. Evidence from those places is that such forms were no longer being manufactured at the time the Whitewater settlements were started. Hence the lack of similar vessels. In general it may be said that the greater variety of vessel shapes in the nonculinary wares is probably due to the fact that the potters were less restricted by functional requirements than they were in the case of the culinary vessels where serviceability and efficiency were essential and as a con- sequence strove for containers that were both pleasing in appearance and useful. This group of wares supplied the vessels upon which decorations were painted and the fact that they were to be the back- ground for designs probably had an effect on the care with which the pots themselves were made. Artisans engaged in developing and per- fecting good designs would also be inclined to give more attention to the objects on which they were to be placed, with the result that better forms evolved. The miniature vessels as a class are considered as part of the non- culinary group, although some of the specimens are of such a nature that they actually exhibit more of the qualities of the culinary divi- sion in the matter of paste and surface finish. Their size is such that 28 BUREAU OF AMERICAN ETHNOLOGY [BULL. 126 they would not have been suitable for actual use in cooking. A ma- jority of the miniature vessels are of the pitcher type. Those with undecorated surfaces and culinary features in their paste and gen- eral appearance occur in several shapes. The commonest form is that with a globular body, cylindrical neck, and wide orifice (pl. 9, f). A close second is the globular-bodied form with short, cylindrical neck and medium-sized orifice (pl. 9, e). Another group comprises those with a globular body and tapering neck with medium to small orifice as illustrated by the specimen d, plate 9. An elaboration of the latter shape is found in examples with longer tapering necks and small orifices such as ¢, plate 9. This form is somewhat later in the sequence than that represented by d, although the latter shape ap- peared sporadically as long as miniature forms were made. An- other characteristic shape in earlier stages was that of the so-called slipper or bird-form pitchers (pl. 9, a, 6). They occur in both the miniature and larger sizes, but are more common in the group of small vessels. They are widely distributed in the Anasazi province and seem to be typical of the early Developmental period. The form persisted into later stages but did not have the prominence then that it did earlier. Later examples are generally in the larger-sized group and are usually painted with some simple decoration. Handles on the miniature pitchers in the unpainted group have the same general characteristics noted for handles on the various examples of the culinary pitchers. Most are of the single-loop variety, round or oval in cross section, and extend from the rim to the shoulder. There are a few examples of the double-loop variety, but none with a simulation of the double loop produced by incising. In this group, as in others, sequence differences are indicated by the place where the handle is attached to the rim. The oldest forms are those where it starts from the extreme margin. Later examples have the attachment slightly below the rim and as the period pro- gressed it was made still farther down the side of the neck. The size range in the vessels in this group was not great. The ordinary pitcher forms (pl. 9, c to f) have a height range of 33% to 51% inches (8.57 to 18.97 cm.). The latter is really somewhat above the average and this particular specimen, ¢, might be considered as a small example in the regular series rather than in miniature. Body diameters are from 3 to 334 inches (7.62 to 9.52 cm.), and orifice diameters from 1%» to 334, inches (3.25 to 8.09 cm.). Wal! thicknesses are from 14 to 14 of an inch (3.1 to 6.3 mm.). Bird- shaped pitchers range from 2 to 234 inches (5.08 to 6.85 cm.) in height and from 8 to 314 inches (7.62 to 8.25 cm.) in length. Body width is from 114, to 154 inches (2.69 to 4.12 cm.). ROBERTS ] ARCHEOLOGICAL REMAINS IN ARIZONA 29 The miniature pitchers in the painted wares are predominantly of the bird shape, although none are sufficiently detailed to be con- sidered effigy forms. The representation is more formalized with the body indicative of the bird shape and the neck placed in a position corresponding to that on the living model, but there was no molding of distinctive features such as the head, wings and tail. In a few cases, as in the examples illustrated in plate 14, a, c, d, the wings and breast are suggested by slight protuberances at the proper places in the vessel wall. Other shapes in this series consist of globular bodies with tapering necks and small orifices (pl. 12, a) and rounded-bottom forms with wide, tapering necks that are almost variations of a mug (pl. 14, 6). The latter probably is the result of difficulty of manipu- lation of the clay in such small objects rather than an intentional attempt at a muglike vessel. Handles in this group are all of the single loop, round or oval cross section, variety. No parallel double loops or simulated double loops were observed. As on the plain examples, the handle extends from the lip of the rim to the upper part of the body in the older forms and from a point on the neck below the edge of the rim to the shoulder on the later types. The normal type pitchers range from 21, to 314 inches (5.71 to 8.89 cm.) in height and from 1% to 21% inches (4.76 to 6.35 cm.) in diameter. The bird-shaped types have heights from 1% to 3 inches (4.76 to 7.62 cm.), body lengths from 2 to 3 inches (5.08 to 7.62 cm.), and body widths from 15g to 214 inches (4.12 to 5.71 em.). Wall thick- nesses in the group as a whole are from %4¢ to 5%¢ of an inch (4.7 to 7.9mm.) The paste texture, material used in tempering, and general structural features are the same as for the larger types of specimens and will be discussed in the paragraphs pertaining to those forms. Large jars, such as were used for storage purposes and as containers for water, are represented only by fragments and portions of vessels. There are no whole specimens, but the pieces in the collection give clear indication of the shapes and types made in the Whitewater Dis- trict. The plain forms associated with Modified Basket Maker culinary potsherds suggest two shapes characteristic of that phase in ceramic development. One is characterized by a body in which the lower portion, extending from the line of greatest diameter to the bottom, is between a hemispherical and a half-oval form. The upper part is slightly flattened and turns upward to make the tapering neck (fig. 4, a). The other shape consists of a globular body with a con- stricted, bulbous neck (fig. 4, 6). There were no indications of handles for these vessels, although similar forms in other localities were equipped with down-raking lug handles and those of the hori- zontal, single-loop form. There were no suggestions of painted 30 BUREAU OF AMERICAN ETHNOLOGY [BULL, 126 decoration on any of the pieces. The paste texture was similar to that of the culinary jars of the Modified Basket Maker series. The clay was tempered with white sand or pulverized quartzlike rock and projecting pieces of this material impart a characteristic stippled appearance to the surface. The surfaces were scraped and rubbed but do not seem to have been polished. Upon completion, after firing, they were given a coating of the “fugitive red.” ** Only approximate measurements can be given for these jars. The type with tapering neck averaged about 15 inches (88.1 cm.) in height and 14 inches (35.56 cm.) in diameter. The necks were from 214 to 21% inches (5.71 to 6.35 em.) long (this is actual measurement as a number of b Ficurn 4.—Modified Basket Maker jar shapes. a them were found) and the orifices from 214 to 3 inches (6.35 to 7.62 cm.) in diameter. Wall thicknesses were from 14, to % of an inch (6.3 to 9.5 mm.). Jars with constricted, bulbous necks do not seem to have been quite as large. They approximate a 1214-inch (381.75 em.) height and a diameter from 111% to 12 inches (29.21 to 30.48 cm.). Necks were 214 to 234 inches (6.35 to 6.98 cm.) high. The maximum ‘t'This seems to be the type of ware recently christened Lino Fugitive Red, Colton and Hargrave, 1937. In the remarks on the type the writers err, however. They state, ibid., p. 198: “The only way this red color could be produced on vessels fired in a reducing atmosphere would be by painting the vessel after firing since the paint would turn gray if fired. This theory is opposed to that offered by Roberts (1929, p. 9, 10).’’ In describing this type in 1929, the writer stated: ‘‘Another characteristic feature of the surface treat- ment was the application of a red wash (‘fugitive red’’ as it is known to the archeologists of the Southwest) to the exterior of vessels. Because it was not fired into the surface and made permanent it is not apparent, or else is very indistinct, on many of the frag- ments found.” (Roberts, 1929, p. 110, not p. 10, as given by Colton and Hargrave.) What obviously happened is that the latter confused two types—one, the vessels with an drange-red to red tone, as described in preceding pages in this report and also on pages 109-110 in the 1929 paper, that resulted from overfiring, and two, those that were treated with a red wash after firing. They are distinct forms. ROBERTS ] ARCHEOLOGICAL REMAINS IN ARIZONA 31 diameter at the swollen portion was 234 to 3 inches (6.98 to 7.62 cm.), and the orifice diameter ranged from 134 to 2 inches (4.44 to 5.08 cm.). Wall thicknesses were the same as for the other style. An interesting structural difference between the two forms is that the tapering neck on the one was made as a part of the jar, the coils being a continua- tion of those used in building up the vessel, while on the constricted- neck examples the neck appears to have been fashioned separately and then welded to the body. The large jars of the painted series have a number of shapes. Those made at the beginning of the Developmental period are similar to the forms found in Chaco I and other variations of the eastern style of the pottery better known as Pueblo I. There is one example of this group in the Kana-a black on white and none for the Little Colorado. One shape is a continuation of the first de- scribed for the Modified Basket Maker (fig. 4, a). The body is slightly more rounding and the neck a little less tapering (fig. 7), but the main contours of the jar are essentially the same. The com- monest shape, judging from the potsherds, was that of an ovoid body with very short, squat neck; really more of a pronounced rim for the orifice than a neck (fig. 5, a). Associated with this was one with an ovoid body with smaller, slightly pointed bottom, a sug- gestion of a constriction in the walls just above the area of greatest diameter, and a definite, though short neck (fig. 5, 6). Following these were shapes that unquestionably were derived from the pre- ceding examples. One was an ovoid form with a slight constriction and a secondary swell in the upper portion of the walls, midway between the area of greatest diameter and the base of the neck, and a short neck with direct rim (fig. 5, ¢). The other tended toward the globular-body form but still had a trace of the constriction and a slight secondary swelling in the upper portion. It also had a short neck with direct rim (fig. 5, d). These four shapes were typical of the Developmental period in the Chaco Canyon. The first (fig. 5, @) 1s represented by vessels from a pit house in the floor of the Canyon,* the structure that yielded charred beams giving the dendrochronological date 777+10,** and numerous sherds from similar jars occur in refuse mounds at sites scattered over this general region. The form is also present in the Long H Ranch area southwest from the Whitewater District ** and occurs in the early Developmental remains in the Piedra area in southwestern Colo- rado.*®° The shape seems to be characteristic for that general stage 32 Judd, 1924, pl. 4, a. 33 Douglass, 1935, p. 51. See Part I, pp. 262-263, 34 Roberts, 1931 b, pp. 124-125, % Roberts, 19380. 32 BUREAU OF AMERICAN ETHNOLOGY [BULD, 126 in the groups that were under a strong influence from the Chaco pattern. The second (fig. 5, 6) also was a Chaco form, but does not appear to have had as wide a distribution as the first. On the other hand, ¢, figure 5, was made at about as many places as the first. It occurs in the Chaco, in the Piedra region, and in the Whitewater. DY Do Ficgurp 5.—Developmental Pueblo jar shapes. The last shape (fig. 5, 7) had a wide distribution and unquestionably was the forerunner of typical body forms in subsequent stages, the globular body with cylindrical neck (fig. 6, @) and the globular body with slightly depressed top and cylindrical neck (fig. 6, 0). The latter represent the ultimate jar developments in the White- water District and also are typical forms for the Great Pueblo ROBERTS ] ARCHEOLOGICAL REMAINS IN ARIZONA 33 horizon in the Anasazi province. Neither of the first two forms seems to have had handles, but most of the others did. The flat-band horizontal handle was the chief form. Some vessels only had one. In most cases, however, there were two placed on opposite sides at approximately the area of greatest diameter or slightly below it. Some of the later types, toward the end of the Developmental period when the globular body with cylindrical neck was the style, were equipped with indented handles, “hand holds” perhaps is a better word. These were made while the walls of the vessel were still green, that is, before they had dried sufficiently to lose their pliancy. The side of the jar was pushed in over an area large enough to admit several fingers of the hand, care being taken not to perforate Ficurn 6.