anthropological records the archaeological ceramics of yucatan

ANTHROPOLOGICAL RECORDS

19

THE ARCHAEOLOGICAL CERAMICS OF YUCATAN

BY

GEORGE W. BRAINERD

UNIVERSITY OF CALIFORNIA PRESSBERKELEY AND LOS ANGELES

1958

THE ARCHAEOLOGICAL CERAMICSOF YUCATAN

BY

GEORGE W. BRAINERD


Vol. 19

UNIVERSITY OF CALIFORNIA PUBLICATIONS


Editors (Los Angeles): R. L. Beals, Harry Hoijer, C. W. MeighanVolume 19, pp. 1-378, figures 1-109, A-D, maps 1-24, charts 1-24, plates 1-3

Submitted by editors May 19, 1955Issued June 6, 1958

Price, cloth, $8.00; paper, $6.00

University of California PressBerkeley and Los Angeles

California

Cambridge University PressLondon, England

Manufactured in the United States of America

CONTENTSPage

Introduction 1...........,,,,,....... 1Earlier studies ................. . . .. . . . . 1Summary of chronology used in this report................................ 3Acknowledgments . , , .. 4

L Techniques and Materials Used,... 5Materials . 5Collecting techniques. 6Excavation and handling of sherds. . 8

II. The Site .10Cob a .10Yaxuna................11

Oxkintok .,.......... 13Dzibilchaltun........... 15Acanceh ....... 18Mayapan...... 21Ticul .23Dzan .. 23Mani .. 24"Mound near Merida" ceramics................................ 25The Puuc sites .............. 26Holactun (Xcalumkin) .32Chiche'n Itz .34

III. Ceramic Descriptions .46Vessel shapes, details of form............................... 47Ware descriptions............................. , . , . . 48

IV. Chronological Analysis............................. , , . ,.,.... 60Theory .... 60Nomenclature in this report..60Stratigraphy......................... 62Ceramic evidence bearing on the Maya-Christian

calendric correlation.....................,,.,.,S64V. The Development of Yucatan Ceramics: A Trial Formulationl.. 66

Pottery-making in modern Yucata'n,,. 66Archaeological materials ......................... , . ..... 69Historical account .71

VI. Yucatan Culture History in View of the Present Study .89Formative stage .89Regional stage.90Florescent stage.............. 92Mexican stage .94Post-Conquest stage .96

Notes ... 97

Bibliography .......... . . . . . . . . . . 101Figures ... 107

Maps .339

Charts ..... 361

Plate s....... 375

[iii]

THE ARCHAEOLOGICAL CERAMICSOF YUCATAN

BY

GEORGE W. BRAINERD

INTRODUCTIONThis monograph is the first of its kind to be devoted

to Yucatan Maya ceramics. Yucatan has for more than

a hundred years been notable for pre-Columbian archi-tectural remains which are among the most spectacularin the New World.1

The reconstruction of the history of the area hasheretofore leaned heavily on post-Conquest documen-tary sources and upon the dates which have been deci-phered from Maya inscriptions, both more readilyavailable sources of information than are usually ob-tainable for prehistoric cultures. The weaknesses in-herent in these sources-confusion and contradictionsin the documents, and restrictive distribution in timeand space for the datable inscriptions-have left majorblank areas and disagreements in the framework ofYucatecan Maya history, and have yielded descriptivematerials weighted toward religion and politics to thenear exclusion of reliable information on such mattersas cultural dynamics, economics, and demography.

Archaeological ceramics are notably good workingmaterial for the recovery of chronology as well as

prime indices for the study of trade, culture borrow-ing, and social patterning. They also allow the datingof associated archaeological finds; thus, although theYucata'n ceramics are drab in comparison with themagnificent architecture and sculpture with whichthey are associated, they are in a very real sense thekey to the discovery of the history of the YucatecanMaya. The Maya were, in Morley's flamboyant butquite accurate phrase, "the intellectuals of the NewWorld." All archaeologists are of necessity romantics,and the chance of learning even a little about such an

advanced and exotic group as the Maya has lightenedthe labor of combing through hundreds of thousands ofgrayish potsherds and reams of gradually yellowingnote s.

EARLIER STUDIES

The setting of the Yucatan ceramic survey can bestbe given by a summary of the results of earlier at-tempts to formulate a historical framework for Yuca-tan. The Yucatan historical or legendary accountsdescribe the peopling of the area by two migrations,one from the east and one from the west (map 1). Thedatable Maya Initial Series inscriptions have been usedin conjunction with the documents to construct a sortof history of colonization, on the assumption that theMaya could be defined exclusively as the people whocarved the dates (Morley, 1946, ch. 4). The conceptevolved that these date-carving, 'Old Empire" Mayafor some reason, hypothesized by Ellsworth Hunting-ton (1914) as a change of climate, abandoned the sitesof the central Maya area in the Peten lowlands and,abandoning also the custom of carving their InitialSeries dates, gradually moved northward into Yucatan,

[1]

their travel being marked by the sites of the Rio Becand Chenes areas. Upon arrival, in a renaissance ofcultural activity, they built the famous sites of thePuuc region and of Chichen Itza and Mayapan. Thechiefs of these three settlements in confederationruled Yucatan, with Mayapan as capital according tothe chronicles, until they disagreed about 100 yearsbefore the Conquest. This period of federation wascharacterized by Mexican rulers and much Mexicaninfluence.2

Major changes in this reconstruction have gradu-ally been made as the results of archaeological find-ings have become available. In addition, attempts toread the abbreviated Maya dates which occur in Yu-catecan ruins have in some cases provided resultswhich fit archaeological finds, and readjustments inthe chronicle accounts have allowed their use toclothe the archaeological framework. The causes ofthe abandonment of the Peten have also been sub-jected to continued scrutiny, but no hypothesis is asyet universally accepted. The exact degree of inter-dependence of these kinds of evidence on Maya his-tory-architectural and ceramic sequences, Yucate-can Maya dates, and chronicles-is hard to assessfrom the literature on the subject, but it can be saidthat in general these reconstructions now can befitted together to give an integrated historical ac-count.

The earliest ceramic sequence reconstructed forYucatan was formulated by Vaillant in 1927 and wassummarized and somewhat clarified by him in 1935.3He found evidence of 'very late Old Empire influ-ence" followed by (1) a carved slate period, followedby a period (2) characterized by fine orange (of Islade Sacrificios type) and plumbate, then a period (3)characterized by incensarios and porous wares,which included the slipped lacquer wares. Periods(2) and (3) were found stratigraphically at ChichenItza; period (1) was placed between (2) and the latestCentral Maya area period, characterized by poly-chrome figure painting, on the basis of stylistic re-semblances between the carved slateware and poly-chrome figures. These periods were named byVaillant in 1935 as (1) Mexican Contact, (2) MexicanOccupation, and (3) Maya Reoccupation. This termi-nology was based on the then current chronicle re-constructions, supplemented by the presence of Islade Sacrificios Fine Orangeware and many architec-tural features as Mexican imports in (2), and thepresence of a redware superficially not unlikeClassic-stage Maya pottery in (3).

The interpretation of the chronicles was alteredto some degree and clarified by Roys (1933) who,as a supplement to his translation of the ChilamBalam of Chumayel, gave a chronology (p. 204), andin his description of the Hunac Ceel Episode con-cluded with an insight verified by later archaeologi-


cal findings: 'After the end of the Twelfth Century wehear nothing more of Chichen Itzaas an important po-litical power, but its sacred cenote continued to be acenter of pilgrimage down to the time of the SpanishConquest" (p. 181).

In 1931 Morley summarized the architectural find-ings of the Carnegie Institution staff at Chichen Itzabylisting a division of buildings there into three sequentperiods. He concluded (1931, p. 107) that the Chiche'nItzca Initial Series lintel which dates at 10.2.0.0.0 Maya,the extreme end of the central Maya occupation, musthave antedated all architectural remains yet found atthe site. He placed the better-known buildings in threechronologically sequent groupings, the first belongingto a 'Maya Period," the second to the transitional, thethird to a "Period of Mexican Influence. " The first ofthese periods is the architectural equivalent of Vail-lant's period (1) above, the third is Vaillant's period(2). On the basis of the excavation of the Temple of theWarriors, Morris (1931, pp. 165-177) was able to breakperiod (Z) into a stratigraphic architectural sequence,and Tozzer (1930) attempted a tentative chronologicalbreakdown of the same period on the basis of the cul-tural content and style of bas reliefs in "Toltec" build-ings at Chiche'n Itza'.

In 1931 Henry Roberts began the study of YucatanMaya ceramics, and continued with four field seasonsin 1932, 1933, 1935, and 1936 (see Roberts, 1931, 1932,1933, 1934, 1935). In 1933 he reported a ceramic se-

quence from Chiche'n Itza which was essentially simi-lar to Vaillant's but which established for the first timea definite contemporaneity between the Puuc ruins andthe earlier ceramics of "Maya" Chiche'n Itzg. Roberts'period I at Chiche'n Itza (corresponding to Vaillant's(1) as given above) is called (p. 87) "Pre-Mexican pe-riod with pottery similar to that of the Labna-Puucregion." In 1934 he verified Vaillant's stylisticallybased hypothesis that period (1) followed the late Cen-tral Maya occupation by the excavation of "Old Empire"style polychrome pottery in lower levels of testtrenches; in 1935, in a more detailed report, he speci-fied the finding of such pottery at Coba and Yaxuna indistinct strata, and at Kabah mixed with the earliestPuuc wares.4 In the same progress report he noted theabsence of plumbate and fine orange pottery in the Puucsites,5 claimed a Mexican period date for the Holactunpottery, and also referred to chronological subdivisionswithin the Puuc and Mexican periods.6

The dating of various inscriptions on Maya Chiche6nbuildings was attempted in 1937 by both Beyer andThompson. Beyer, relying on an assumed stylistic se-

quence based chiefly on varying forms of two glyphs,assigned to the inscriptions a chronological orderwhich he fitted to the chronicles. In several instancesthis sequence was in conflict with then known archae-ological placements, and additional discrepancies in ithave been discovered during the ceramic studies re-

ported here. Thompson, using a new theory of decipher-ment conforming to dating systems known to have beenin use in Yucatan at the Conquest, dated the Chiche6nItza and certain other Yucatecan inscriptions in a

shorter, earlier time span which in gener.al fits thearchaeological sequence. Both sets of dates discardedMorley's 1931 placement of the Chichen Itza InitialSeries lintel as earlier than the Maya (Puuc equivalent)period at Chiche'n Itza'. Morley's early placement ofthis date was made reasonable by the use of the12.9.0.0.0 Maya-Christian correlation, which was aban-doned by him and by most other workers during the1930's.