—Great Pueblo jar shapes. the wall, and thus provide an easy means for lifting it. This treat- ment was common in the Chaco Canyon, particularly during the Great Pueblo period, and in fact was quite widespread in the Anasazi area. Surfaces on these jars were scraped and smoothed and all traces of the coils from which they were made were obliterated. In most cases the vessels had a slip, a wash of “liquid” clay containing some kaolin to lighten the color, that was applied to the surface after it was smoothed. This served as a background for the decoration. Both the slip and the paint for the design were added to the vessel before it was fired. The slip on the earliest forms is irregular in quality. It tends to be thin in places and thick in others with scat- tered flecks of undissolved material. Later the slip became more consistent, was applied evenly and, because it was appreciably thicker, attained an almost enamellike quality. Subsequently a thin- ner mixture was used and there is a tendency for the base gray color of the vessel to show through. With the introduction of the slip the practice of polishing vessels with smooth stones came into vogue and as a result the wares have a more finished appearance. Some- times the polishing was done before and sometimes after the applica- 34 BUREAU OF AMERICAN ETHNOLOGY [BULL, 126 tion of the slip wash. Occasional specimens suggest that they were polished both before and after and that their smooth, shiny surfaces are attributable to that factor. The paste texture in most of these vessels is medium to fine; some of the later examples are particularly fine and compact. Pulverized potsherds supplied most of the ma- terial used in tempering the clay and this seemingly enabled the potters to knead the mixture into a less granular type of paste. Breakage occurs along sharp lines and there is little crumbling of the edges of the fracture. As a matter of fact, the edges on many potsherds from vessels of this class are so hard and sharp that they will cut a person’s finger if handled carelessly in the washing process. Many of the fragments show a dark core to the paste. In some cases it is so wide that a cross section appears as a dark body bordered by two thin lines of white, in others it is merely a narrow streak in the center. Some potsherds do not exhibit the feature at all. Variations in firing are responsible. Most of the vessels in this group were large. For some of the shapes the measurements are based on partial specimens, but enough is present to make the figures fairly accurate. Those with the ovoid- body shape (fig. 5, a) were from 14 to 16 inches (35.56 to 40.64 cm.) in height, 11 to 13 inches (27.94 to 33.02 cm.) in diameter, and orifice diameters were from 3 to 4 inches (7.62 to 10.16 em.). Asa rule the orifice was more oval than circular in contour. The rim around the orifice varied from 1% to ¥4 of an inch (3.17 to 6.85 mm.) in height and had approximately the same thickness at the base. It tapered slightly to the lip and measurements for the latter average one-half millimeter less than those for the base. Wall thicknesses ranged from 75 to 23g, of an inch (5.5 to 9.1 mm.). The jars with ovoid body, tapering bottom, and slight constriction in the upper walls (fig. 5, 0) had a height range from 15 to 17 inches (38.1 to 43.18 cm.), diameters from 12 to 14 inches (30.48 to 35.56 em.), neck heights of from 1% inch to an inch (1.27 to 2.54 em.), and orifice diameters from 214 to 314 inches (6.35 to 8.89 cm.). The latter more consistently approximated the circular form than did the other group. The somewhat similar shape (fig. 5, ¢) was made in sizes 14 to 16 inches (35.56 to 40.64 cm.) in height, 1314 to 1514 inches (84.29 to 39.37 cm.) in diameter, neck 14 to 34 of an inch (6.8 to 19 mm.) in height, and orifice diameters of 214 to 314 inches (5.71 to 9.52 cm.). Wall thicknesses in both the 6 and ¢ shapes were from 134, to % of an inch (5.1 to 9.5 mm.). The globular body with slightly depressed top (fig. 5, d) is a shape in which the height measurements are consistently less than the maximum diameter. Heights in this group range from 1114 to 1314 inches (29.21 to 34.29 cm.) and diameters from 1214 to 1514 inches (31.75 to 39.37 ROBERTS ] ARCHEOLOGICAL REMAINS IN ARIZONA 35 em.). The stubby necks range from 1% of an inch (1.27 em.) to 1 inch (2.54 cm.) in height. Orifice diameters are from 21% to 334 inches (6.35 to 8.89 cm.). Wall thicknesses are from 34g to 2%, of an inch (4.7 to 9.1 mm.). The jars with globular body and cylindrical necks (fig. 6, @) have a height range between 141% and 1614 inches (36.83 and 41.91 cm.), diameters from 1634 to 185 inches (42.54 to 47.30 em.), and neck heights from 1 to 2 inches (2.54 to 5.08 em.). Orifice diameters are from 21% to 31% inches (6.35 to 8.89 cm.) and wall thicknesses from 7, to 34 of an inch (5.5. to 9.5 mm.). The group with globular bodies and slightly depressed tops (fig. 6, 6) are from 11 to 131% inches (27.94 to 34.29 em.) in height, 1334 to 1834 inches (34.92 to 47.62 cm.) in diameter, 34 inch (1.90 em.) to 1384 inches (4.44 cm.) neck heights, and 234 to 4 inches (6.98 to 10.16 cm.) for the orifice diameters. Wall thicknesses range from 134, to 2364 of an inch (5.1 to 9.1 mm.). Pitchers of the standard sizes in the nonculinary wares have a va- riety of shapes. One curious form has a three-lobed body with cylin- drical neck and medium-sized orifice (pl. 10, a, 5). There were only a few examples of this body type, but enough more were repre- sented in the potsherds to show that it was definitely a part of the local complex and not accidental. The texture and surface finish on these vessels is strongly suggestive of Modified Basket Maker, yet the asso- ciation in every case was with typical Developmental Pueblo vessels. These pitchers may be an actual survival from the earlier horizon, pieces that were kept as heirlooms or antiques because of their odd shape. On the other hand it is possible that they were made in the early stages of the Developmental and, because much of the Modified Basket Maker technique was still in use, give the appearance of belong- ing to that period. The shape is not common in the Anasazi area.*® It is suggestive of some of the bulbous-legged vessels from the Mexican area, although here the lobes are actually body segments rather than legs. In this the form is more like that of some of the jars found in the Hohokam area and in Arkansas,** but it is not as highly developed. The shape probably owes something to Mexican influence and is gen- erally believed to have originated in Middle America,** although it is more common for jars and vases than for pitchers. Variations of the so-called gourd-type pitcher are present in the series. The Whitewater examples do not follow the vegetal model as closely as those from some localities, but they do suggest the form. The bodies as a rule are globular or globular with slightly depressed 36 A painted example, black on white, was found at Starkweather ruin in the Mogollon district (Nesbitt, 1938, pl. 31, b) in a later horizon than that of the Whitewater specimens. 87 Gladwin, 1930 b, pl. 11; Harrington, 1920, pl. 88 a; Moore, 1910, fig. 34. 38 Vaillant, 1932, p. 13. 36 BUREAU OF AMERICAN ETHNOLOGY [ BULL, 126 tops. The neck portions are continued and bent downward to make hollow handles for the vessels. It is in the latter feature that they simulate the gourd or squash. The example ec, plate 10, is much like some of the other types of pitchers except for the handle which actu- ally is a part of the neck. On this specimen the handle is attached to the body of the vessel at the shoulder but in many cases it is not. On both @ and e, plate 10, it is free at the lower end. The complete. handles are missing from both of the latter specimens, yet the sides of the vessels show that the handles were not attached. In the case of ¢ the fractured end was smoothed and the pitcher continued in use after the tip was broken off. Specimen c is a good example of a modified Basket Maker pitcher with a decoration. The Kana-a black on white is represented by d, plate 10, and the Chaco I by e. The common pitcher shapes were those with globular bodies, com- paratively short necks with approximately vertical sides, and medi- um-sized orifice (pl. 12, ¢, d; pl. 16, b); globular bodies with longer necks and medium-sized orifice (pl. 12, e, 7; pl. 16, a, e); globular bodies with medium-length necks and large orifice (pl. 15, d); glob- ular bodies with tapering necks and wide orifice (pl. 15, b, e) ; globu- lar bodies with a depressed top forming a definite shoulder at the base of the neck and a cylindrical neck with medium-sized orifice (pl. 15, c; pl. 16, c, f). Shapes not so prevalent, although occurring in sufficient numbers to indicate that they belong in the complex, are the squat bodies with depressed top forming a shoulder and long cylindrical neck (pl. 16, d); bird-shaped (pl. 14, e; pl. 15, a); ring- bottomed bodies (pl. 14, g, 2); and the slipper or moccasin type sometimes considered a modification of the bird shapes (pl. 14, /). Most of these shapes are present at sites of comparable horizon in the Anasazi province. The surface treatment and style of decora- tion on these vessels show some differences. Suffice it to say that there are Chaco, Kayenta, Little Colorado, and Tularosa examples with the Chaco predominating. Im all cases the surfaces were scraped and smoothed, generally treated with a polishing stone, and a majority were given a slip before the design was applied. Some do not have it, however, particularly those of the Modified Basket Maker series and some in what is here considered the Little Colorado group. The slips show the same characteristic on the pitchers that they did on the large jars and the same is true for the paste. Han- dies are of the loop, round to oval cross section, flat band, concave band, and on a few examples the effigy type. The latter are zo- omorphic in character and are generally considered as typical of Tularosa ceramics. As on the other types of pitchers previously described, the handles start from the extreme margin of the rim, on a line with the lip, and from the side of the neck below the rim, and BUREAU OF AMERICAN ETHNOLOGY BULLETIN 126 PLATE 9 MINIATURE VESSELS FROM MODIFIED BASKET MAKER AND BEGINNING DEVELOP- MENTAL. PUEBLO HORIZONS. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 126 PLATE 10 NONCULINARY PITCHERS. a and b are trilobed body type; c, d, and e, variations of gourd-shape vessels. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 126) PEATE eC PITCHERS AND CANTEEN FROM THE PAINTED WARES. These are examples of the Little Colorado series. BULLETIN 126 PLATE 12 BUREAU OF AMERICAN ETHNOLOGY VARIOUS FORMS OF BLACK-ON-WHITE PITCHERS. ROBERTS ] ARCHEOLOGICAL REMAINS IN ARIZONA 37 extend to the body in the approximate region of the shoulder. On some examples, as in the case of those with the long necks, the attach- ment of the lower end of the handle is on the neck. The welding and riveting processes seem to have prevailed in fastening the handle to the body. There is a rather wide size range in some of the groups, while others do not show any marked variation. ‘The gourd-type pitchers are from 4% to 71% inches (11.11 to 19.05 em.) in height, 314 to 514% inches (8.25 to 13.97 cm.) in diameter, and the orifice diameters are from 1 to 2 inches (2.54 to 5.08 cm.). Wall thicknesses are from 34¢ to ¥Y, of an inch (4.7 to 6.8 mm.). The three-lobed pitchers have heights from 314 to 434 inches (8.89 to 12.06 cm.), and maximum diameters from 314 to 334 inches (8.25 to 9.52 cm.) ; diameters of the individual lobes range from 134 to 2 inches (4.44 to 5.08 cm.). Orifice diameters are from 114 to 11% inches (38.17 to 3.81 cm.). Wall thicknesses are from 34¢ to %» of an inch (4.7 to 7.1 mm.). The globular-bodied forms with short necks have heights from 454, to 6 inches (10.95 to 15.24 cm.), diameters from 314 to 51% inches (8.25 to 18.97 cm.), and orifice diameters from 11% to 214 inches (38.81 to 6.85 cm.). Wall thicknesses are from %4, to 14 of an inch (4.7 to 6.3 mm.). The globular-bodied vessels with medium-length necks have heights from 434 to 814 inches (12.06 to 21.59 cm.), diameters from 37% to 63 inches (9.84 to 16.19 em.), and orifice diameters from 28% to 4 inches (6.03 to 10.16 em.). Vessel walls range from %4¢ to 54g of an inch (4.7 to 7.9 mm.) in thickness. The globular-bodied forms with medium-length necks and wide orifices, in proportion to vessel size, have heights from 634¢ to 7114, inches (15.71 to 19.5 cm.), diameters from 45¢ to 55g inches (11.74 to 14.28 em.), orifice diameters from 2154, to 314 inches (7.46 to 8.25 cm.). Wall thicknesses are from 7» to 1445 of an inch (5.5 to 8.7 mm.). Pitchers with the globular body and tapering neck with wide orifice range from 71 to 85% inches (18.09 to 21.90 cm.) in height, from 6 to 83% inches (15.24 to 21.27 cm.) in diameter, and the orifice diameter from 314 to 414 inches (8.25 to 10.79 em.). Wall thicknesses range from 7,5 to 23¢, of an inch (5.5 to 9.1 mm.). Pitchers in the globular body with shoulder and cylindrical neck group have a height range between 714 and 81% inches (18.41 and 21.59 cm.), diameters from 534 to 63% inches (14.60 to 16.19 cm.), and orifice diameters 3549 to 4 inches (7.95 to 10.16 cm.). The wall thicknesses are 1364 to 3% of an inch (5.1 to 9.5 mm.). The pitchers with long cylindrical necks and squat bodies have a height