In this account Eric Thompson's 'Co6rdination ofthe History of Chichen Itza with Ceramic Sequences in

Central Mexico" (1941) must be mentioned. Thomp-son pointed out that plumbate pottery, which is datedon the basis of chronicles as not later than 1300 A.D.in the Valley of Mexico, is found in occupational de-posits of some of the buildings judged to be amongthe latest at Chichen Itza, and an unbroken plumbatejar was found in such a position as to suggest that itwas deposited after the fall of the Caracol Tower.With these data Thompson buttressed Roys' sugges-.tion (1933) of the early decline of Chichen Itza', andfrom his 1937 readings of Puuc and Maya Chichendates suggested dating schemes for the beginningsof the newly identified Toltec period in Mexico.

Thus at the beginning of this study, three sequentperiods had been recognized in the ceramic materialsof Yucatan, although they had not been detailed inpublication: (1) a Maya or pre-Mexican period repre-sented by pottery from the Puuc sites, and from the"Maya style" buildings at Chichen Itza; (2) a Mexicanperiod or period of Toltec influence, represented bythe majority of the pottery at Chichen Itza'; and (3) a

period variously called Maya Reoccupation, or MayaResurgence by Vaillant, and found in superficial,postoccupational deposits at Chichen Itza' and else-where. The first of these periods was believed eitherto postdate or slightly overlap the end of the InitialSeries period on evidence of Thompson's readingsof the Maya Chiche'n dates and, more firmly, byRoberts' 1935 determinations and by others fromSan Jose', British Honduras, published by Thompson(1939, pp. 149-150). A few ceramics datable to theearly part of the Initial Series period had been lo-cated in collections from Yucatan, but no earlierceramics had been recognized.

In his "The Maya New Empire" (1938), Morleyelaborated his earlier reconstructions of the historyof this part of the northern Yucatan Peninsula. Hedefined Maya culture as characterized by "the typi-cal architecture and unique hieroglyphic writing,"which he believed were introduced together from thesouth in the sixth century A.D. as evidenced by theearliest northern Initial Series stela at Tulum, fol-lowed by later dates also in the eastern part of thePeninsula. He noted that the Yucatecan inscriptionsare late and clearly related to those of the Peten,and therefore must be derived from them, and as-

sumed that the architecture was likewise derivedfrom the south. The distinctive, indigenous characterof the Yucatan slateware pottery was discussed byMorley, who suggested that it denoted the early pres-ence of a people who were probably Maya-speakingagriculturalists of Maya racial stock who adoptedsouthern stone architecture, hieroglyphic writing,astronomic learning, and religious philosophy (Mor-ley's 'Maya Culture") but retained their local ce-ramic craft, and developed the Yucatecan florescenceof the Maya culture. Morley documented this storyof the introduction of Maya culture into Yucatin byLizana's account of a Maya tradition of the populatingof Yucatan by two migrations, a "lesser descent"from the East, and a "greater descent" from theWest, and later (1946, p. 76) fitted the 9.2.0.0.0Oxkintok lintel (Pollock, 1940) to this reconstruction.In his "The Ancient Maya" (1946), Morley amplifiedthis reconstruction, suggesting that the Chenes ruinsrepresent a second wave of migration northwardfrom the Peten (pp. 80-81) and that the sites of thePuuc region, of Maya Chiche'n Itza, and of Mayapanwere all settled at about 10.9.0.0.0, 1000 A.D. (seehis Table V) by people of Mexican extraction. Thisperiod, which is made to correspond to the archae-ological Puuc Period (period 1 above), is followed

2

BRAINERD: THE ARCHAEOLOGICAL CERAMICS OF YUCATAN

at about 1200 A.D. by a period of the Ascendancy ofMayapan (archaeologically the Mexican period, period3 above) which ends with the fall of Mayapan ca. 1441A.D.7

Morley's scheme depends upon Initial Series Mayadates and documentary sources for its chronologicalframework, and upon dirt archaeology only for detail.The chronological placements of his Puuc and Mexicanperiods do not fit the archaeological framework, andhis cultural dynamics, notably the relationships be-tween the Peten, Chenes, and Puuc areas, and amongUxmal, Chich6n Itza, and Mayapan, also do not fit theceramic evidence.8 Thompson (1945), using many ofthe same data, has produced a scheme archaeological-ly more acceptable. As may be seen later in this re-port, the three sequent ceramic periods first hypothe-sized by Vaillant have fitted into the longer time scaleunearthed during the ceramic survey.

SUMMARY OF CHRONOLOGY USED

IN THIS REPORT

To aid in the reading of the subsequent sections, thechronological outline which has emerged from the ce-ramic survey will be summarized here. A chronologi-cal chart is given as chart 22; the detailed evidencefor its formulation constitutes the bulk of this report.

Because of the large area covered and the consider-able time depth reconstructed, I have decided to usegeneralized names for major sequent culture divisionsin Yucatan, and have fitted the collections from varioushorizons and sites into this more general scheme.These divisions are meant only for Yucatan, and fitcertain peculiarities in Yucatecan cultural history.They differ somewhat both in name and duration fromothers used elsewhere in the Maya area. The majordivisions have been called stages in accord with Dr.Kidder's usage (Kidder, Jennings, and Shook, 1946, pp.1-9). Their names have been chosen in an effort to de-signate general characteristics of each, and at thesame time to avoid confusion with the increasing num-ber of such terms now in use in Mesoamerican archae-ology.

The sequence chosen runs as follows:Yucatan Formative stage.-This stage is divided

thus far into three substages, Early, Middle, and Late,on the basis of evidence for distinctively separate ce-ramic assemblages running sequentially in that order.The Early Formative substage was found only in theMani Cenote; the Middle Formative substage has beensampled only in the Chenes area, and thus is not de-scribed in this report. The placement of my Late Form-ative substage was first made by the establishment ofclose correspondences with the Uaxactun Chicanelphase which dates pre-Initial Series (or pre-Classic)at that site, and was later confirmed stratigraphicallyfor Yucatan. Some ceramic diagnostics are the almostexclusive use of monochrome in slipped wares; theprominence of flat-bottomed, allover - slipped bowlswith flaring sides and heavy, out-turned rims; the pres-ence of tubular-spouted small jars. The dates of thisstage are estimated roughly as 1500 B.C. to perhaps100 A.D.

Yucatan Regional stage.-Several distinctive ceram-ic assemblages of this stage are known. It shows sucha variety that sharp regional as well as chronologicaldifferences are evident. The Regional Flaky Redwareassemblage, with its included Incised Dichrome andPolychrome, is the earliest of these; the OxkintokMonochromes seem to be a distinctive western mani-

festation and must overlap at least the latter part ofthe Flaky Redwares in time. The Regional Monochromesof Acanceh and Yaxuna seem to date later, in majorpart with Tepeu 1 at Uaxactun (9.9.10.0.0-9.12.10.0.0Maya calendar, 623-682 A.D.). Following them strati-graphically at Acanceh, and also found at Dzibilchal-tun, is Red on Thin Grayware, which occurs in asso-ciation with slatewares. Some of the general ceramiccharacteristics of the stage are the use of red, orange,and cinnamon-colored monochromes, and of polychromesin some regions. Distinctive vessel forms includecurved-bottomed bowls, many with lower face un-slipped, and some with a ridge or flange low on thecurved side. Small teat-shaped tripod legs are thecommonest support, but there is wide variety. Slippedjars commonly have a short, sharply outsloping rim,and often a low-placed angular shoulder. There is con-siderable regional variation in both wares and formsduring this stage, as well as evidence of chronologicaloverlapping in some regions between this stage andthe next in sequence. With reasonable certainty, thisstage lasted later than 9.16.0.0.0 (751 A.D.) in north-west Yucata'n. Its duration, even its presence, is un-known from the sites of the Puuc, Chenes, and Rfo Becareas. Its beginnings in northern Yucatan must havedistinctly preceded the advent of Tzakol-style poly-chrome, estimated in the central Maya area at A.D.278.

Yucatan Flor escent stage.-Ceramic as semblagesof this stage are dominated by the slatewares. Thinand Medium Slatewares, accompanied by Medium Red-ware, seem to have predominated in the assemblagesall over the northern part of the Peninsula during thelater part of this stage. To the southwest another dis-tinctive ware, Holactun Slateware, is common. Cer-tain slateware vessel forms believed to belong to theearly part of the stage were used contemporaneouslywith Regional stage Red on Thin Grayware at Dzibil-chaltun and Acanceh, thus suggesting an origin atleast as far south as the Puuc area for the slatewares,with a later diffusion into northwest Yucatan. In thesites sampled thus far, this stage can be dated as be-ginning earlier than 9.16.0.0.0, but there is as yet noclear evidence of a beginning date earlier than9.12.10.0.0 (the beginning of Tepeu 2 at Uaxactun;no Tepeu 1 pottery has been found in demonstrablypure Florescent deposits). This stage is distinguishedfrom Regional by its slatewares; its late end is de-limited primarily by changes in forms caused byMexican innovations, and by changes in the types andorigins of its tradewares, although gradual warechanges also occurred. Its end is estimated as be-tween 10.3.0.0.0 and 10.8.0.0.0 (889-987 A.D.), andwas probably nearly contemporaneous all over Yuca-tin, although there may have been a slight chronologi-cal lag in the Puuc region in comparison with Chich,6nItza.

Yucatan Mexican stage.-This stage is divided in-to three substages, each recognizable by a widespreadand distinctive ceramic assemblage, and is followedby a century which, although historically documented,is not thus far recognizable ceramically.

The stage was initiated by the Early Mexican sub-stage. Medium Slateware rather similar to that ofthe Florescent stage, but inclined toward a white,opaque slip, dominates the slipped wares, accompa-nied by Medium Redware but with no thinwares. Thesmaller vessel forms and the decoration show strikingchanges from Florescent styles; most innovations are

unquestionably copied from X Fine Orange, a potteryimported from central Veracruz, which sometimesruns as high as 10 per cent of samples. Plumbate

3


pottery occurs as a rare trade item. This substage isrepresented in our collections by materials almostexclusively from Chichen Itza, where all the "Toltec"period architecture is contemporaneous to it.

The Middle Mexican substage is distinguished bythe advent of Coarse Slateware as the major slippedpottery. This ware differs from Early Mexican Medi-um Slateware in the employment of a coarse-textured,calcite-tempered paste. The opaque, white slip contin-ued to be used, and sloppy, rather exuberant, blackpainted decoration. X Fine Orange was still importedduring this substage; plumbate importation had prob-ably ceased. The subphase saw the building of a fewminor architectural additions to the Toltec periodbuildings at Chichen Itza, the beginnings of the mainMayapan occupation, and occupations at Dzibilchaltunand Acanceh.

The Late Mexican substage is characterized by thepresence in quantity of Coarse Redware, the paste ofwhich is similar to that of the earlier Coarse Slate-ware. The vessel form changes through the Mexicanstage are gradual, but several of the Late substageforms are quite distinctive from those of the EarlyMexican substage. Mayapan Fine Orangeware, prob-ably traded from the Tabasco-Campeche coastal area,has influenced local form and decoration. During thisperiod a new type of unslipped jar appeared, distin-guished in both form and finish from the precedingform which had lasted with but minor changes from theFlorescent stage. A new unslipped cooking pot, thecaldero'n, also began to be used during this period andhas lasted until now. Late in the substage the manufac-ture of figurine incensarios was introduced, almostcertainly from the west, suggesting evangelization toa new set of religious concepts.

At this point there is an uncertainly documented gapof about a century in the ceramic sequence, followingthe historically documented fall of Mayapan and pre-ceding the Spanish conquest. This period has beentermed the Interreanum by Morley.

Post-Conquest stage.-This stage is characterizedby an impoverishment of the pre-Columbian repertory,and by a few innovations. Coarse Redware is still theprevailing manufacture. The unslipped jar does notseem to have survived the Conquest; the calder6n musthave replaced it. Two new types of decoration, paintingof conventionalized floral motifs in brown on an un-slipped exterior, and embossed inscriptions in Euro-pean script on an unslipped band, began at some timein this stage. The pre-Conquest practice of painting intempera colors on previously fired unslipped potterypersists on modern Lacandone incensarios, and uponwhistles used in Yucatecan fiestas. Animal figures aremade each in several molds and luted, perhaps in anold technique. Neither slip casting, glazing, nor anywheel techniques have become a part of the Yucata'nceramic folk craft. Our chief archaeological samplesof this period come from the Mani monastery, datingback for an unknown span from about 1830 A.D. Small-

er samples come from the Mani and Telchaquillo ce-notes, and a small but certainly early sample fromDzibilchaltun.

ACKNOWLEDGMENTS

This study was started in December, 1939, and be-gan my acquaintance with Maya archaeology; my back-ground at that time was in the archaeological ceramicsof other areas. My debts have been great, first to Dr.A. V. Kidder who started me and has since sustainedme with constant understanding and encouragement.Maya archaeology, I have found, is an unusually com-plex field. I owe much to characteristically candid andgenerous discussions with those members of the Car-negie Institution staff with longer experience in theMaya field, particularly to Anna Shepard, who intro-duced me to Yucatan ceramics.

In Yucatan Don Rafael Regil graciously allowed meto photograph and record his collection. Sr. Jose A.Erosa Peniche, resident inspector of the InstitutoNacional de Antropologia e Historia, Direci6n deMonumentos Prehispanicos, was helpful in arrangingpermissions. Sr. Alfredo Barrera Vasquez as direc-tor of the Yucatan State Museum, and as my guide ontwo field trips, gave valued time and advice. Sr. RaulPavon Abreu extended the courtesy of the CampecheMuseum and allowed me to study collections fromJaina. Many people in Yucatan, both property ownersand laborers, gave the constant courtesy and interestso characteristic among residents of the state. Timo-teo Canul and Eugenio Mai increased my efficiencyby devoted labor as field foremen, and in handling ofceramics. Sr. Juan Germon did valued work in 1942in the sorting aned tabulating of ceramics.

Of the several draftsmen who prepared the draw-ings for reproduction, Sr. Isaac Esquiliano, Mr. ElmerReising, Mr. Leo Prince, and Mr. Robert Quinn wereparticularly helpful. Miss Tatiana Proskouriakoff al-so prepared several pages of the more difficult draw.ings with unusual skill, Mrs. Betty Bell improved themanuscript by an editorial reading, and Mrs. MargaretHarrison checked the bibliography.

I am indebted to the Southwest Museum for the useof library facilities, and for a month's time which wasallowed me toward the completion of this manuscript.I owe thanks to the Me'rida State and Federal Museums,the Peabody Museum at Harvard University, the Uni-versity Museum at Philadelphia, and several ownersof private collections for permission to illustratespecimens in their possession. The Committee onResearch, University of California, Los Angeles, al-lowed me grants that permitted statistical studieswhich, although they are not published here, haveclarified my understanding of Yucatan ceramics. Tomy wife I owe not only the considerable effort she hasspent on various tasks connected with the field workand report, but her forebearance during the fourteenyears it has taken me to complete it.

4

I. TECHNIQUES AND MATERIALS USEDThe unusually varied nature and sources of the ce-

ramics discussed in this report make a careful listingadvi sable.

MATERIALS

I. Pottery from archaeological excavation (almost ex-clusively sherd material). See map 2 for locations.a. Collections resulting from the clearing and re-

pair of the various buildings at Chichen Itza' bystaff members of the Carnegie Institution. Bothwhole vessels and fragments resulted.

b. Potsherds from trenches at Chich6n Itza dug byH. B. Roberts. This material was meant to aidin the stratigraphic placement of the above-floorsamples which form the bulk of pottery obtainedfrom building excavation and restoration. TheRoberts collections are relatively small.

c. Sherd collections and whole specimens fromwidely scattered localities, obtained by workersin the course of the Institution's studies in Yuca-tgn since 1924.

d. Collections from trenches dug in ruins investi-gated by the Architectural-Ceramic Survey dur-ing the years 1932-1936. Collections were dugby Roberts from the Puuc sites of Uxmal, Sayil,Labna, Kabah, Sabacche, and from Yaxuna, Coba,and Holactun. Several collections came fromsites in the region between the Puuc hill rangeand the city of Campeche, but their smallnessand heterogeneity has made their study imprao-ticable at this time.

e. Material dug by the author from trenches in var-ious Yucatan ruins and from neighboring watersources. These excavations were conducted aft-er the study of all previously excavated material,to the end of determining stratigraphy thus farlacking, and of covering regions and time peri-ods only hinted at in the previously mentionedcollections which were almost all from ruinsbearing impressive standing architecture. Pot-tery from Oxkintok was obtained in 1940, andfrom Yaxuna, Dzibilchaltun, Acanceh, Mayapan,Ticul, Dzan, and Mani and environs in 1942.Nearly all of the ruins sampled in this work pro-duced evidence of long occupation. An effort wasmade in each case to assure as representativea sample as possible for each site.

Most of the whole vessels excavated by the Carne-gie Institution are divided between the National Mu-seum in Mexico City and the Federal Museum inMerida. A large sample of the sherd material isstored in Me'rida; small sherd study collectionshave been furnished interested institutions in andoutside of Mexico.

II. Whole vessels from museums and private collec-tions were used to supplement the excavated mate-rial in the determination of vessel shapes and de-signs often left doubtful because of the badly brokencondition of the documented excavated sherds. Afew well-documented museum collections furnishedadditional information of regional and chronologicalvalue. The principal sources were as follows:

Collections at Me6rida, Yucatan, includingthose of the State Museum, the collection of the

[5]

late Don Rafael Regil, and other smaller pri-vately owned collections. The majority of thespecimens in these collections are undocu-mented as to provenience, the remainder aredocumented only to a general geographic loca-tion. About 1,000 specimens were studied andphotographed by the author, and a number ofthese have been illustrated in the text. TheState Museum collection of whole specimensfrom graves at Dzebtun gives a valuable pic-ture of pottery of its period, and samples ofall types have been illustrated.

Peabody Museum of Harvard University. Amajority of the specimens here came from ex-cavations at Chich6n Itza and in the Puuc re-gion, made by E. H. Thompson about 1900. Col-lections from Labna, Mound 6, and the locallymade pottery from the Sacred Cenote at Chi-che6n Itza, have been illustrated as groupscharacterizing their respective periods. Con-siderable pottery from Peabody Museum hasbeen illustrated as supplementary to sherdmaterial, as have a few previously unpublishedspecimens from private collections in theUnited States and Mexico. Specimens alreadyillustrated and described in the literature havebeen discussed and reillustrated wherever di-rect comparison seemed advisable.

III. Notes made by G. C. Vaillant and by H. B. Robertsand held by the Carnegie Institution have been usedfor the description of certain important collectionssince mixed or lost. These citations are individual-ly attributed in the text.

IV. Maps locating all pottery excavations as nearly aspossible are included to indicate the exact sourceof the material. The majority of these have beenredrawn from more detailed maps by CarnegieInstitution workers, some of which are thus farunpublished. It should be emphasized that themaps as here published make no pretense of ade-quacy in architectural or other specialized details,and were drawn primarily to locate the potterytrenches in relation to the other features of thesites. The source is shown in the caption of eachmap. It is hoped that these maps will eventuallyallow the ceramic data to be linked with their ac-companying architectural and other cultural mani-festations, without which the descriptive archaeol-ogy of Yucatan is now sadly one-sided. Only aftersuch use will the chronological scale, which hasbeen the primary purpose of this study, achieveits full application.