range from 9 to 10 inches (22.86 to 25.40 cm.), body diameters from 514 to 6 inches (13.33 to 15.24 cm.), neck diameters from 37% to 5 inches (9.84 to 12.7 em.). Wall thick- nesses range between 14 and 1145 of an inch (6.38 and 8.7 mm.). The bird-shape pitchers have a height range from 45¢ to 834 inches (11.74 154468—40-—4 38 BUREAU OF AMERICAN ETHNOLOGY [BULL. 126 to 22.22 cm.), a body length from 534 to 1034 inches (14.60 to 26.35 em.) , body width from 4% to 714 inches (12.88 to 19.05 cm.) , and orific diameters from 284 to 584 inches (6.98 to 14.60 em.) Wall thicknesses range between 14 and 3¢ of an inch (6.35 and 9.52 mm.). Ring-bot- tomed pitchers have heights from 23% to 4 inches (6.03 to 10.16 cm.), maximum, diameters from 234 to 4 inches (6.98 to 10.16 cm.). Ori- fice diameters are from 1 to 11% inches (2.54 to 3.81 cm.). Wall thick- nesses are from 34¢ to 14 inch (4.7 to 6.3 mm.). The slipper or moccasin forms range between 414 and 5 inches (10.79 and 12.7 cm.) in height, 51 and 6 inches (13.01 and 15.24 cm.) in length, 214 and 3%% inches (6.35 and 8.57 cm.) in width, and have orifice diameters from 134 to 214 inches (4.44 to 5.71 cm.). Wall thicknesses are between 34, and Y, of an inch (4.7 and 6.3 mm.). The canteens or small-mouthed jars do not have a great variety of shapes. There are two main forms; one consists of those exhibiting modifications of a pear shape (pl. 11, f; pl. 17, a, 6, figs. 20, 25) and the other the group with globular bodies and slightly depressed tops (pl. 17, ec, d, f). Most of the pear-shaped vessels have direct orifices, although a few have a low rim bordering the opening as in the case of the one illustrated in figure 20. The other group is characterized by short cylindrical necks. The pear-shaped type is the older form. As in the case of the large jars and the pitchers, the surfaces of the can- teens were scraped and smoothed, and frequently polished. Most of the vessels have a slip, although a few of the earlier ones were not given the wash of “liquid” clay before the decoration was applied. This is especially true for those belonging to the Little Colorado group. Paste qualities are similar to the corresponding wares in the jars and pitchers. Handles are either of the simple lug variety with a vertical perforation, possibly provision for suspension with a cord, as on f, plate 11, and ¢, plate 17; or of the single, short-loop type such as illustrated by a, 6, d, and 7, plate 17. The lug forms as a rule were simply welded to the side of the vessel when the clay was moist, but an occasional example indicates that the riveting process was used. The loop types all appear to have been attached by the riveting process. The size variation in the canteens is not great. Fragments from vessels of this type when compared with the whole specimens show little deviation from the sizes represented. The pear-shaped group in the collection have heights from 414 to 7445 inches (11.43 to 17.83 cm.), diameters from 4535 to 7 inches (10.56 to 17.78 cem.), orifice diameters from 514, of an inch (2.02 cm.) to 134 inches (4.44 cm.). The wall thicknesses are from 34, to 9% of an inch (4.7 to 7.1 mm.). In the second group the heights range from 334 to 4% inches (9.52 to 12.38 cm.), diameters from 414 to 4% inches (10.47 to 12.38 cm.), and BUREAU OF AMERICAN ETHNOLOGY BULLETIN 126 PLATE 13 c, d, Closing stage of same period. BLACK-ON-WHITE WARE PITCHERS. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 126 PLATE 14 BIRD-SHAPED AND RING-BOTTOMED PITCHERS. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 126 PLATE 15 pee i és sunning BLACK-ON-WHITE WARE PITCHERS. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 126 PLATE 16 LATE DEVELOPMENTAL AND EARLY GREAT PUEBLO BLACK-ON-WHITE PITCHERS. ROBERTS ] ARCHEOLOGICAL REMAINS IN ARIZONA 39 orifice diameters from 3344 of an inch (1.30 cm.) to 134, inches (3.01 cm.). The wall thicknesses are the same as for the other series. Seed jars or seed bowls, as they are sometimes called, occur in two shapes. The prevailing form was a globular body with a flat top and medium-sized orifice (pl. 17, e; fig. 832). The other form had a glob- ular body with only slightly depressed top and medium-sized orifice (fig. 31). For some reason this type of vessel does not appear to have been made in very large numbers. In the Chaco canyon and more northern parts of the Anasazi province the seed jars constitute a large percentage in the vessel forms, but here, at the Village of the Great Kivas on the Zufii Reservation,®° and at the Long H. Ranch *° they were not numerous. What significance, if any, this condition implies is not quite clear. The form was not rare in later horizons at other sites in this region and farther west. Perhaps it had its greatest development in the northern parts of the province and the spread to this district was slow. All of the examples found here belong to the Developmental period, there are none from the Modified Basket Maker horizon. The surfaces of these vessels were scraped and smoothed and polished in varying degrees. On some there is a good polish, on others only a suggestion of the use of the stone. The texture of the paste and its other qualities corresponds to the other groups. Some of the seed jars have small nubbin handles with vertical perforations, but most of them are without the feature. The seed jars with globular bodies and flattened tops were rather small as a group. There are no specimens approaching the large sizes found in other districts and none of the potsherds from this type vessel indicated much deviation from the range in the collection. Heights are from 214 to 314 inches (6.35 to 8.25 cm.), body diameters from 334 to 554 inches (9.52 to 14.28 cm.), orifice diameters from 134 to 214 inches (4.44 to 6.35 cm.). The globular-bodied vessels with slightly depressed top are somewhat larger. They range from 4 to 6 inches (10.16 to 15.24 cm.) in height, from 61% to 8 inches (15.55 to 20.32 cm.) in diameter, with orifice diameters from 23/4 to 314 inches (6.98 to 8.89 cm.). Wall thicknesses in both groups fall within the same range, *4¢ to 4. of an inch (4.7 to 7.9 mm.). The scoops and ladles in the Whitewater series show some interest- ing forms. The scoops are of the type where the handle is merely a prolongation of one side of the bowl, a form sometimes called the half-gourd ladle because they are so similar to the well-known dippers made by bisecting such vegetal objects lengthwise. There are several variations in the shapes in this group. The simplest one has an 39 Roberts, 19382, p. 108. 4° Roberts, 1931 b, p. 125. 40 BUREAU OF AMERICAN ETHNOLOGY (BULL. 126 ovoid contour with a broad, shallow handle and no indication of a break or differentiation between the bowl and handle portions beyond a slight indentation of the rim (pl. 18, a). The second shape was derived from the first and is characterized by a definite delimitation of a bowl through building up the wall for several rows of bands before fashioning the handle. This shape is sometimes called the cap-type scoop because it resembles that article of apparel when held in an inverted position (pl. 18, 6). The third form is more nearly like that of a half gourd. The handle is still a prolongation of one side, but it forms a slight angle with the bowl and the latter in its lower portion is distinct from the handle (pl. 18, ¢). Ladles are all of the bowl and handle type. One form suggests a development from the scoop type in that the concave handle is a prolongation of the bowl, although it is separated from it by a partition that actually completes the bow! (pl. 19, @). Other forms have solid handles at- tached to the bowl. Some, as on 0, plate 19, are flat with a slight concavity on the upper side and join the side of the bowl below the rim. Others are roughly rectangular or oval in cross section and are attached to the bowl at the edge of the rim (pl. 19, c). Both scoops and ladles were made during the Modified Basket Maker and Developmental Pueblo periods and the vessels exhibit the same char- acteristics of paste texture and quality as the other containers of the periods. The earlier forms did not have a slip and in many cases were not decorated. The later examples were given a slip and some form of design was painted in the bowl and on the handle. The simple form of scoop appears to have been the first made, but the handled types were not long subsequent. At the end of the Develop- mental period ladle handles were of the hollow, tubular variety. This form persisted in the Great Pueblo horizon. No whole specimens of this style ladle are in the collection, although its presence is demon- strated by handle fragments and by portions of bowls with pieces of handle on them. The simple ovoid-shaped scoops range from 15¢ to 214 inches (4.12 to 6.385 cm.) for bowl height and from 354, to 414 inches (8.41 to 10.79 cm.) for bow] diameter and from 4% to 5% inches (12.38 to 13.97 cm.) for total length. Thickness is from 34, to 44 of an inch (4.7 to 6.3 mm.). The cap-type scoops have a bowl height from 17% to 8 inches (4.76 to 7.62 cm.), bowl diameters from 334 to 514 inches (9.52 to 13.33 cm.), and total lengths from 4% to 6% inches (12.38 to 16.51 cm.). Wall thicknesses are within the same range as for the first type. The third scoop shape ranges from 114 to 21/4 inches (3.17 to 5.71 cm.) in bowl height, from 2 to 514 inches (5.08 to 13.01 cm.) for bowl diameters, and from 43 to 834 inches (11.11 to 22.22 cm.) for total length. Wall thicknesses range from 1% to 3% of an inch ROBERTS ] ARCHEOLOGICAL REMAINS IN ARIZONA 41 (3.17 to 9.52 mm.). Ladles with the scoop-type handle such as a, plate 19, have bow] heights from 154, of an inch (2.88 cm.) to 134 inches (4. 44 cm.), bowl diameters from 2 to 214 inches (5.08 to 6.35 em.), and total lengths from 4%, to 61% inches (10.31 to 16.51 cm.). Wall thicknesses from 34 to 1142 of an inch (4.76 to 8.73 mm.). The bowl] and solid-handle ladles have bowl heights ranging from 1%¢ to 2 inches (3.65 to 5.08 cm.), bowl diameters from 27% to 534 inches (7.80 to 14.60 cm.), and total lengths from 614, to 914 inches (15.39 to 24.13 cm.). Wall thicknesses are from 34¢ to 3¢ of an inch (4.76 to 9.52 mm.), handle widths from 34 of an inch (1.90 cm.), to 114 inches (8.81 cm.), and handle thicknesses from 14 to 44 of an inch (6.35 to 12.7 mm.). There are two types of bowl shapes in the collection. One is the form generally called deep, the other is the shallow. The deep bowls are those in which the height is greater than half the diameter. In the shallow group the height is less than half the diameter. The lat- ter class predominated throughout in the Whitewater District. Some of the deep bowls had full-rounded bottoms and sides that were almost vertical for some distance below the rim (pl. 20, a). Others had somewhat more sharply-rounded bottoms and curved sides with the beginning of a slight incurve to the wall just below the rim (pl. 20, 6). There is some doubt that the latter were local products. Certain features suggest that they may have been trade pieces brought into the district from farther west. Practically all of the shallow bowls had sharply-curved bottoms and walls that followed a regular curve to the rim (pl. 21). Rims in all cases were direct, rarely flat- tened on the lip, and for the most part were undulating and uneven around the periphery. The interior surfaces of the vessels were scraped and smoothed, often polished, and the exteriors were smooth, banded, or corrugated. Smoothed exteriors (pl. 20, a, b; pl. 21, d) predominated, but banded forms (pl. 21, ¢) were not rare. And banded surfaces with tooled grooves between the bands and pinched-in designs (pl. 22, c) were made from time to time. Occasional speci- mens exhibit an allover indented-corrugated exterior (pl. 21, a). Modified Basket Maker bowls have no slip and some of the early Developmental Pueblo vessels also lack the feature, but its use became general soon after the beginning of the latter stage. All interiors were so treated and in most cases the wash of “liquid” clay was ap- plied to the exterior. Sporadic examples occur throughout the series, however, with no outside slip. The bowl (pl. 20, a) is a typical Modified Basket Maker specimen. It exhibits the eastern characteristics of an iron pigment, light- colored pulverized rock particles used in tempering, and an interior that is much smoother than the exterior. It represents the wares 42 BUREAU OF AMERICAN ETHNOLOGY [BULL, 126 belonging to the closing days of the period and unquestionably is related to those found in the Chaco district and regions farther north. Bowls 6 and ec, plate 21, are early Developmental Pueblo. The Chaco I type is well illustrated by 6 which also is like the majority of the wares from the pit houses at the Long H Ranch, that have been designated as Kiatuthlanna black on white.*? The banded example is more in line with the Little Colorado form of the early Developmental, although its counterparts may be seen in collections from the Chaco Canyon and along the Puerco River east from the Whitewater District. The paint in the decoration is an iron mixture and corresponds to the group in that respect. An interesting feature of the banding is that the point of juncture for each band is plainly visible and the loops of clay made but a single circuit of the periphery. They were true bands, not coils. The bowl with the tooled bands and indented pattern (pl. 22, ¢) is a late Developmental form. The indented-corrugated example (pl. 21, a) is Great Pueblo and the smoothed exterior with painted decoration (pl. 20, 6) is from the same horizon. The latter vessel does not seem to be a local product, however, and probably came from one of the centers farther west.