As may be seen from the sources enumeratedabove, the report is to some extent a patchwork con-structed from objects obtained at various times inthe course of work aimed toward various purposes.Despite Roberts' considerable labors to integrate thematerials available to him, a complete reworking ofall material by the present author was necessary dueto two causes. A large proportion of the excavatedpottery in storage at Chich6n Itza had become mixedor lost its labeling during the years 1936-1940. Thecollections from Chich6n Itza, totaling perhaps100,000 sherds selected from collections of severaltimes that number, were in storage at the site in


1940 when this study was begun. The climate of Chi-ch4n Itza is poorly adapted to the dead storage of per-ishable materials. Although the pottery had remainedin good condition, the wooden boxes containing it hadbeen riddled by termites, and the whole collection hadbeen appropriated as dwelling quarters by the abundantrodent, reptile, and insect life which characterizes theregion. More serious was the destruction of labels.The boxes had been labeled with filing cards thumb-tacked to their fronts. All paper sizing, save that un-der the tack heads, had been eaten from these cards,presumably by silverfish and cockroaches. With thepaper sizing went the penciled labeling, except wherethe pencil point had dug very deeply. Before my arriv-al, an effort to save the labels had been made by sub-stitution of a series of brass number tags, and thus aconsiderable amount of information still remainedstored on paper labels in a safe. The labeling on brownmanila bags had escaped damage, although rodents hadin many cases cut up the bags. Even worse than theloss of the labels was the mixing of collections throughthe disintegration of their containers. Sherds frombags and boxes had often dropped through into the col-lections below them.

Despite these ravages, a considerable number ofcollections were still identifiable. However, at leastin some cases, certain fine wares and noteworthy spec-imens had been extracted from their collections andsorted by categories, and few of these sherds had beenseparately labeled. This fact introduced disquietingdoubt as to the validity of presence-absence criteriain the samples. To this were added questions as towhether a part of any sample, previously sorted, wasamong the unlabeled collections, thus leaving a se-lected sample from which I would extract a false con-clusion.

With a keen sense of the irreplaceability of thesecollections, I spent several months in analyzing thetraces of inscriptions on labels, correlating Roberts'tabulations with the original material available, andchecking his field notes, which for the earlier part ofhis study were excellent, to determine what collectionswere unaccounted for. Many collections were reassem-bled by matching fragments between them.

Parts of Roberts' notes had been misplaced, includ-ing the illustrated keys to most of his tabulations, andRoberts was unavailable for consultation to clarify hisincomplete notes. After several months spent largelyin tracing and relabeling collections with the aid ofwhat notes were available, I retabulated all the potteryof traceable provenience.

Upon taking stock of information available from thepottery samples, it became evident that large colleo-tions belonging to two major architectural periods con-stituted the great bulk of the sample (see periods (1)and (2) above), but that tantalizingly small collectionsobviously belonging to several quite separate horizonswere also present. This situation seemed to cry foradditional excavation aimed toward a general samplingof the northern part of the peninsula, with emphasis onsites which seemed different from those already rep-resented.

The first attempt to round out the samples wasmade in 1940 when, taking advantage of the presenceof the Yucata'n Architectural Survey at Oxkintok, I dugtest trenches there (Brainerd, 1940). The season of1942 was spent covering scattered sites, with particu-lar emphasis on obtaining materials from little-knownceramic horizons.9 During the three months spent atthese excavations, we were fortunate in documentingseveral new ceramic periods, thus completing in skel-eton outline a chronologic sequence extending from

times well antedating the earliest Maya stelae untilthe Conquest.

In June, 1943, work on the Yucatan ceramics washalted by the war and has been continued only at vari-ous intervals since the autumn of 1946, when I beganteaching and museum work in Los Angeles.

COLLECTING TECHNIQUES

The methods used in gathering pottery for thissurvey were governed by factors which it may bewell to discuss.

The terrain of Yucatan is flat to gently rolling inmost parts, and has but a thin coating of soil whichis burned over every few years in the course of agri-cultural procedures. Yucatan is a Karst area withunderground drainage; there are no surface streams.The country, when not cleared for agriculture, iscovered by a mass of thick, thorny, low-lying growth,difficult to see through and necessitating the constantuse of a machete to penetrate. Pottery can rarely befound on the surface of the ground; burning and weath-ering of the pottery in the tropical climate, as wellas the difficulty in seeing it through vegetation, makessurface collecting impracticable. The scarcity andconsequent value of soil has made the deposition ofdeep, well-stratified refuse deposits a rare occur-rence.

The characteristics of the Yucatan Maya civiliza-tion further limit and determine archaeological pro-cedures. The relative simplicity of the pottery incomparison to its accompanying elaborate architec-tural remains should be emphasized. Finer chrono-logical subdivisions would be easier to determinefrom more elaborate pottery. It is very clear thatthe pottery we are studying is the common ware ofthe population, and that during most of our time spanpottery-making was not a major art. Although finelymade wares are present in varying quantity, they al-ways constitute but a small percentage in collections.

The masonry buildings, which are the most strikingfeature of Maya archaeological sites, were commonlyplaced on massive raised masonry terraces, whichwere usually plaster-floored and were kept clean.The population may be assumed to have lived simply,as they do now, in flimsy pole and thatch huts. Somewere probably near the temples but the majority, dur-ing most time periods, seem to have lived at a dis-tance, convenient to their fields. Owing to the Mayaagricultural system, the fields were necessarilyscattered more or less evenly over the whole country-side. House sites are difficult to locate, and theirshort occupation, the extremely shallow earth, levelland, and luxuriant vegetation seem to have preventedthe accumulation of localized domestic dumps, andthe climate and shallow soil cause rapid disintegra-tion even of pottery.10

In Yucatan, doubtless owing to the scarcity andconsequent value of earth and to the enormous quan-tity of limestone easily available from the raisedmarine reef which composes the peninsula, the fillsof the Maya temples and terraces of most periodswere made of clean chunks of limestone unmixedwith dirt or debris, thus eliminating another possiblesource of pottery of particular chronological value.The long-continued use of ceremonial centers hasaided the obtaining of stratified sequences, but hasmade more difficult the obtaining of short-period de-posits. It has caused a mixture of earlier pottery toappear in nearly every collection, and has weightedour samples toward the later part of the time scale.

6


The main sources of archaeological pottery wereas follows:

In ruin debris and on floors of rooms. -Althoughmuch material of this sort resulted from the extensiveCarnegie Institution program of building restorationat Chiche'n Itza, the quantity of pottery relative to de-bris moved was small, and very few short-period col-lections resulted. This was due presumably to the longperiod of occupation permitted by the durable construc-tion of masonry buildings, to the undesirability of ex-cavation in deposits sealed by later construction whenthe preservation of the later structures was a primarygoal, and to the fact that many buildings stood long aftertheir major period of use. A constant, long-continuedsecondary use which often culminated in conversion tosimple shrines or temporary shelters seems to havebeen the rule. It is likely that rooms were kept cleanduring their primary period of occupation, and thereis almost no evidence that masonry buildings wereever constructed for domestic use.

Stratigraphic trenches due in inside corners at thebase of terraces (for example, see map 7). -Suchtrenches provided the largest and most heavily concen-trated collections found at ruins. They represent theonly true refuse dumps found. Pottery broken on ter-races and in buildings seems to have been thrown offthe terrace sides into these relatively inconspicuouscorners. Stratigraphy is not found in the majority ofthese collections, probably due to salvage of earth foragriculture as well as to the Maya practice of constant-ly enlarging buildings and terraces, using previousstructures as cores. Frequent additions to terraceswould cause the covering of old dumps and the begin-ning of new in more peripheral positions, and thus thedeposits obtained in the superficial excavations under-taken in this survey usually represent only the latestoccupational phase of the site. Evidence along severallines points to long earlier occupations being but poor-ly represented in the ceramic samples.

Shafts sunk through bases of early building plat-forms.-The largely demolished early 'acropolis" inAcanceh (see map 8b) was trenched along its presentperiphery. The previous removal of a large part of thepyramid for use as building stone made its core ofearly deposits easily accessible to us. These wererich and well stratified. Here, as elsewhere in Region-al stage ruins, we noted the use of cultural debris asfill in the construction of subsequent buildings. Thispractice seems not to have been followed in Florescentor later stages.

Chultun deposits.-Chultunes are jug-shaped cisternsdug into the limestone to collect water from surround-ing paved floors. They often contain a mixture of waterjars broken or lost in use, and of general rubbish; stra-tigraphy was seldom apparent in those dug. In the chul-tunes we found only debris dating from the Florescentstage and occasional fragments from later periods. Wehave records of chultunes in numbers only from thePuuc area and from the north edge of the Puuc rangeat Ticul, Dzan, and Mani; they are very common inthese regions. It seems likely that in Yucatan they arecharacteristic of, and limited to, the Florescent stage;in sites of this period they are most common when nonatural water sources are nearby.

Deposits in cenotes and in caves containing springs.-In the absence of streams and lakes, cenotes andcaves were the principal source of water in areaswhere they are found. The open, large-mouthed cenoteis less common in some areas investigated than areextensively ramified caves. In the branches of thesecaves, often at considerable distance and depth fromthe mouth, are found springs which seem to lie at the

ground water level (see map 12). Along paths leadingto the water, there are deposits sometimes as muchas two meters deep, containing a large proportion ofbroken pottery. Stratigraphy is sometimes mixed,sometimes excellent; pottery horizons occurring sel-dom or not at all in our neighboring ruin collectionsare represented by collections found in stratigraphicsequence. A drawback to the study of cenote collec-tions, particularly evident in certain periods, is aprevalence of water vessels over other shapes. Thismakes cenote collections somewhat difficult to com-pare with ruin samples of the same date. However,the size of collections and length of occupation repre-sented by the stratigraphy of cenote excavations makethem our best source of stratified deposits, and theyare likely to represent the complete ceramic historyof the sites in which they are found.

Burial and cache pottery.-This pottery, includingnearly all of the whole museum specimens so usefulin studies of pottery shapes, seems to have come fromtwo main sources, buildings and cenotes, as parts ofcaches or of secondary burials. A number of suchspecimens were found in Carnegie Institution excava-tions at Chiche'n Itza' and are documented. The major-ity of those from elsewhere are in museums or pri-vate collections. They have been discovered duringthe destruction of ruins for stone and lime by the in-habitants of the country. It is impossible to learn theexact provenience of most of such specimens, or theexact associations in which they were found. Many ofthose which are documented were found with disartic-ulated human bones in or near them, and sometimesassociated with flaked stone artifacts, shell ornaments,and jade. Thus secondary burial must have occurredin Yucatan. Although the vessels in these tombs andcaches are usually of better than average quality andsometimes include elaborate pieces, they do not con-stitute an entirely separate repertory of vessels madeonly for funerary or ceremonial use. Many of thesevessels therefore can be used in direct comparisonwith documented sherd material. They provide valu-able information on decoration and vessel form other-wise difficult to obtain.

No pretense can be made that our ceramic collec-tions are exhaustive either as to period or as to typeof locality, and certainly not every period is repre-sented in several areas. However, because of recog-nizable continuity of techniques and styles throughour sequences, we are confident that no major hori-zons within them have been missed. One type of localecompletely slighted in the survey has been the coastalshell heap. Such heaps occur on the north Yucatancoast and should certainly be investigated in the fu-ture. Also, it is certain that early horizons have beenslighted. Many ceremonial sites were in continuousoccupation for 2,000 years or more. The scarcity ofsoil prompted later people to reuse early rubbish.The relatively heavy recent occupation of Yucata'n hasdestroyed much evidence, particularly at major popu-lation centers. Early material is harder to obtain thanlate.

It should be reported here that I do not believe theYucatan ceramic sample is biased to any considerabledegree by containing artifacts used only by membersof the upper class of Maya society."1 The bulk of thepottery was dug from where it could be found-nearthe ceremonial centers. But pottery from watersources tallies well in general complexity of formand craftsmanship with that at the masonry ruins. Iam convinced that the refuse dug was left by run-of-the-mill Maya pottery users. There will be somediscussion later in the text on such inferences of

7


Maya social structure as have been attempted. Fortu-nately for the purposes of this monograph, which arein major part frankly descriptive and chronological,enough elaborate pottery, both local and trade, wasfound to allow several of the close stylistic anchoragesso necessary to the fixing of the time-space frameworkwhich is a prerequisite to all proper culture reconstruc-tion.

EXCAVATION AND HANDLING OF SHERDS

The criteria for the placing of pottery trenches onsites have been described. Trenches were excavatedin cuts of 25 cm. to 50 cm. depth, generally followingwhatever evidence of depositional plane was available,including the top earth level, the angle of rest of sherdsand flat stones, and earth strata showing in trench sideprofiles. Unfortunately, few trench walls showed moredetail than a humus line, possibly due to the shallow-ness of trenches and to root disturbance by the rankvegetation. Evidence of floors was noted carefully whenpresent, and new collections were always started be-low floor level.

Sherds were sacked in the field, labeled as to trenchand cut, and then washed and stamped individually withprovenience, either at the site or later in Me'rida.Stamping was done with movable rubber letters in a

small holder, with stamp-pad ink. This marking is rap-id, does not require skilled or even literate labor, andshows satisfactorily on most Yucatan pottery. Afterseven years' storage in the light, fading was observedon some labeling. Carbon ink is recommended in futurework to prevent this difficulty.

No screening was done at the trenches; the charac-ter of the local soil makes this operation difficult, andthe kinds of remains characteristic of the area are

easily located in earth without screening. The workmensaved all sherds they could see. In later sorting, allsherds under one inch in longest dimension were dis-carded, except for sherds of rare or foreign wares. Itis believed the samples remaining are but little biasedby selection.

Various procedures to avoid sorting errors previous-ly described by the author (Beals, Brainerd, and Smith,1945, pp. 164-166) were used, the most important ofwhich was the sorting of types from the combined lotsobtained from each site. This procedure, permittedonly by the individual labeling of each sherd, is in myjudgment the only way to obtain the objective, unbiasedsorting necessary to close chronological work.

The initial stage of ceramic analysis in a new area

is necessarily taxonomic. The procedure followed was

that of isolating, and objectively sorting, the materialinto successively smaller categories, usually mutuallyexclusive. The final result is a large series of sortingcategories, each of which in theory depends upon a suf-ficient number of characteristics to assure against thelikelihood of its duplication by chance in any culturalcontext not its own.

After an initial sorting into wares, based upon thesurface appearance of the pottery, the categories were

based when possible-and it was seldom found impos-sible-on whole vessel forms, so that the classificationwould allow direct comparison with whole vessels. Thisprocedure, perhaps not necessary to the original chron-ologic study, later allows much better comparison withpottery of adjoining areas, the study of change ihroughtime in pottery shapes, and estimates of the place ofceramics in the culture. Unusually complete and nu-

merous records of pottery forms were made as illus-trations, including details such as sherd orientation,

vessel radius, and extent of slip when present. Some6,000 specimens are illustrated in the plates. When-ever possible, whole vessels or reconstructed vesselshave been introduced on plates with their sherds toaid in visualization of whole specimens from the docu-mented sherd material, and a series of summarizedpages of whole vessel shapes has been included (seefigs. 104-109).

During the classification, description, and tabula-tion, every effort was made to attain objectivity. Thefrequent intergrading found was described, is notedboth in the illustrations and in the text of the typology,and is further analyzed in sections on various attri-butes used in the sorting. Most of the types representconcurrent variations in more than one recognizablesorting criterion; variations in a single criterion nor-

mally were handled by a series of drawings showingform, slip, or diameter variation, or descriptively,as by a series of color readings. The theory underly-ing this sorting system is that while variations in allsorting criteria are normal in all groups of artifacts,concurrent variations of several technologically unre-

lated criteria are evidence of a conscious effort tomake recognizably different kinds of artifacts, or ofa cultural difference between the makers of the arti-facts so distinguishable. Interrupted variations, attri-butes which showed bimodal distribution curves in thematerial, were also considered valid reasons for sep-arate pigeon-holing, since they suggested distinct dif-ferences in procedure of manufacture, and thus mir-rored either conscious choice differences in the cul-tural conditioning of the makers, or differences inmaterials used.

The text of the section on ware descriptions is notaimed at exhaustive description of each category butat concentration upon characteristics showing varia-bility in the samples, and upon those chosen as diag-nostic in the classification. On the other hand, a con-stant guard was kept against basing the classificationupon any preconceived notion of what was importantor unimportant to the craftsmen. Aid toward this at-tempted objectivity was given by the elimination as

sorting criteria of certain types of variation knownto be inherent in the process of manufacture of pot-tery of this general class-variations known to be de-pendent upon a poorly controlled technique, ratherthan upon difference in technique.

The sherd tabulations, as may be guessed fromthe above description, are unusually voluminous.Eighty separate sorting groups were not unusual forthe pottery of one site; over 200 were used for Chi-che6n Itza. Even then, certain gaps in description were

keenly felt in the later analysis. One reason for suchfine typological subdivision was the fact that thesherds, having been left in Yucatan after tabulation,were not available for rechecking during the remain-der of the analysis. Another reason for numerous

categories was the attempt to keep rim sherds iso-lated from body sherds in order to allow comparisonwith results of other workers, some of whom use onlyrims for counts. Since the typology was formed, as

are all typologies, before an understanding of the cul-ture could be reached, an exhaustive record of allpossibly significant factors was desired. Such a rec-

ord was obviously not attained, and can never be at-tained. It was felt that no loss due to too much subdi-vision was possible, since the categories could belater combined at will, as they invariably were, dur-ing the process of analysis.

The above-described procedures and principlesare not novel, and most of them are used with varyingemphasis by all workers on archaeological ceramics.

8


The general classificational hierarchy used wassimilar to that used on most American archaeologicalceramics.

1. Subdivision into slipped and unslipped pottery.2. These two groups each divided into wares on the

basis of paste characteristics, surface treatment, andsurface color, all based on megascopic examination.The wares defined here are broadly drawn in compari-son to those of most workers. This was felt to be ad-visable in a study with such a broad regional and tem-poral coverage as is this.

3. Each ware subdivided into vessel forms, jarsbeing those of markedly restricted orifice (slip, if pre-sent, not being found on interior); basins, bowls of var-ious types, plates, cups, all having a relatively largeorifice (slip, if present, always found on the interiorand sometimes on exterior). Other rarer forms wereseparated when recognized.

4. Sherds belonging to each vessel form then classi-fied according to variation in rim form, design, if pre-sent, distribution of slip, base form, etc. The waredescriptions and plate captions give the details of thisclas sification.

The author does not hold with the extremist viewthat all classifications must of necessity be arbitrary,but is firmly convinced that there is a natural hierarchyof combinations of attributes in single specimens whichthe pottery classification should mirror,12 and that theabove-described kind of pottery classification does soin this, as in all the other varied collections he hasworked upon. He does not wish this statement to com-mit him to a closely genetic theory of the origin of histypes, nor to a denial that many factors, such as deco-rative style, may cross borderlines of a hierarchicclassification. Random assortment of traits in artifactswill not allow the definition of types. Were there notclear tendencies toward combination on individual potsof such traits as paste character, texture of slip, andsurface color, the definition of a ware would be mean-ingless, as would the defining of a vessel shape, werenot certain shapes of rim, body, and base found togeth-er in a preponderant number of vessels in the productsof a given culture. These patterns or combinations of

traits on single specimens and groups of specimensallow us to describe a cultural repertory or, converse-ly, to establish the cultural provenience of a specimen.Such placement would also be impossible were thesepatterns not complicated enough to guarantee that theywere unique to a given time and place. It is the com-plicated nature of these trait combinations which makespottery such a sure and sensitive tool for culture place-ment. The description of wares and types as markersin the complex variation encountered in ceramics isan organizational step without which the first technicalstage of cultural analysis would be impossible.