** Burnished-black interior bowls and redware bowls tend to the shal- low type. Some approached the deep form with heights approxi- mately the same as half the diameter; this is particularly true for those with the gray exterior and burnished-black interior, but most of them belong unquestionably to the shallow category. The sur- faces on these vessels were scraped and smoothed and generally pol- ished. Some indicate that they were polished before the slip was applied and afterward as well, producing a very smooth and shiny surface. A few fragments from burnished-black interior, corrugated- red exterior bowls and black-on-red interior corrugated-exterior bowls were found on the Great Pueblo refuse mounds but there are no whole vessels in the collection on which to judge the shape type. Most of the black-on-red fragments indicate the type of pottery characteristic of early Developmental Pueblo ceramics. The general color tone is a light red and the pigment in the decoration has a brownish cast and a semitransparent quality that is typical. This same type of red-with-black decoration occurs in the Chaco Canyon and as far northeast as the Piedra District in southwestern Colo- rado.** The form is also found at sites in southeastern Utah where 41 Roberts, 1931 b, pl. 17 e, pls. 19, 21. #2 Gladwin, 1934, figs. 3, 4. 48 This may be an example of the so- ueie Holbrook black on white (Mera, 1934, pp. 9-10) or some related variant of the so-called Deadman’s black on white (Hargrave-Colton, 1932, pp. 15-16). 44 Roberts, 1930, pp. 188-139. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 126 PLATE 17 BLACK-ON-WHITE WARE CANTEENS. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 126 PLATE 18 SCOOPRMEYPES: BUREAU OF AMERICAN ETHNOLOGY BULLETIN 126 PLATE 19 LADLE FORMS. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 126 PLATE b, Great Pueblo bowl. DEEP BOWLS. 20 ROBERTS ] ARCHEOLOGICAL REMAINS IN ARIZONA 43 the percentage of black on red is so much greater than that for other forms that the type seems likely to be a diffusion from that region.*® Later examples of the red-with-black decoration belong to several different groups. One is characterized by a light-reddish color and decorations composed of heavy line, rectangular fret figures, a style of design frequently found on vessels of the black-on-white wares. This form occurs at sites along the Puerco and is quite prevalent in the Chaco Canyon. The decorations are suggestive of those on the so-called Showlow black on red,** but the other characteristics differ too much to consider the Whitewater examples an affinity of the type. Another ware corresponds in general to that called Puerco black on red,*? although there seems to be some confusion over what the char- acteristics of this pottery actually are. A third form is that of the Wingate black on red.** The few sherds from indented-corrugated exterior and decorated-interior red bowls seem to be of the type called Wingate corrugated.*® The paste texture, qualities of fracture, material used in tempering, surface finish and general treatment of the bowls in the black-on- white group correspond to similar characteristics in the other vessel forms belonging to the different series represented. The paste in the Modified Basket Maker bowls was medium in texture although tend- ing to be somewhat finer than that in the large jars and pitchers. White sand or pulverized light-colored rock was used in the temper- ing and particles of the material project from the surfaces to produce the characteristic stippled appearance. Color is gray to white, al- most a chalk white in a few examples. Breakage is a rough and crumbling fracture. Surfaces scraped and smoothed but not actu- ally polished. No slip, yet the interior surfaces were sometimes smoothed to the extent that fine particles were brought to the surface as a float and suggest a thin slip. Decoration applied with iron paint that fired from a reddish brown to a brownish black and occasionally a good black tone.*° The exterior on some bowls was given a wash of “fugitive red.” More may have been treated in this fashion and the pigment may have disappeared as the result of weathering. No ex- 4 This ware is possibly a variant of the type called Deadman’s black on red (Colton, 1932, p. 11; Hargrave-Colton, 1932, p. 18; Colton-Hargrave, 1937, pp. 71-72). It is similar to, but not wholly identical with potsherds designated Pueblo I and illustrated by Guernsey, 19381, pl. 61. 40 Haury, in Gladwin, 1931, pp. 27-28, pl. 31, lower. 47 Hawley, 1936, p. 47, described but not named in Hawley, 1934, pp. 43-44; name with- out description in Gladwin, 1934, fig. 4. The form is redescribed in Colton-Hargrave, 1937, pp. 120-121, but the accompanying sherd illustrating the type does not answer Hawley’s description. It is more like the first unnamed form mentioned above. 48 Gladwin, 1931, pp. 29-381. 49 Mera, 1934, p. 11. 50 This type is the eastern form of Modified Basket Maker, sometimes called La Plata black on white (Hawley, 1936, p. 23). 44 BUREAU OF AMERICAN ETHNOLOGY [BULD, 126 amples of the western form, Lino black on gray,*' are present in the collection although it is not improbable that sporadic sherds of the type will be found in the district. The Chaco I bowls have a medium paste texture. The potters used pulverized light-colored stone, presumably quartz, and ground potsherds for the necessary aplastic in tempering the clay. Color ranges from a light gray to a good white with an occasional dark streak in the core. Fractures are somewhat sharper than in the case of the Modified Basket Maker vessels but have some tendency to crumble. Surfaces are scraped and smoothed with some indications of polish on the interior; exteriors tend to be rough. Slip was applied to both interior and exterior on most vessels, although it was omitted in a number of cases. Slip quality is irregular with thick and thin spots. The designs were painted on with an iron paint that ranges from brownish black to black. Some of the specimens in this group bear considerable resemblance to some of the better Modified Basket Maker vessels and it is quite possible that a form called White Mound black on white,®? believed by some to be a late Modified Basket Maker type occurring in this region, actually is one of the variants in the early Chaco I of the Developmental Pueblo pottery. The following form in the Chaco series, one that grew out of the Chaco I, was given the working designation of Transitional in the strat- igraphic studies of the Chaco ceramics. Since then it has been called Red Mesa black on white ** and Gallup black on white.°¢ The term Transitional was used because the ware apparently bridged the gap between what were then called Pre-Pueblo and Early Pueblo types, later Pueblo I and Pueblo II. The paste texture is fine and compact and considerably harder than that of the form from which it was derived. Ground potsherds with a small showing of light- colored particles, probably pulverized rock that may have been in the sherds employed for the purpose, were used in tempering the clay. Breakage occurs along well defined lines and the edges of the fracture are straight and sharp. The color ranges from a me- dium gray to a good white with some examples exhibiting a gray streak in the core. Surfaces were scraped, smoothed and polished. An unusually thick slip was applied and given such a high polish that it attained an enamellike quality which is readily recognized. The slip varies in color. On some specimens it is a flat white sug- 51 Hargrave, 1932, p. 12. 8 Hawley, 1936, p. 23; Haury, 1936 a, fig. 6, No. 2. 53 Gladwin, 1934, p. 20, No. 35, fig. 4, No. 35. 5 Hawley, 1936, p. 33, pp. 42-43. There seems to be an error in identification in this instance because the description given for Gallup black on white does not fit the Transitional form but applies to other Chaco types. Also the horizon given is much later. ROBERTS ] ARCHEOLOGICAL REMAINS IN ARIZONA 45 gestive of an ivory finish, on others a thin white that apparently is a lighter form of the first. Some bowls have a cream or warm white tone, almost pinkish in cast, and others are a gray white. The lat- ter probably is attributable to insufficient whiting in the mixture. In the earlier forms of this thick slip there was a slight tendency for it to crackle. The slip was generally applied to both the interior and exterior surfaces of bowls, but an occasional example is found where it was omitted from the exterior. The designs were painted with a thick iron pigment that is predominantly a good black, although in places it may range from a brownish black to a brown- ish red. The transitional ware, which seems to have been a rather short- lived one, was followed by a form clearly derived from it yet lacking many of its better qualities. This particular characteristic is so marked that the pottery representing the class was given the working designation of Degenerate-Transitional by the investigators studying the Chaco Canyon ceramics.®> Why there should have been such a pronounced deterioration in the wares is not known. The occurrence in the Chaco was generally considered to be a purely local trend, but with sites located in the more distant reaches of the Chaco range of influence showing the same features, it must be attributed to some widespread motivation not yet recognized. When first noted, vessels belonging to this group were thought to be those made by poorer potters and to represent an inferior form of the Transitional or Red Mesa black on white. The stratigraphic tests, however, demonstrated that it was later in the sequence although the two occur together for a short interval and thus suggest better and poorer examples of the same ware. ‘The paste texture is not as good; it is medium to coarse and tends to be flaky in appearance. The clay was tempered with ground potsherds and coarse particles of sand or pulverized rock, often dark in color and suggestive of a form of shale or lignite. Lines of breakage are more irregular and the edges of the fracture rough, al- though not crumbly. The color is generally gray with a dark core. Surfaces were scraped and smoothed, but rarely polished. The exter- iors of bowls frequently show breaks and scratches and extremely coarse spots. The walls are uneven and tended to warp out of shape before being fired. The slip was applied to the interior and either omitted entirely or else restricted to a narrow band just below the rim on the exterior. Sometimes the slip was merely sloshed on the exterior or the potter wiped her hand across the bottom of the bowl leaving a smear of the “liquid clay.” The slip is thin, the base color % This type is presumably the one that is now known by the name of Escavada black on white (Hawley, 1936, pp. 32-83) and Unpolished Chaco black on white (Hawley, 1934, pp. 36-38). s A6 BUREAU OF AMERICAN ETHNOLOGY [BULL, 126 of the paste showing through on some vessels, and ranges from a dirty gray to a chalky white in character. The pigment in the de- signs is an iron paint and a deep black, like a lampblack, where the color is good, or a dirty rust color where not properly fired. All three of the forms, Chaco I, Transitional or Red Mesa, Degen- erate-Transitional or Escavada, should be considered as subtypes of a single ware, one that is typically of the Chaco pattern. On all three the rims are direct and in general are pinched somewhat thinner than the vessel wall. They are either tapering or slightly rounded in cross section. Around the periphery they tend to be uneven and wavy. Handles were occasionally placed on the exterior of the bowls in this group. They were not restricted to any one particular form but appear sporadically on all three. In some cases one handle was at- tached to the side wall a short distance below the rim, other examples have two placed at opposite sides. The handles are of the loop va- riety. Some are roughly circular to oval in cross section, but the majority are of the flat-band type. They were attached by the rivet- ing method. Growing out of the Chaco-pattern series just described was another Chaco-influenced style of pottery in which the bowls were shallow with a tendency to flattish bottoms and only slightly curved sides. The paste texture is medium. The clay was tempered with some coarse material, but ground potsherds predominated. The paste was kneaded in a more thorough fashion than in the case of the Escavada or Degenerate-Transitional group and was somewhat more compact and slightly harder. Surfaces were scraped and smoothed, although the exteriors on some examples were not well finished, but the polish- ing stone was sparingly used. The slip tends to a chalky-white tone and there was some retention of the practice of merely sloshing the exterior. Most of the vessels, however, were given a good exterior slip. The pigment in the design was a combination iron and carbon paint and fired a deep black.