It should be emphasized that this stage, the chrono-logic and regional placement of material, is but a pre-liminary to the cultural study which should follow. Oncethe culture material has been classified and dated, itis available for study; at this point the cultural mechan-ics must be worked out afresh, using either individualspecimens or copious records for the finer definitionneeded.13 It is but fair to say here that I have not beenable, for want of time as well as for want of accessto the pottery during later stages of the work, to carrycultural reconstruction into the finest possible detail.The voluminous tabulations, and samples of the pottery,are still available and merit future work. A furtherapology as to method of presentation should be made.Since information was gradually accumulated from astarting point in 1940 in a new field, and back-checkingof early recording was often impossible, the level ofworkmanship in this report is more uneven than couldbe wished. Technological analysis of material, whichcould have solved several of our most perplexing prob-lems, was abandoned after a preliminary trial, due tothe press of time and the wide variety of ceramicshandled. The detailed analysis of the many ceramicassemblages described here has not been attemptedfor want of time, although such analysis by statisticaltechniques has been started on several of them (seeRobinson, 1951; Brainerd, 1951). In the interests ofeconomy, the detailed tabulations are not presented.They will be kept on deposit, available to the smallnumber of students who may wish to use them.

9

II. THE SITESCOBA

Coba is unique as the most northerly well-preservedsite of a clear-cut Central Maya type. The Peten simi-larities in architecture, site planning, sculpture, andinscriptions are strikingly close, and the large sizeand excellent preservation of the site have allowed de-tailed study (Thompson, Pollock, and Charlot, 1932).The nineteen stelae found at Coba' give a dating rangein the Maya calendar from 9.9.0.0.0 to 9.12.10.5.12,with some stylistic evidence of earlier dates amongthose not directly decipherable. There is also architec-tural evidence of a Late Mexican substage occupationsimilar to the main occupation at Tulum (ibid., pp.129-130). Surprisingly, because of its peripheral posi-tion, the architecture and inscriptions have been judgedequal to if not a shade more advanced than those of thePeten of the same period (ibid., p. 196).

Henry B. Roberts excavated pottery samples fromGroup B at Coba' during April, 1932 (see map 3). Sixtrenches were dug, yielding 10,099 sherds, of whichthe rims and vessel legs were brought in. As far as Iknow, I tabulated this complete collection. It totaled2,238 sherds.

Pottery of three time horizons appears in the colleo-tions; two of these are represented only by traces. Themain period is a close equivalent of the Late Regionalassemblage from Yaxuna (Yaxuna III). The overwhelm-ing majority of the Coba' collections seem to be pure ornearly pure deposits clearly of this period, although theassemblages are not identical with those from Yaxuna.

Twenty-nine fragments were classifiable as slate-ware, although 19 of these were atypical. There issome suggestion among them of color intergrading in-to mottled Medium Redware like that from Yaxuna III.This suggests that they may well antedate our better-known slatewares. The remaining fragments fit wellinto the Yaxuna and Maya Chiche'n Itza' (Late Classicstage) slateware collections (compare fig. 49, , 1-6,with figs. 15, 67); one fragment (fig. 49, m. i) in bothrim form and color is suggestive of Holactun slateware.All of these, except for two dubious sherds, come fromstrata which, judging from Roberts' notes, probablyconsist of postconstructional wall slump. Thus, al-though a small but rather consistent number of slate-ware sherds comes from upper levels of these trenches,there is no positive evidence that slateware comesfrom under the plaza floors which were pierced bynearly every trench. Two of the typical Yaxuna-Chich6ntype slateware fragments show lichen staining, an evi-dence of their having come from the surface. The mosttenable explanation is that slateware-using people wereon Coba Group B in small numbers before the plazaretaining walls collapsed, but after the construction ofthe group. The position of slateware in the stratifieddeposits at Yaxuna reinforces this sequence.

Figurine incensarios (for examples of the generaltype, see figs. 98-101), occurred on or near the soilsurface in several trenches, and constitute a majorpart of a sample taken from inside the standing vaultof structure 9. This structure is classified by Pollockas an East Coast style shrine of a type known fromTulum (Thompson, Pollock, and Charlot, 1932, p. 41),and thus shortly antedates the Conquest. Coba is knownto be held sacred by the present-day Maya, and theruins have very likely been the scene of many pilgrim-

[10]

ages to leave small offerings in the past. A group ofincensarios, probably of relatively recent date, wascollected by Thompson. One of these (fig. 100, O) isshown. Figure 100, f is a modern Lacandone incen-sario. Stylistically, the Coba' vessel is closer to thisthan to the Figurine Incensarios, and therefore maybe suspected to be post-Conquest. The absence ofLate Mexcican Coarse Redware in the Coba collectionssuggests that Group B at least was not inhabited dur-ing this late period and that the shrine of this periodwas perhaps only used by pilgrims as a place of wor-ship.

The ceramic collections from Group B (the largestgroup at Coba') thus point to intensive occupation dur-ing a relatively short period which the stela datesplace as 9.9.0.0.0 to 9.12.10.5.12 (613-682 A.D.). Thisdating is strongly supported by Robert E. Smith's re-port on the Coba' pottery, made at Chichen Itza in1940. Smith dates 119 of the 240 polychrome sherdstabulated (see figs. 2, b 31, a) as definitely Tepeu 1of Uaxactun type, and sees no sign of another periodin the remainder of the polychrome, much of whichwas too finely fragmented or badly worn for positiveidentification (see charts 2-4 for these period datings).Smith notes that twenty-six rounded bowls, mainlymonochrome but some possibly worn polychrome (seefig. 2, a, ., 7, 21), appear more like Tepeu 3 but maywell be Tepeu 1. Twenty-six conoid-footed tripodplates with low basal molding (Yaxuna type mottledMedium Redware, fig. 2, f and .) fit Tepeu 1 in form.Eight red-slipped jar rims, he states, look more likeTzakol than Tepeu. Thus 145 sherds, most of themelaborate polychrome and thus closely diagnostic, fitTepeu 1, which conforms to the Coba stela dates, whileonly 34 sherds of more generalized type may equatewith other Peten horizons.14 This is a strikingly closeagreement, considering the distance between Coba andthe Peten, and is equaled only at Yaxuna. The ceram-ics therefore support the other Peten similarities andsuggest that the common people as well as the rulershad southern affiliations.

There is stratigraphic evidence of ceramic changein the collections but it is of a very gradual nature.The unmixed collections of the major period show in-crease with time in the percentage of coarse pasteware. Regional Coarse Redware in lower levels oftrenches 1, 2a, 2b, and 5 runs 32.8 per cent of a totalof 256 sherds; in the high levels of these trenches itruns 53.1 per cent of a total of 1,335 sherds. Theclose similarities with the Peten noted by Smith areall among the fine wares, but such wares are so nu-merous as to make local manufacture likely. Thecoarse pottery, however, also shows certain similari-ties to southern forms. Similar redware dishes (seefig. 2, a, b) were found at San Jose' and Benque Viejo(Thompson, 1939, pp. 100-101, fig. 47; Thompson,1940, figs. 11 and 13), where the corresponding typesare dated in periods III a and b, which in turn aresynchronous with Tepeu 1 at Uaxactun. Another sug-gestive parallel with the central Maya sites is theuse of thumbed, sub-rim encircling fillets (see fig.2, e) on Tepeu basins at Uaxactun (Smith, 1936a,Tepeu 5, 6, 8). Plates from Panama (Lothrop, 1942,part 2, figs. 8, 9) are roughly comparable to thosefrom Cobg (fig. 2, b) and to unslipped forms from Ya-xuna (fig. 1, b, 9, 10), though the distance makes this


correspondence uncertain and cross-dating hazardous.None of the above similarities are identities. Al-

though the coarse wares were certainly locally madeand therefore likely to be less cosmopolitan in typethan the more elaborate wares, it seems probable thateven the latter were locally made, although they mayhave been made by more traveled craftsmen.

Qualified by the above observations and analyses,we can confirm the Coba Group B collections as repre-sentative of a short, heavy occupation ranging in timefrom the Maya dates 9.9.0.0.0, or somewhat earlier,to 9.12.0.0.0. A further sample taken from this ruin,preferably including material from the other architec-tural groups, slhould considerably clarify the ceramicrepertory of the above period, in addition to provingwith greater certainty the stratigraphic position ofslateware on the site. Late Mexican substage potteryshould also be present at least in the vicinity, judgingfrom ar chite ctur al indications.

The markedly close relationship of the ceramics ofthis site to those of the central Maya area is shared tosome degree by Yaxuna III, and to a lesser extent bythe Late Regional ceramics of Acanceh. Coba' Poly-chrome, however (fig. 3, a), is a ware not shared bythe other two sites.

YAXUNA

Pottery collections from Yaxuna were studied in1940; these collections were excavated by H. B. Robertsin 1932. Roberts' ten trenches have been plotted fromhis notes on map 4, revised from the survey made byO'Neill and Stromsvik (O'Neill, 1933). Roberts notesthat "the most important information obtained fromthese two sites [Sayil and Kabah] was the discovery inthe lowest strata of two trenches at Kabah of polychromesherds of the same general type as those from HolmulV and Uaxactun III [Tepeu]. Polychrome sherds ofwares assignable to the above two periods have beenfound under stratigraphic conditions at two other sitesin the Yucatan Peninsula; namely Coba and Yaxuna.The Coba' and Yaxuna sherds were found in strata whichdid not contain any of the standard Yucata'n-Maya [slate]wares" (Roberts, 1935, p. 127).

Roberts' Yaxuna pottery was examined by Robert E.Smith at Chiche'n Itza' in 1940; he reported sherds ofChicanel, Tzakol, and Tepeu types in the collections.The collection in storage at Chiche'n Itza' at this timeconsisted of 272 sherds, separated by trench but notby level. An additional sample collection from theseexcavations was found at the Peabody Museum; it con-

sists of 89 sherds. The fact that the original collectionshad evidently been combined made the checking of strat-igraphic locations impossible.

The relatively large proportion of Peten-like mono-

chromes and some polychromes, and the presence ofPuuc-style architecture and slateware sherds, sug-gested that excavations at Yaxuna might allow an equa-tion of the Yucatan and Peten chronologies (see Thomp-son, Pollock, and Charlot, 1932, pp. 197-205). Thecauseway connecting Yaxuna and Coba', together withthe proximity of Yaxuna to Chichen Itza', suggested thelikelihood of ceramic connections among these threesites, and the stelae at Coba# promised a chronologicalanchor of sorts (see ibid., pp. 205-206).

About two weeks were spent in excavation at Yaxunain February, 1942. Fourteen trenches were excavated(numbered 21-34, see maps 4 and 5). Several trencheswere located in an attempt to augment and clarifyRoberts' ceramic sample from the environs of thenorth building group; another series from the cenote

to the northwest of the ruins gave a less complex de-posit, representing in nearly pure form the periodleast well known from the other site collections. Struc-ture 8 was sampled and the trenches found to containlittle but Formative pottery. Of these trenches, 29 and30 gave large, pure Formative collections. This mound,consisting of a high platform bearing evidence of sev-eral probable structures, measures approximately 60m. by 130 m. at the base, altitude over 20 m., and vol-ume about 50,000 cu. m. It would certainly repay fur-ther investigation, as would the whole site.

The Yaxuna chronologic sequence.-There is ceram-

ic evidence of four distinct periods at Yaxuna, eachcharacterized by a change in the predominant slippedware. Although large unmixed collections were not ob-tained for each of these periods, there is reason to be-lieve that such deposits exist on the site. The earliestof the periods, Yaxuna I, belongs in the Yucata'n LateFormative substage and compares quite closely withthe Chicanel phase at Uaxactun, showing a few resem-

blances to Mamom ceramics. Following this period isone characterized by Flaky Redware associated withthe less frequent but more diagnostic Incised Dichromeand with Black Trickle on Redware: Yaxuna II, Yuca-ta'n Early Regional stage. Next in order, a phase mark-edly similar to that of Coba' Group B is well repre-sented: Yaxuna III, Yucata'n Late Regional stage. Fi-nally, the ceramics show the slateware period, alsorepresented by the only standing architecture: YaxunaIV, placed in the Yucatan Florescent stage.

The order of this sequence is amply supported bystratigraphic changes in the proportions of wares, al-though the last two periods are not represented by un-

mixed collections. The slipped wares of four excavatedtrenches have been graphed to demonstrate this se-

quence (charts 2-4). Trenches 24 and 25 show transi-tions from predominantly Late Formative deposits(Yaxuna I) into the Dichrome-Flaky Redware period(Yaxuna II), nearly missing the Coba' Group B period(Yaxuna III), and ending with slateware predominant(Yaxuna IV). Trenches 21 and 26, which adjoin andhave therefore been combined (chart 4), begin theirhistory with a stratum (cut C) showing predominantYaxuna I pottery. Cut B shows the Yaxuna III warespredominant. The deposit from bottom to top shows a

gradual increase in Medium Slateware. Although Ya-xuna II is but slightly represented, this deposit thusshows Yaxuna III wares following Yaxuna I and II waresin strong percentage. This sequence is weakest betweenthe Yaxuna III and IV, but support for the III-IV se-

quence has also been found at Coba' and at Acanceh.These placements gain further support from theirequivalents in the Peten sequence, which are discussedlater in this section.

The cenote collection, as stated above, showed a

relatively pure, short-period occupation called Yaxu-na II. Of the 3,905 sherds tabulated from the cenotetrenches, 180 (4.6 per cent) were of Thin or MediumSlateware, mostly atypical from those of the Puuc sitesin form and finish. Percentage of this slateware de-creased, although indecisively because of the small-ness of the sample, from top to bottom of trenchesthrough the four cuts taken. Although some intergrad-ing between the cenote monochrome wares and mono-chromes of the later ruin deposit were noted, no ce-note monochrome fragments were distinctive enoughto be attributed definitely to Yaxuna III.

Chronoloaic equivalence with Peten phases. -RobertSmith examined the greater part of the Yaxuna collec-tion made in 1942, and selected 236 sherds out of thecollection of 8,909 as identical with or very similar toUaxactun sherds.

I 1


Ruin

Similar to Identical with

Uaxactun Pottery

33

30

1848374

The overwhelmingly Yaxuna II deposit from

note, it will be noticed, contains Uaxactun Chicanel

and Tzakol correspondences in nearly equal numbers,

suggesting at first glance that Yaxuna U should

placed chronologically between the Chicanel

kol phases at that site. It should be borne in mind, how-

ever, that the Chicanel-like specimens are all

orated monochrome and were probably locally made,

while the Tzakol-like specimens are polychromes,

rarity of which in all Yaxuna collections suggests

they were traded in from a distance. Since it has

been suspected (see, for example, Thompson,

224 and table 17) that there is a partial or complete

disjunction in the occupation of Uaxactun between

Chicanel and Tzakol phases, and since there is

tic evidence that the decorated type of this period,

cised Dichrome, is stylistically linked and probably

antecedent to Tzakol-style polychrome (see fig.

tion), Yaxuna II can most likely be placed in this

The distribution of the Peten-like sherds in the

of trenches 24, 21, and 26 (included in the "ruin"

umn of the above table) have been shown for direct

comparison to the distribution of the native wares

(chart 4). The right-hand columns show the percentages

of sherds attributed by Smith to each of the Uaxactun

phases. Total slipped sherds in each cut are shown

the figures to the right of the graphs. The trench

sherds were not seen by Smith; the Peten-like

in this trench are at any rate too few for significant

percentage comparison.

Late Formative monochrome is predominant

bottoms of these trenches, and Smith has classified

most of it (62.5 per cent) as of Uaxactun type,

phase. The next local period in order is that ce-

note sample. Flaky Redware shows as major percent-

ages in the middle levels of trenches 24 and 25,

nearly absent in trenches 21 and 26. The graph

equivalents for trench 24 (chart 3) shows a

nance of Tzakol pottery corresponding with

closeness in distribution with Flaky Redware,

ing the smallness of the Peten-style sample. This

currence supports the evidence of the cenote trenches.

The third period, characterized by Medium and

Redwares closely similar to those from the Cobaes

cavations, is predominant in levels a and b of

21 and 26, but nearly or entirely absent in the

two trenches. The Peten graph for trenches 21 and

shows a predominance of Tepeu 1 in these two upper

levels.Thus the first three sequent periods at Yaxuna

with reasonable certainty be equated approximately

with the Uaxactun periods Chicanel, a period

Chicanel and Tzakol which Robert Smith recommends

calling Matzanel, and Tepeu 1. Above them lies

slateware. Only one or two sherds of Peten-like

tery dating later than Tepeu 1 (believed to have

about 9.12.5.0.0) were identified at Yaxuna. The

plest hypothesis to explain these facts is to date

slateware (equating with the building of the Puuc-style

ruins) as having begun with Tepeu 2 times,

sume a dropping off of cultural relationships

the Peten and the Yucatan areas at this time to

the absence of later Peten-style wares. An overlapbetween the Peten occupation and the Northern Yuca-tan sites has long been suspected (Thompson, 1939,pp. 231-232, 240); the sequence from Yaxuna suggeststhat the overlap is more marked, and dates earlier,than formerly suspected.

An alternative explanation may be presented withsome degree of plausibility. The absence of Peten in-fluence in the slateware period at Yaxuna may be dueto a disjunction in the stratigraphic column. The sitemay have been deserted and remained uninhabited un-til after 10.2.0.0.0, the estimated end of the potterychronology in the Peten, and then reoccupied by slate-ware-making peoples. This reconstruction is decidedlyless likely than that of a continuous occupation. First,there is no evidence in the several trench profiles ofabandonment of the site, and all the deposits sampledwhich show this time interval evidence the admixtureof pottery to be expected in a site under continuous oc-cupation. Stylistic evidences for continuous occupationare also convincing; the slateware, which is the pre-dominant and diagnostic ware of the Puuc period, oc-curs in considerable proportion even in low levels ofthe trenches. This suggests early beginnings of thetechnical characteristics used to define slateware.This theory is strikingly supported by the paint tech-nique and shapes of the Trickle on Flaky Red potteryof early Regional times, which show close similarityto later slatewares (see fig. 7). The majority of theYaxuna slateware sherds are from jars and basins,and in most cases show intergradation into earlierwares in paste and slip as well as in form. The sameis true of such bowl sherds as occurred. Thus a con-tinuous stylistic tradition is indicated throughout theYaxuna deposits. Also, there are certain similaritiesin form between ceramics of the Florescent stage andUaxactun ceramics of Tepeu 2 and 3 times, the demon-stration of the exact nature of which must await pub-lication of Smith's report.

It seems likely from the Yaxuna pottery collectionsthat the Florescent stage there coincides in its begin-nings with Tepeu 2 at Uaxactun. This placement hassince been supported by my excavation at Chenes sitesduring 1949. This placement is an importantlink inthe synchronization of Northern and Central Mayachronologies. The chronological position of the FlakyRedware assemblage, and its developmental implica-tions in the history of Maya pottery, is further dis-cussed elsewhere.

Notes on architecture and sculpture.-A carvedstela was found (fig. A) in the south end of trench 28,facing 3 degrees east of north; its dimensions were:height 1.18 width .72 in., thickness .39 m. Itseems to have been reset in the position in which itwas found, judging from the broken edges at the baseand at the lower, eastern corner. It was set flush withthe face of a step having a .42 m. riser and a .40 m.tread, which may have been the bottom member of astaircase ascending the north face of structure 4. Thetop of the stela had been broken off at ground level.The front of the stela, as well as the step, bore sev-eral coats of plaster, each painted red. The plasterwas applied heavily to the front of the stela to forma smooth surface, obliterating the carving. The sidesand back of the stela were carefully shaped, but boreno plaster. The red paint appeared to have been ap-plied in at least two coats, the first dull and orangetinged, the final coat a strong purplish-red in colorand showing specular reflections (imported specularhematite?). Fragments of carved plaster bearing thesame paint coatings were found in the trench, probablyfallen from a building atop the substructure to the south.

CenoteSimilar to or Identicalwith Uaxactun Pottery

No sherds identifiedas later than Tepeu 1Tepeu 1 5Tzakol 35Chicanel 26Mamon

12


On the south side of the North Group, just west oftrench 1, an uncarved stela lay on the fallen terraceslope, and near the southeast corner of the smallestof the five ruined buildings to the southwest of it, ajamblike sculptured stone (fig. B) was still in placein the wall face. At the foot of the south face of sub-structure 7 were noted badly fractured fragments ofwhat appeared to have been one or more stelae. Ap-proximately 150 meters northwest of the northwestcorner of the North Group, there was a ruin containinground column drums. A single column drum was foundat the west side of the small square buildings south-west of trench 1. On the west slope of substructure 4were strewn small cylindrical stones about 35 cm. indiameter, each with one side flattened as though theyhad been engaged to a wall surface (colonnettes). Sev-eral elements of a carved stone mosaic wall decora-tion were found just west of the standing building, struc-ture 5.