** The rims are direct and the edge or lip is rounding. Associated with the above type, and probably related to it, was one in which the bowls are shallow with slightly curved bottoms, straight to moderately curved sides and direct rims that are tapering with slightly rounded or flattish lips. The paste texture is fine, ight gray to white in color with a tendency to gray streaks in the core, The clay was tempered with light-colored sand or pulverized rock, prob- ably quartz, and ground potsherds. Less of the latter were used in this ware than in some of those that just preceded it. Breakage tends to be along straight lines with a sharp-edged fracture, al- 56'This form seems to correspond to one of those that has been included in the group designated as Puerco black on white (Gladwin, 1931, pp. 24—26). ROBERTS ] ARCHEOLOGICAL REMAINS IN ARIZONA 47 though projecting particles of the material used in tempering make for rather rough surfaces. The interior and exterior surfaces were scraped and smoothed, but they were not polished until after the application of the slip. The slip ranges from a yellowish white to a chalky-gray white. The decoration was applied with an iron paint that produced a heavy, dull-black hue. This ware has a counterpart in the Chaco Canyon ceramics and at numerous sites showing a marked Chaco influence. The Kana-a type bowls are shallow, although in most cases the shape has the same contour as in the deep style, with slightly rounded bottoms and sides that rise with a moderate curve more nearly ap- proaching the vertical than in the ‘case of the other forms. The paste texture is medium to medium-fine. Light-colored sand or pulverized rock, quartz in all probability, was used in tempering the clay. Paste color usually a light gray with a dark streak in the core. Breakage tends to be somewhat irregular and the edges of the fracture are crumbly. Surfaces were scraped and smoothed, but only the interior polished. The scraping marks are frequently present on the exte- rior. The slip is generally a gray-white color and on many examples is crackled. The decoration was applied in a carbon paint that has a black tone but has the appearance of being thin or somewhat washed out, Associated with the Kana-a is a form of bowl that appears to be a Little Colorado variant of the type. The bowls are shallow with flattish bottoms and walls that tend to rise almost vertically. Rims are direct and uneven. The paste texture is medium to medium-fine. The clay was tempered with pulverized rock, occasionally with some white sand. Breakage tends to be irregular and the lines of -the fracture crumble. Surfaces scraped and smoothed, interior slightly polished. A slip was applied in some cases and not in others; when used, it was a thin mixture with a good gray-white tone. The paste color of the vessels is so light that it does not affect the slip hue. The designs were painted on with a carbon pigment that is black, although it tends to appear thin and slightly washed out and may have a slight purple tinge in spots.** This ware probably bears the same relation to the Kana-a black on white that the other form of Little Colorado early Developmental Pueblo does to the Chaco I. 57 Probably is the same as that named Gallup black on white (Hawley, 1936, pp. 42-43), that has been erroneously identified with the Red Mesa black on white (Gladwin, 1934, p. 20, No. 35, fig. 4, No. 35), or the Transitional black on white. 58 In some respects these vessels suggest a prototype for the so-called Deadman’s black on white (Hargrave, 1932, pp. 15-16), although it is possible that it may be an early form of the Holbrook black on white (Mera, 1934, pp. 9-10). The matter of paste color is not in agreement with that given by Mera, yet seems to fit that of Colton-Hargrave, 1937, pp. 235-236. 48 BUREAU OF AMERICAN ETHNOLOGY [BULL. 126 The bowls of the Little Colorado Developmental Pueblo group belong in the shallow category. They have slightly rounded or flat- tish bottoms and sides that are only slightly curved, rising obliquely from the bottom. ‘The paste texture is medium to medium-fine. The clay was tempered with sand and pulverized rock, light brown or gray in color, and some ground potsherds. The paste texture is medium and tends to be somewhat flaky in some cases. The paste color may be a light tan or a grayish white. Dark streaks in the core are rather rare. The surfaces were scraped and smoothed, the interior receiving a light polish. Scraping marks occasionally are noted on the exterior. Use of the slip was not consistent. Some interiors have it, others do not, but the polishing of the interior brought a float to the surface that suggests a slip. Breakage tends to be irregular and the edges of the fracture are crumbly. Decorations were painted on in an iron pigment that fired from a brownish black to a dull black.®® The bowls with blackened interiors, gray, gray-brown, brown-red, red, and white exteriors, exhibit considerable variation in structural characteristics. Those with the gray exterior have medium to medium-coarse paste that has a rather spongelike appearance. Sand was used in the tempering of the clay. It appears, under a hand glass, to be the same kind of light-colored material as that in the Modified Basket Maker and early Developmental Pueblo series. As a matter of fact, the paste is quite similar to that of the early Developmental culinary vessels. The color is a dark gray on the earlier specimens and medium to light on the later ones. There is a thin, black line along the edge that was the inside of the bowl. This line is slightly irregular as the smoke from the smudge that was used in the blacken- ing process seems to have penetrated more deeply in some areas than in others. Surfaces were scraped and smoothed and the interior was given a semipolish. The rims were direct and tapering and the black from the smudge sometimes penetrated the paste completely along this portion of the vessel wall. Breakage is along irregular lines and the edges of the fracture tend to crumble. The main distinction between this form and later examples is that the latter have a some- what finer paste texture, were tempered with sand and some potsherds, °° Certain features, particularly color and paste, in this group suggest the so-called Dead River black on white (Mera, 1934, pp. 8-9) yet it is not clear in descriptions of the latter what the pigment type was, and if the statement Kana-a designs includes Kana-a carbon paint, the present comparison would not be valid. Various designs called Kana-a are also typically Chaco I or Kiatuthlanna, the main difference being that of the carbon or iron pigment used in applying the decoration. If Mera’s Dead River black on white has an iron pigment, the Whitewater Little Colorado Developmental specimens may well represent a variant of that form; if not, they probably are a hybrid comparable to it in that the vessel proper was derived from the same sources while the style of decoration is attributable to influence from the Chaco pattern. ROBERTS ] ARCHEOLOGICAL REMAINS IN ARIZONA 49 have thinner walls, and a smoother interior finish. ‘The exterior gray is also consistently lighter in shade. No slip appears on any of these vessels and the black is dull, not shiny like that of subsequent examples. The paste texture of the black-interior, white-exterior bowls is fine and hard. Ground potsherds were used in tempering the clay. The paste color is gray with a white streak on the exterior side and a black one on the interior. The smudging penetrated rather deeply into the paste. Surfaces were scraped, smoothed, and lightly polished. The rims are direct and tapering with a rounded lip. Breakage is along regular lines and the edges of the fracture are sharp. In some cases the exterior was treated with a slip, in others it was rubbed sufficiently to bring the fine float to the surface and produce a pseudo- slip. The interior black is either dull or lightly burnished ; it does not have the high gloss of the blackened surfaces of red bowls of con- temporaneous manufacture. In places the smudge penetrated entirely through the thinned portion of the vessel wall and shows on the ex- terior in small splotches just below the rim. The paste texture of the brownish-gray exterior group is similar to that of the light-gray exterior specimens; it is medium with a rather flaky appearance. In tempering the clay the potters employed a light-colored sand and some ground potsherds. Surfaces were scraped and smoothed, scraping marks are occasionally present on the exterior, and the interior was lightly polished. Paste color is gray. The exterior is slightly mottled with gray and brownish-gray areas, the interior a dull black. Rims are direct and tapering with rounding or flattish lip. Smudge splotches appear on the exterior just below the rim. Breakage is along somewhat irregular lines and the edges of the fracture have a tendency to crumble. The brown-red group is noticeably different in paste texture. It is fine and hard, although somewhat striated in appearance. The material used in tempering is not distinct. Occa- sional pieces of light-colored sand are present, but are too sporadic to have functioned well in the capacity of a binder. Finely pulverized potsherds may have been used. Surfaces were scraped and smoothed, scraping marks sometimes show on the exterior, and the interior was lightly polished. In most cases a slip was applied. Rims are direct and slightly tapered with rounding or flattish lips. Breakage is along regular lines and the edges of the fracture are sharp but have a slight tendency to crumble. The interior black has a slight gloss and the exterior ranges from brown-red to reddish splotches.*° The bowls with red exterior are characterized by a medium to medium-fine paste 6 In many respects these bowls correspond to the type called Woodruff Smudged, (Mera, 1934, pp. 6-7) that centers along the Little Colorado River south of Holbrook, Ariz., and west of the Petrified Forest, but there is just enough variation to make close correlation questionable. The Whitewater form may represent a local variant of a basic type from which the Woodruff was also derived, which accounts for the similarity. 50 BUREAU OF AMERICAN ETHNOLOGY [RULD, 126 texture, although it tends to be somewhat flaky in appearance and there are sporadic small cavities in the core. Material used in temper- ing the clay consisted of some ground potsherds, light-colored sand that may have been in the sherds prepared for the purpose, some dark, angular material suggestive of a shale or form of lignite, and small particles of an unidentified white substance that is not unlike flakes of shell yet does not seem to be that substance. The color of the paste is gray to black from the center toward the side that was the smudged interior of the bowl and brown to reddish brown from the center toward the surface that was the exterior of the vessel. Surfaces were scraped and smoothed, lightly polished on the exterior and well polished on the interior. Rims are direct, slightly tapering and with flattened lips. There are smudge splotches on the exterior along the thinned-wall area just below the rim. Breakage is sharp and regular with only a slight tendency for crumbling along the edges of the frac- ture. The interior ranges from a grayish black to a good black with a medium gloss. The exterior is a variegated brownish red to red. In most cases a slip was applied. All of the bowls with black in- terior, regardless of the particular group to which they belong, are characterized by irregular smudged splotches or firing clouds on the exterior. The paste of the early Developmental Pueblo black on red, the first in this series of wares, is medium to fine. The clay was tempered with light-colored sand and some ground potsherds. The surfaces were scraped and smoothed, exteriors were not polished and interiors were given a semipolish before the slip was applied. The paste color is gray to gray brown. Rims are direct and slightly tapering with a flattened lip. Breakage occurs along regular lines with sharp edges to the fracture and only a slight tendency toward crumbling. Interiors are a light red that has a suggestion of stippling due to the projecting particles of the light-colored tempering material. The slip has a tendency to be thin and in places shows marks of the mop, the folded fabric used to apply the wash of “liquid clay.” The exteriors were not always treated with a slip and when this was omitted they are characterized by large gray splotches, firing clouds, and small areas with grayish-red hues. Those with a slip have an exterior color similar to that of the interior, with evidences of the mop and large firing clouds. The pigment used in the decorations was a mineral paint ranging from a brownish black to a thin blacix in tone. In places it has a slight metallic sheen suggestive of graph- ite in its appearance. The paste of the red bowls with decorations suggestive of Show- low black on red * is medium to fine in texture. Ground potsherds, 6 Haury, in Gladwin, 1931, pl. 31, lower. ROBERTS ] ARCHEOLOGICAL REMAINS IN ARIZONA 51 some sand, and some dark-colored angular material were used in tempering the clay. The paste color ranges from a pinkish brown to a brownish red in color. Surfaces were scraped and smoothed, with medium to good polish on both exterior and interior. Rims are direct and slightly tapering with flattened lips. Slp was ap- plied to both interior and exterior and is rather thick. It has a tendency to flake off or become somewhat powdery. ‘The color is a good red, although not quite a maroon. The designs were painted in an iron pigment and are a dull black. The bowls in this group have flattish bottoms with slightly curved, vertical sides. The ware has some of the characteristics of both the Wingate black on red ® and the Puerco black on red ® but has sufficient variation from each to set it apart. There undoubtedly is some relation between all three forms, however. The Whitewater black on red comparable to the Wingate black on red has a medium to fine texture. The clay was tempered with ground potsherds and some sand, and a small percentage of light- colored angular particles that may be some kind of pulverized rock. The paste color ranges from a grayish pink to cream. Surfaces were scraped, smoothed, and lightly polished. A heavy red slip is present on both interior and exterior surfaces. It has a tendency to flake off. The color is a good red or a maroon red. The rims are direct with a slight taper and rounded or flattish lips. The rims are uneven around the periphery. Breakage tends to be somewhat irregular with some crumbling of the edges of the fracture. The paint used in the decoration was an iron-carbon mixture and is a dull black. The bowls in this group have rounded bottoms and naturally curv- ing sides and border on the line between the deep and shallow forms. The only difference between this group and those with a corrugated exterior “* seems to be in the treatment of the exterior. Paste qual- ities and other characteristics are the same. There is considerable range in the sizes of the bowls as a group. Some of the special types do not have a marked variation, while others include the limits of the class as a whole. All forms cover a diameter range from 334 inches (9.52 cm.) to 111% inches (29.21 em.), and heights from 184 to 614 inches (4.44 to 16.51 cm.). The Modified Basket Maker examples have diameters between 434 and 7 inches (12.06 and 17.78 cm.), with most of them falling in the range from 51% to 61% inches (13.97 and 16.51 cm.). Heights are from 21% to 41% inches (6.35 to 10.47 cm.). Wall thicknesses are from 545 to %o of an inch (3.9 to 5.5 mm.). None of the larger sizes 6 Gladwin, 1931, pp. 29-31. 6 Hawley, 1936, p. 47; Colton-Hargrave, 1937, pp. 120-121. 6 Mera, 1934, p. 11; Colton-Hargrave, 1937, p. 119. 52 BUREAU OF AMERICAN ETHNOLOGY [BULL, 126 commonly found in this horizon farther north and east are repre- sented in the collection. Chaco I bowls have diameters from 4 to 1014 inches (10.16 to 26.03 cm.), heights from 134 to 41% inches (4.44 to 11.43 em.), and wall thicknesses from 349 to %. of an inch (2.4 to 5.5 mm.). The Red Mesa black on white, the Chaco Transitional form, diameter range is 514 to 834 inches (13.33 to 22.22 cm.), height range 2% to 414 inches (6.35 to 10.79 cm.), and wall thicknesses from %4 to 1344 of an inch (3.5 to 5.1mm.). The subsequent form, Esca- vada black on white, or Chaco Degenerate-Transitional, has diameters varying from 334 to 914 inches (9.52 to 24.13 cm.), heights from 134 to 4% inches (4.44 to 11.43 cm.), and wall thicknesses from 346 to 4 of an inch (47 to 6.3 mm.). The final type in the Chaco series, the one probably corresponding to Gallup black on white, has bowl diameters from 5 to 111% inches (12.7 to 29.21 cm.), heights from 21% to 514 inches (6.35 to 13.383 cm.), wall thicknesses from %4 to 136, of an inch (3.5 to 5.1 mm.). The Kana-a black on white bowls in the collection have diameters from 334 to 81% inches (9.52 to 21.59 em.), heights from 1% to 35% inches (4.76 to 9.2 cm.), and wall thicknesses from 7%4 to 15¢4 of an inch (2.7 to 5.9 mm.). Most of these vessels, however, are in the diameter range of 8 to 814 inches (20.32 to 20.95 cm.), and height range from 3%, to 3%¢ inches (7.77 to 9.05 cm.). The Little Colorado variant of Kana-a black on white bowls have a diameter range from 6 to 834 inches (15.24 to 22.22 em.), heights from 234 to 3% inches (6.98 to 9.84 cm.), and wall thicknesses from 14 to 14 of an inch (3.1 to 6.35 mm.). Bowls of the Little Colorado form considered a variant of the Chaco early Developmental black on white range from 5 to 9 inches (12.7 to 22.86 cm.) in diameter; 2 to 314 inches (5.08 to 8.89 cm.) in height, and have wall thicknesses from 545 to %,» of an inch (3.9 to 7.1 mm.). The black interior bowls as a group have a diameter range from 334 inches to 9% inches (9.52 to 25.08 cm.), heights from 134 to 4% inches (4.44 to 12.38 cm.), and wall thicknesses from 5» to 7%. of an inch (3.9 to 5.5 mm.). The black on red wares are too fragmentary for measurements. DESIGNS ON THE PAINTED VESSELS Decorations painted on the surfaces of vessels in the nonculinary group were applied to the exteriors of the large jars, pitchers, can- teens, and seed jars, and to the interiors of scoops, ladles, and bowls. An occasional design is noted on the exterior of bowls, but consider- ing the group as a whole this treatment is rare. When it does occur 6 Morris, 1927; Roberts, 1929. 6 Hawley, 1936, pp. 42-43; Hawley, 1934, pp. 88-41, called semipolished black on white in the latter. ROBERTS ] ARCHEOLOGICAL REMAINS IN ARIZONA 53 it is generally on specimens belonging to the late horizons rather than to the older levels. The outstanding features in the decora- tions are the elements used and the style of patterns prevailing at different stages in the ceramic sequence in the Whitewater District. These correlate closely with developments in other portions of the Anasazi province and despite the factor of the time lag previously discussed ** show the same order of progression. The main differ- ence is in the interval between the appearance of certain forms. Toward the end of the Developmental Pueblo period the succession was more rapid and the life span of each style was shorter than in nuclear parts of the province. This probably is attributable to an acceleration in diffusion that tended to offset the former lag and bring the peripheral precincts into closer synchronization with gen- eral trends by curtailing the duration of forms rather than omit- ting them altogether, a thing which happened in some districts. During the earlier stages in the Whitewater District the elements used in decorations were simple dots, series of dots framed by bordering lines; zigzag, parallel, stepped parallel, ticked and dotted, squiggled or undulating lines; solid triangles, ticked triangles, triangles with dots, stepped triangles, scalloped triangles, triangles with rectilinear tips or hooks projecting from the apex, triangles with curvilinear tips or volutes, triangles with open spaces enclosing a circle, dot, cross or other small figure; terraced or stepped symbols with short lines rising vertically from the riser to each step or extending horizontally from the tread; volutes or curvilinear scrolls and ticked or dotted volutes; triangular and rectangular frets, curvi- linear frets, interlocking frets, and running frets; checkerboard pat- terns; and occasional simple representations of human or animal figures by means of single line drawing such as children often make. These elements were combined in various ways to make typical and pleasing patterns. Subsequently heavy, solid figures became the vogue with broad lines used as meanders or frets, enlarged and lengthened triangles suggestive of pennants or, when projecting from broad lines, leaves of plants, and heavy scrolls and curvilinear symbols contributing to the boldness of the decoration. Accompany- ing this type of pattern was one characterized by contrasting solid and hachured figures. Both forms appear to have been a direct development out of earlier decorations. In the following discussion of the designs there will be no attempt to interpret the so-called symbolism, to find esoteric or mythological significance in the various combinations of elements or repetitions of the motif. A study of that nature belongs in the field of speculative art and, as the results 8 Part I, pp. 262-263. 154468—40. 5 54 BUREAU OF AMERICAN ETHNOLOGY [BULL, 126 obtained would be based for the most part on guess work, it is not deemed advisable to undertake such a treatment here. Consideration of stylistic features and general characteristics of the patterns and the sequence of their development is sufficient to demonstrate horizon differences and the effects of various influences. The large jars had two zones of decoration. The major one was around the body of the vessel, the minor one around the neck. In some cases the main pattern extended from the base of the neck to the bottom of the jar, in others it was little more than a band around the shoulder or area of greatest diameter. On rare examples an all- over decoration was used and the design extends from the rim of the aperture down the neck and over the entire body. This form appeared late in the Developmental Pueblo period and is largely re- stricted to the solid or heavy type of pattern suggestive of plant leaves. No indications were found of decorations on the jars of the Modified Basket Maker wares. The early Developmental Pueblo forms were characterized by bands around the body and rather sim- ple figures on the necks. A typical example of the form is that on the jar (fig. 7), which also illustrates the somewhat modified body shape derived from the Modified Basket Maker globular vessel with tapering neck. The main elements in this decoration are triangles, parallel lines, and stepped parallel lines. ‘The series of opposed triangles separated by the set of parallel-stepped lines is character- istic. They are combined with chevronlike figures to make a pattern that is repeated four times, two erect and two inverted, in the band. The failure of the potter to make the lines meet at the corners in the parallel-stepped series separating the opposing rows of triangles is a feature frequently observed. The ornamentation on the neck con- sists of two symbols, placed at opposite sides, formed from parallel- stepped lines or, if preferred, from parallel meanders. These two figures are pendant from a single line encircling the neck. Just below the rim is another encircling, single line. No complete vessels in three of the other body shapes of the early Developmental period (fig. 5, a, 6, c) were found and it is not possible to illustrate their decorations. The portions of the jars indicate, however, that they had the same type of designs as are present on similar vessels found in the Chaco Canyon pottery of the same horizon * and other sites under the Chaco influence. Not all of the decorations in this group have the bordering line at the bottom of the band, however. Some of them have an irregular lower border composed of stepped-parallel lines like the designs on some of the pitchers, figure 10, for example. As a matter of fact this style is more characteristics of the jars belonging to the Red Mesa black on white than the plain band form. 6 Judd, 1924, pl. 4, a and DB. ROBERTS ] ARCHEOLOGICAL REMAINS IN ARIZONA 55 An illustration of a more elaborate pattern composed of a number of simple elements is that shown in figure 8. This decoration is painted on one of the jars of the globular body with slightly de- pressed top and secondary swelling in the upper zone type (fig. 5, d). There are two main patterns on the vessel. One occurs on the prin- Fieurn 7.—Design on large black-on-white jar. cipal body zone, the other on the upper portion. The neck is miss- ing hence it is not possible to tell whether it was decorated or not. The decoration around the body proper consists of four similar, although not identical, chevron figures pendant from a line en- circling the jar just above the shoulder. Each of the pendant chevrons is made up of a row of three fret figures in hachure, the framing lines of which are straight and form a rectilinear fret, and the composing lines are squiggled or undulating. Hachure figures 56 BUREAU OF AMERICAN ETHNOLOGY [BULL, 126 with squiggled composing lines are characteristic of the Develop- mental period, particularly its earlier stages. Below this are three, in one example only two, rows of opposed triangles with dotted tri- angles, also opposed, at the ends of the rectilinear tips. In two of the figures the top row has three, the second two, and the third one of these motifs. The other two have four for the top, three for the second, and one has two for the third while the other has no third POQYH 4 Fane oo” Ans FicgurE 8.—Design on late Developmental Pueblo jar. row. At the lower tip of the chevron on three of the figures is a triangular-shaped area of squiggled hachure. This is missing from the other figure, the one that also lacks one row of the opposed tri- angles. A series of 10 parallel lines, that is to say as nearly parallel as the potter could draw them with a brush on a curved surface, forms a continuous lower border for the pattern. In a sense this treatment is comparable to that where parallel-stepped framing lines are employed; the general effect is the same. The decoration on the upper part of the jar consists of three main figures in which there ROBERTS ] ARCHEOLOGICAL REMAINS IN ARIZONA 57 are panels composed of opposed terraced figures with short lines or tips projecting from the corners. The panels are separated by series of parallel lines, the number in each series varying from 8 to 12. At the lower right-hand corner of the main figures is a triangular- shaped area of squiggled hachure similar to that in the decoration around the body proper. The panels of opposed terraced elements and the series of parallel lines start obliquely across the field of dec- oration but, due to the curve of the jar, the last series to the left in each of the main figures is practically vertical and serves to separate one figure from another. This particular jar represents a late phase in the Developmental period while the preceding example (fig. 7) is from the beginning stage. No complete jars with decorations in the solid, heavy type of pat- tern or the contrasting solid and hachure style i design were recov- ered. The patterns, as indicated by the potsherds and portions of vessels, were comparable to those from the Village of the Great Kivas on the Zufi Reservation °° and other sites where there was a strong Chaco influence. These vessels belong to the end of the Developmen- tal Pueblo and the beginning of the Great period. Fragments from jars found in the refuse in the rooms and on the dump at the third unit dwelling *° indicate that designs consisting of rectilinear mean- ders and running frets composed of straight-line hachure were the predominant style of decoration when that structure was occupied. Both the type of hachure and the general motif of the designs are typically Chaco in character. The other forms are more like those considered as characteristic Little Colorado or Tularosa decorations. There were several design zones on the pitchers. As in the case of the large jars, the neck formed one zone, the body another. In some styles of decoration the entire vessel, body and neck, was treated as a single zone. The latter was not common, however, until about the middle of the Developmental period and continued into the subsequent Great Pueblo stage. The handles were generally treated as a sepa- rate zone for decoration. As previously mentioned, painted pitchers in the Modified Basket Maker horizon are rare. The only example in the present collection is the one illustrated on plate 10, c. The decoration in this case consists of a simple band around the body of the vessel. The main elements are an upper and a lower framing line and a series of vertical cross lines from which triangles with curved tips or hooks project. ‘There are two of these on each line. Some of them are so poorly drawn that they suggest thick-based hooks rather than triangles with hooks. The pigment is a thin, brownish black. 6 Roberts, 1932, figs. 19, 20. 7 Part I, pp. 227-244. 58 BUREAU OF AMERICAN ETHNOLOGY execution. medium. [BULL, 126 Considered as a whole the design is good in conception but poor in The potter apparently was working in a none too familiar An example of a single-band decoration covering a greater part of the body of a pitcher is illustrated by the specimen d, plate 11. The elements in this design are merely parallel lines and ticked lines ar- oe’ eee onee Peeeeree® oooocnere? . eee? ae. ponneveee® eee a gscsarracere®® Ficurn 9.—Decoration on early Developmental Pueblo pitcher of the Little Colorado series yet quite effective. ranged in such a fashion that they form triangular figures separated by series of parallel, oblique lines. The pattern is extremely simple, A combination of two simple bands, one around the neck and the other around the body, makes an effective decoration for one of the full-bodied, tapering neck pitchers (fig. 9). The design around the neck consists of a broad band of squiggled hachure with fairly heavy composing and framing lines. This was a popular form of decoration for the neck in the earlier phase of the Little Colorado type of Devel- ROBERTS] ARCHEOLOGICAL REMAINS IN ARIZONA 59 opmental Pueblo vessels at this site. Another example of it appears on the pitcher 6, plate 11. The body bands are different, however. On the first example (fig. 9) the motif consists of a series of concen- tric squares separated by sets of parallel Jines. ‘There is no consistency in the number of concentric squares in each unit of the pattern, one has three, two of them four, and the others five. The same holds true for the separating lines which occur in groups of four, five or six. In the center of each square, though, is an oblique-ticked line. Despite the irregularity of the units the pattern as a whole gives the effect of an orderly decoration. The band around the body of the second pitcher (pl. 11, 6) is composed of opposed rows of simple triangles separated by a dotted line. Below the band are two plain, approxi- mately parallel lines encircling the lower portion of the body. On both of these vessels the pigment in the decoration is a brownish black in some places and black or brown in others. The firing was not sufficiently controlled to produce an even color throughout. Designs that are more characteristic of the early stage of the Devel- opmental period than those just described are exemplified by vessels dande, plate 10. The first is one of the Kana-a black-on-white series. It has a simple band around the neck consisting of opposed triangles with short, straight tips. Between the upper design and that around the body is a single line. The band around the body is composed of four units in which the elements are contiguous triangles and open triangles with an enclosed dot. ‘The lower portion of the design is a series of three roughly-parallel stepped lines. The thick and thin character of the lines with a tendency to blurred edges is typical of the Kana-a decorations. The second pitcher (pl. 10, e) also has a body decoration consisting of four units bordered by a series of three parallel-stepped lines. The elements in the units are dotted triangles and dotted lines. The design on the upper part of the vessel differs somewhat from the first in its position as well as in the nature of the motif. Instead of encircling the neck, which in this case was shaped to represent the handle of a gourd, the pattern in the form of a panel extends along the top. The main elements in the units are opposed, dotted triangles. The dots on some of the triangles, however, are so large that they produce the effect of a scalloped triangle. The units are separated by sets of parallel lines. Inasmuch as a portion of the combination handle-neck is missing, having been broken while in use by the makers and the edges smoothed down to make a new aperture, the number of units in the panel cannot be determined. Judging from the average length of the handle-necks on such vessels it seems that three is the probable number, yet there may have been more. The same style of design as that on the bodies of d and e, plate 10, but with somewhat different elements in the pattern, was 60 BUREAU OF AMERICAN ETHNOLOGY [RULL, 126 used on the pitcher illustrated in figure 10. In this case the same motif was employed for both the upper and lower decorations, al- though there is some difference between the units in each. The design’ around the neck has three units composed of opposed-scalloped tri- angles. These are framed by a series of three parallel-stepped lines. FicurE 10.—Design on Developmental Pueblo pitcher of Chaco type. The bottom decoration, that around the body of the vessel, also has three units. Two of them are composed of opposed triangles sepa- rated by two squiggled or undulating lines, the third has opposing- scalloped triangles that are separated by two lines that appear to be squiggled but which actually are two lines crossed by zigzag lines in which a number of the corners are rounded as the result of too rapid ROBERTS |] ARCHEOLOGICAL REMAINS IN ARIZONA 61 brushwork. Here again are the three parallel-stepped bordering lines at the bottom of the design. This style of decoration is typical of the early Developmental period both on the Kana-a black on white and on the Chaco I or Kiatuthlanna black on white and occurs on vessels other than pitchers. Figure 11.—Pitcher design from middle Developmental Pueblo phase. The opposed triangles and parallel-stepped line motif was also used in other styles of design. They sometimes were combined in the units in a band decoration and did not present the irregular lower border noted in the preceding examples. A good illustration of this is the pattern around the body of the pitcher shown in figure 11. There are five units in this band. In the upper left-hand corner of each are 62 BUREAU OF AMERICAN ETHNOLOGY [BULL, 126 opposed triangles alternately dotted and ticked. On some of them the dots are so large that they suggest a scalloped rather than a dotted triangle, but they were actually placed on the element as dots. The brush was a little too full and the pigment spread where the dot touched the triangle and produced the scalloped effect. The latter, however, when intentionally drawn was obtained by a continuous undulating movement of the brush. At the lower right-hand corner of each unit is a figure composed of two opposed, plain triangles. The space between the two figures with the triangular elements is filled with concentric, 8-sided, rectilinear symbols made up of parallel- stepped lines. The upper zone on this pitcher has a heavy-line recti- linear meander suspended from a line encircling the neck just below the rim. Pendant from this same line, between the places where the meander touches it, are solid triangles with short, straight-line tips. The handle has a simple pattern formed from two zigzag lines that cross each other in such a manner that they make open diamond figures. The ornamentation was completed by painting the lip of the rim. This particular pitcher and the style of decoration represents a slightly later phase than the others on which the opposed triangles and stepped-line elements were employed. ‘The trend toward heavy, broad lines such as that in the meander around the neck began about mid- way of the Developmental period. The use of lines crossed by zigzag lines as a separating figure be- tween rows of opposed triangles, as noted in the units in the band on the body of the vessel shown in figure 10, was rather common and led to modifications that at times produced curious effects. This was the case in the design around the body of the pitcher illustrated in plate 12, d. Here the main band is composed of two rows of op- posed triangles separated by a line of erect and inverted triangles that resulted from filling in the angles produced by the drawing of a straight line with an overlying zigzag. The brushwork is not par- ticularly good on this specimen, but it demonstrates one method of design treatment that was not unusual. Other forms of the triangle element are shown in the decorations on pitchers @ and 0, plate 13. On the first specimen the dotted tri- angle with curved tip or volute is the chief element in both the upper and lower bands. The latter are made up of a series of units in which the dotted triangles are opposed and the curved tips or volutes interlock. ‘The units are separated by sets of double parallel lines. The second example (0) has two bands in which the units are very suggestive of those described in the decorations on some of the pre- ceding pitchers. The main elements in the units are opposed tri- angles, scalloped triangles, stepped triangles with interlocking curved tips, and dotted lines. The distinctive feature in this particular case ROBERTS | ARCHEOLOGICAL REMAINS IN ARIZONA 63 is the omission of the bordering stepped-parallel lines. This is a trend which started about midway of the Developmental period and reached its culmination in the heavy, balanced designs composed of opposing figures that prevailed in the latter part and continued over into the Great Pueblo stage. At about the time the stepped-parallel lines were beginning to disappear there was an increase in the use of Ficurn 12.—Decoration on pitcher of Little Colorado series from middle Developmental Pueblo stage. straight parallel lines of the type that were employed in band designs to separate the units in the pattern. The-difference between these and the earlier forms is that they tend to become a part of the unit rather than part of a frame to set it apart. This is shown in the deco- ration around the body of the vessel in figure 12. The design is of the band type with L-shaped units composed of a series of opposed- dotted triangles and parallel-straight lines. The particular way in 64 BUREAU OF AMERICAN ETHNOLOGY [BULL, 126 which the latter are used is believed by some to be the genesis of the type of hachure decoration that is called the Little Colorado or Tula- rosa style.**_ Others are inclined to question the idea, but that such may have occurred is not beyond the realm of possibility. The band around the neck of this vessel is composed of a series of opposed- ticked triangles. These triangles are of the elongated shape that seems to have developed into the pennant and leaflike symbols of a slightly later phase. Elongated-scalloped triangles, dotted triangles, terraced figures with projecting dotted lines, opposed triangles, and parallel border- ing and framing lines are the elements from which the patterns Fiaurp 13.—Design on bird-shaped pitcher. decorating one of the bird-shaped pitchers were formed (fig. 13). This vessel also furnishes an interesting example of several design zones. One is around the body, another is on the flattened upper portion, a third is around the neck, and a fourth is on the handle. The band around the body is composed of six units separated by series of straight-parallel lines. The number in each set varies from 8 to 12. Each unit consists of corner triangles, opposed-terraced fig- ures with projecting dotted lines, and oblique-parallel lines sepa- rating the terraced elements. Two lines separate the corner triangles 71 Gladwin, 1931, pp. 19-23. ROBERTS ] ARCHEOLOGICAL REMAINS IN ARIZONA 65 from the terraced figures in each case and four or five intervene between the opposed-terraced portions of the design. The decora- tion on the upper part of the body extends in a band around the base of the neck and continues back to the pointed end of the body with its stubby representation of a bird’s tail. The band is not con- tinuous. It merely runs out near the base of the tail. There are Ficgurp 14.—Middle Developmental Pueblo pitcher decoration. six complete units, one that was partially finished, and one con- sisting solely of the space without any elements, in this design. Opposed triangles, opposed-scalloped triangles on the ends of lines projecting from plain triangles, dotted triangles, and parallel fram- ing lines are the elements in the pattern. An interesting factor about this part of the ornamentation on the vessel is that the extensions on each side from the neck back to the pointed end are suggestive of the folded wings of a bird. The pattern around the neck of the vessel just below the rim consists of a series of opposed triangles 66 BUREAU OF AMERICAN ETHNOLOGY [BULL, 126 with projecting lines that support opposed-scalloped or dotted tri- angles of the pennantlike form. The ornamentation on the handle consists of a rectilinear fret figure. The lip of the rim was also painted on this vessel. Two band decorations on another pitcher, one around the neck and the other around the body, are almost identical with that around the neck of the vessel just described (fig. 14). In the designs on the present example the brushwork was not as well done as that on the bird-shaped pitcher, but the decoration is quite Figure 15.—Decoration on ring-bottomed pitcher, side view. effective. The lip of the rim was painted and there was some kind of design on the handle, as indicated on the stub that was left when the major portion was broken off. Not enough remains, however, to indicate the type of decoration. A simpler form of the opposed tri- angle with projecting lines supporting additional opposed triangles, in this case neither dotted nor scalloped, is seen in the band design around the body of the ring-bottomed pitcher (pl. 14, g; fig. 15). This is a good example of a simple pennant style of element. The decoration is further augmented, however, by the oblique, ticked lines connecting some of the other elements. The squiggled lines ROBERTS | ARCHEOLOGICAL REMAINS IN ARIZONA 67 around the neck of the vessel were more crudely drawn than many examples of that motif and contrast rather sharply with the better delineation of the elements in the main decoration. The lip of the rim was painted with a solid line and the handle has a single, median line extending from the rim to the point of juncture with the body (fig. 16). The handle line shows clearly that it was made with two strokes of the brush and that the potter failed to join the ends properly. It is curious that as large an area as the flattened top of this pitcher was left without any decoration. This is rare. Advan- eat Ficurn 16.—Top view of ring-bottomed pitcher. tage was usually taken of such spaces and fragments from other pitchers of this type show that designs were placed there. The other ring-bottomed vessel illustrated (pl. 14, h) was so treated. The decoration on this pitcher is rather simple, consisting as it does of rows of triangles and triangles pendant from the rim. The paint- ing in this design was not well done and exhibits one of the char- acteristics of the Escavada black on white in that respect. The use of dotted triangles and dotted-line elements in the stage following the shift from the parallel and stepped-parallel bordering and framing line style of decoration is well shown by one of the pitchers from approximately the middle of the Developmental horizon (fig. 17). On this specimen there are two bands encircling the body 68 BUREAU OF AMERICAN ETHNOLOGY [BULL, 126 and one on the neck that does not make the complete circuit, an open space occurring beneath the handle. The two bands around the body are the same with units composed of opposed-dotted triangles sep- Ficurn 17.—Chaco series middle Developmental Pueblo pitcher design. arated by dotted lines that connect the projecting tips of the tri- angles. The design around the neck differs in that there are two of the dotted lines in each unit instead of the one. The rim of this vessel is painted and there are two lines extending from the rim to ROBERTS ] ARCHEOLOGICAL REMAINS IN ARIZONA 69 the shoulder on the outer edges of the handle. The latter is slightly concave and the lines serve to emphasize that feature. This particular pitcher is unique in that the body has an oval horizontal section rather than a circular one. It was flattened from front to back in the making and there is a slight shoulder on either side just above the top of the upper design band. Due to this feature it was neces- Figure 18.—Decoration on Little Colorado pitcher from middle Developmental Pueblo stage. sary to tip the vessel somewhat more than usual in making a drawing of it and for that reason part of the body bands appear to follow an oblique direction when actually they are horizontal. At about the time that the vogue for parallel and stepped-parallel lines was on the decline in the Chaco style of design, it took a differ- ent trend on some of the vessels indicative of a Little Colorado derivation. The lines became much heavier. A typical example is 154468—40—6 70 BUREAU OF AMERICAN ETHNOLOGY [BULL, 126 that of figure 18. On this vessel the body and neck zone bands of decoration meet, making a single pattern from the rim to well down toward the bottom of the pitcher. A difference in the character of the design distinguishes the two zones, however. The units in the band around the body are made up of opposed-tipped triangles, tipped triangles and the parallel-straight lines arranged in a pattern similar to some of those already described. The units in the band around the neck are a series of triangular figures composed of a solid triangle in one corner and a series of six parallel lines that make one form of hachure. The fact that the lines run in opposite direc- tions in alternate units produces an impression of a folded fabric or woven style of design that is considered characteristic of the Little Colorado and of the type of pottery that has been called Puerco black on white.72. This type of decoration is not confined to pitchers, but appears on other vessel shapes as well. Heavy-line elements and modifications of the triangle to suggest leaves and pennants were combined in various ways to make suitable decorations. A simple form of heavy-line design is illustrated by one of the pitchers in the Little Colorado series (pl. 12, e¢). This vessel is one of the group with two zones of decoration. Around the body is a band composed of six units. Three of them are made up of three rectangular figures, merely broad lines, placed in a horizontal position and three are composed of a series of parallel lines or, as they are sometimes called, units of horizontal hachure. The decora- ration on the neck consists of four pendant triangular figures with parallel lines making the same style of hachure. On the handle is a single double-ticked line, that is, the ticks cross the line instead of being placed on one side. Designs composed of the pennant and leaflike elements generally cover a greater part of the vessel, extending from rim to bottom. Two characteristic examples of this type of decoration are shown in plate 18,c and d. The first specimen (c) has a design that is good in concept but poor in execution. The brushwork was not well done and some of the triangular elements are not sufficiently elongated to make good examples of the pennant or leaf forms of element, yet they do suggest them. The second is a much better illustration of that feature. This same style of decoration is found on pitchers of the red-and-black wares and is typical of the late Developmental and early Great Pueblo ceramic stages. There is considerable ques- tion about the derivation of this type of pattern and its proper affinities. Some regard it as a typical Little Colorado motif, others 7 Gladwin, 1931, pp. 17-23. ROBERTS ] ARCHEOLOGICAL REMAINS IN ARIZONA valk think it basically Tularosa in its character,’* and the problem is con- fused by the fact that it was a popular style in the Chaco Canyon ceramics at a slightly earlier chronological period than that of the other districts, although the cultural horizons agree. In the Chaco wares it is one of the group comprising the so-called Gallup black on white. Gladwin in the discussion of the Puerco black on white postulates that this form of decoration preceded that in which solid elements are balanced by similar figures in hachure.* Stratigraphic evidence in the Whitewater District and in the Chaco Canyon demonstrate that this was not the case. The combination solid and hachure designs appeared shortly before and subsequent develop- ments were concurrent. Of course, if the style is definitely of Little Colorado origin it may have had a different position in the sequence there and its place in the Chaco series be due to a lag in influence. However, there is sufficient Little Colorado material in the White- water series to give an indication of the stylistic trends and this points to a slightly later inception for the form. As things stand at present, the evidence is not clear enough to warrant conclusions on the source of origin. The Whitewater phase gives the impression that it came into that district from the east with some of the Chaco styles, as the first examples are on Chaco type vessels. This would seem to indicate that the question is one concerning the Chaco and Tularosa districts. One of these pitchers has a handle of the zoo- morphic type (pl. 18, d) that is regarded as characteristic of the Tularosa wares. Another variation of the allover style of decoration composed of pennant or leaflike elements is that on the pitcher illustrated in figure 19. This design emphasizes one of the features that is often consid- ered typical and that is the negative character produced by the arrang- ing of the elements in such a way that the light background strikes a balance with the painted areas and forms new figures. This is appar- ent in the series of open rectangular figures at the right of the design. There is the possibility that this particular trait is a Little Colorado development as it was prominent on some of the later wares and also was comparatively rare in the Chaco series. Hachured elements in decoration occupied a rather prominent place in the development of various patterns. The earliest form of the hachure is characterized by rectilinear and curvilinear-framing lines 7% In the Little Colorado wares it is grouped with the Puerco black on white (Gladwin, 1931, pl. 25) and in the Tularosa as one of the Reserve black on white group, the latter being considered as an affinity or derivation of the Puerco, a form that just preceded the fully developed Tularosa. % Gladwin, 1931, p. 25, pl. 29. (he BUREAU OF AMERICAN ETHNOLOGY [ BULL. 126 with widely-spaced squiggled composing lines. Subsequently the composing lines became straight and progressively closer together. Examples of the early form occur on pitchers }, plate 12, and e, plate 11, as well as on the necks of the vessels previously described (fig. 9; pl. 11, 6). The design on 6, plate 12, consists of a series of pendant Figure 19.—Heavy element type of design from latter part of the Developmental period. triangles bordered by squiggled hachure, an early combination of the solid and hachure elements. The decoration on e, plate 11, is later in the sequence and exhibits fairly well developed characteristics. Around the body of the vessel are two zigzag bands of hachure in which some of the composing lines are squiggled and others are straight. Above these is a band of roughly diamond-shape figures filled with ROBERTS ] ARCHEOLOGICAL REMAINS IN ARIZONA 73 squiggled-hachure lines. This is surmounted by a similar band that does not completely encircle the vessel, while above the latter is an- other that does. Pendant from the rim is a series of triangular fig- ures with the squiggled lines making the hachure. This vessel is one of the Little Colorado group, the early form, with Chaco influence in the design. Diamond-shaped figures with straight-line hachure were also used in band patterns as illustrated by c, plate 14. Most of the straight-lined hachure occurred in fret figures (pl. 14, e,; pl. 15, 0, ¢, e; pl. 16, f); or in meanders (pl. 16 c, d). In the case of the mean- ders the hachure was frequently balanced by painting in heavy op- posing lines (pl. 15, d; pl. 16, 7). Sometimes a number of different- shaped rectilinear figures were used in making the design as in the case of d, plate 16. Filling of the tips of triangular-hachured figures was a common practice at one stage in the Chaco Canyon, but was only sporadic in the Whitewater District (pl. 15,¢c). All of the fore- going examples are characteristic Chaco style hachure. The main criterion for the Chaco style is the use of composing lines that meet the framing lines at an angle. Sometimes the composing lines cut ebliquely across the field enclosed by the framing lines, at others they were drawn straight across it. Both forms may be found in a single design. The Little Colorado or Tularosa hachured designs are characterized by composing lines that run parallel to the framing lines. This has been called longitudinal hachure in contrast to the Chaco oblique hachure.* The earliest example of the style in the present series is a pitcher from the initial phase of the Developmental Pueblo period (pl. 16, e). The decoration on this vessel is in two bands, one around the body and one around the neck. The design on the body consists of six units separated by sets of vertical, parallel lines. The elements composing the units are solid triangles bordered by series of parallel lines. The triangles fill the lower left and the upper right corners and the bordering lines run parallel to the hypotenuse of each of the tri- angles with a space between the sets of lines. Due to the curve of the body wall this space is irregularly shaped in most of the units. One unit is incomplete. It contains only the upper triangle and its border- ing lines, the lower being omitted. The band around the neck of the pitcher is composed of two rows of opposed hachured and solid figures. One row is pendant from the rim, the other rises vertically from a line drawn around the vessel at the base of the neck. The figures are com- posed of longitudinal hachure with dotted triangles on either side at the base. SSeS TS LS > >= y SVs Senn er sé S S = ay > ~ SS SS SS SS] WAR SS x