The synchronism of these architectural features withthe pottery periods can be assigned with some degreeof probability. The subfloor deposits under the carvedstela contained slateware pottery (Yaxuna IV) mixedwith earlier wares. The stela seems to have been reset;it is thus likely that it may date earlier than these de-po sits.

The standing building, 5, at the North Group has thedistinctive veneer masonry characteristic of the Puucruins, and therefore must date in period IV. The groupof small buildings below it to the west, containing thesculptured jamb and column drum, must also date tothe same period on the basis of masonry style. Thegroup to the northwest of the North Group is Puuc instyle, identified by the cylindrical column drums whichcharacteristically occur centered in door openings inPuuc buildings. The small engaged columns are of atype which, although found in the Puuc, are perhapsmore characteristic of the Chenes area. All of thesearchitectural features are usually found in slatewarehorizons and therefore probably date Yaxuna IV.

It should again be emphasized that the most strikingarchitectural feature found at Yaxuna i3 structure 8, alarge pyramid which yielded ceramic samples of LateFormative date. This mound is of roughly the samesize as one of Miraflores phase recently excavated atKaminaljuyu (Shook and Kidder, 1952). This structure,which would seem to have been approximately contem-poraneous to structure 8 at Yaxuna, was built almostentirely of adobe and contained richly furnished tombs.The Yaxuna mound is composed mainly of rough lime-stone; the absence of any Late Formative pottery amongsome 1,000 complete specimens examined from Yuca-ta'n collections suggests that rich tombs of the LateFormative substage are not characteristic of Yucata'narchaeology, and that mound 8 was thus built solely asa building substructure. It should be remarked that re-cent opinion is inclined to date the Pyramids of the Sunand Moon at Teotihuacan to late in the Formative stage(Kidder, Jennings, and Shook, 1946, p. 252). The evi-dence in Yucataon is for a continuity of culture betweenthe Formative and Regional stages, with no change inthe location of ceremonial sites (see Brainerd, 1951).

OXKINTOK

About a week in March, 1940, was spent in excava-tion at Oxkintok in company with Pollock and Shook,who were at that time working on the architecture andstelae of that site (Shook, 1940).

Four ceramic trenches were dug (see map 6), twoof them (1 and 3) in a lower court upon which backs the

building with the Initial Series lintel which dates9.2.0.0.0, the earliest Initial Series date yet found inYucatan. Both trenches penetrated shallow deposits,but both yielded sherd collections from under the mor-tar floor of the court. The above-floor pottery fromthese trenches is judged contemporaneous with depos-its from the Puuc sites, and is discussed in that con-nection. The locations of the trenches were chosen inhopes of anchoring the ceramic sequence accuratelyto the Initial Series date. Trench 1 showed only Puucoccupation. The trench 3 subfloor pottery contains con-siderable early material. Trench 4, dug off the westend of the terrace bearing the Initial Series buildings,shows only a superficial layer of Puuc period pottery.The lower refuse shows increasing mixture of earlypottery, with only one slateware sherd below the floor.

The following tabulation shows the percentages ofslateware (the characteristic slipped pottery of thePuuc sites) in relation to the total sample of rim andsignificant body sherds. Slateware constitutes about75 per cent of the sample in typical Puuc collections.Variations below this percentage evidence early pot-tery as a constituent of the collection.

Trench 1 aFloorTrench 1 bTrench 2 a-cTrench 2 d-gTrench 3 aFloorI'rench 3 bTrench 3 cTrench 4 aTrench 4 b (east end)Trench 4 b (west end)Trench 4 c (east end)FloorTrench 4 c and d (west end)

Slatewar epercentages

78.4

76.614.20.0

75.7

44.54.5

73.950.449.066.0

Totalrims102

60409760185

24719739423611491

0.8 802,875

Trench 2, dug about 200 meters to the west of theInitial Series lintel group off the slope of a major plat-form which bears a large pyramid, yielded 2.50 metersof rich deposit, with some 1,200 rim and significantbody sherds. The platform retaining wall proved to bein good condition, preserved by the sterile earth of apostoccupational slump. Remains of a thin mortarfloor adjoined it but extended only a few centimetersfrom the platform base. Just under this floor lay aborder of wall stones at right angles to the superim-posed platform, probably the facing of a low terrace.These structures extended only through the southeast-ern 4.5 meters of the trench. They were not disturbed,in accord with our policy of obtaining samples with aminimum disturbance of architectural structure onthe ruins. In this case, the lower terrace border stonesproved to have been laid on the native sascab (softlimestone) which sloped sharply downward to the maxi-mum depth of the trench. This deep deposit of culturaldebris, by far the deepest we located outside of a ce-note, is markedly at variance with the normal soildepth in Yucatan, which ranges from zero to six ornine inches. This area seems to have been a basin,bounded to the northwest by a limestone ridge. It mayhave been formed by quarrying for limestone and sas-cab for building purposes.

Although the deposit was not sealed in by a floor,there is every evidence that at least its lower levelsaccumulated during a relatively short period and re-

13


mained undisturbed by later activity. Levels b and ccontained some slateware mixed with previously un-encountered wares, while levels d, e, f, g, runningfrom 1.40 m. to 2.50 m. in depth from surface, containno slateware but a homogeneous assemblage of potterywhich includes a relatively small number of fragmentssuggesting the Uaxactun Tzakol phase. Robert E. Smith'sreport on these lots of pottery follows. It has beenslightly paraphrased, and references to my illustrationsof the types he discusses have been added in brackets.

In general the monochrome wares are Petenishbut fail to be exactly like Uaxactun, except for avery few sherds too fragmentary to be absolutelycertain; one of these is a small, pedestal-base black-ware vase, unslipped on exterior except just belowthe lip, interior slipped, eggshell thin, and Tzakol3 [fig. 12, i j,L15. Several rim sherds may belongto brown-blackware tripod cylindrical bowls [fig.12, j, ] with hollow slab feet, one present [fig. 13,.gJ. Some are vertically fluted, and some plain orwith one or two horizontal incised lines below lip onexterior. Orangeware bowls have rounded sides,horizontal grooves below lip, vertical fluting, andmodeled human faces. One blackware bowl withrounded sides has incised geometric design done be-fore slipping [fig. 12, jL 18]. All of these are Uaxac-tun types, providing the bases are Uaxactun in style,and they are not present.

The near-Peten types are: (1) Basal flange bowlswith flaring sides, beveled lip and ring base. Manyof these have red and black on orange decoration oninterior [fig. 9, gJ. Some of these may be Tepeu 1basal molding types. (2) There is also a smallergroup of basal flange bowls; one flange is triangularin section with rounded lip. (3) Bowls with roundedsides, probably with ring bases; these run more toblack than to red, whereas the basal flanges arenearly all red in monochrome ware [fig. 13, i 1 m;fig. 12, d-hJ.

As can be seen from the above, the only close syn-chronism with the Peten seems to be with Tzakol 3, thelast subphase of Tzakol, and there is a suggestion ofTepeu 1, though Smith emphasizes the lack of complete-ly reliable correspondences.

If late Tzakol is accepted as the correct placing ofthe collection, it should date between 9.2.0.0.0 and9.8.0.0.0 according to present estimates of Uaxactundating. The Oxkintok lintel dates 9.2.0.0.0.

The material examined by Smith probably covers

only a short time-span sample. The collections lumpedtogether here from trench 2, levels c to g, showed no

stratigraphy. The relatively small number of waresand shapes further argue the homogeneity of the sam-

ple. Illustrations shown of Oxkintok Regional potteryare all from this collection.

The placing of this sample in the local ceramic se-

quence is a complex problem. None of the major waresof this deposit have been located in any other collectionfrom Yucatan in quantity, and under conditions allowingstratigraphic placement. Hence any placing must de-pend on stylistic criteria, similarities in shapes anddecorative features with other collections, rather thanon positive ware identifications.

The collection contains no marked Chicanel-likefeatures in form, decoration, or slip type; this is incontrast to the early collections from most other Yu-catan sites. It also contains no Incised Dichrome nor

marked similarities to its accompanying Flaky Red-ware, which in some samples shows Chicanel similari-ties and closely follows Chicanel in time. It must there-

fore postdate both these periods. On the other hand,the early Oxkintok collection shows considerable sim-ilarity in ware and shape with the main Dzibilchaltunceramic assemblage, which in turn can be documentedas preceding the Florescent stage deposits from thePuuc sites. The Acanceh excavations show a strati-graphic sequence of vessel forms from an assemblageresembling Yaxuna III into Dzibilchaltun-like pottery.The Oxkintok pottery fits best stylistically at the be-ginning of this sequence. No stylistic transition fromearly Oxkintok strata into the later Puuc period isdocumented from these trenches, in spite of the grad-ual increase of slateware at trench tops shown in thetable, and thus these trenches show mixed rather thantransitional deposits. The placement of the Oxkintokceramic assemblage earlier than Dzibilchaltun is alsosupported by: (1) the absence at Oxkintok of fine gray-ware found at Dzibilchaltun, which shows close stylis-tic similarities with Piedras Negras wares whichlikely date about 9.18.0.0.0 (see fig. 28, g); (2) by ab-sence at Oxkintok and presence at Dzibilchaltun offine orangeware, distinct though similar to that fromthe Puuc sites (fig. 59, s); and (3) by the absence ofPeten tradewares at Dzibilchaltun, considered in rela,-tion to our findings elsewhere that Peten trade ceasedafter Tepeu 1 times. Since Dzibilchaltun predates thePuuc sites and shows close stylistic similarities withthem, Oxkintok must predate Dzibilchaltun.

The evidence for placing the Coba collections atabout 9.10.0.0.0, the date of the stelae there, is clearand straightforward; the Peten tradeware in this col-lection is preponderantly Tepeu 1. The Coba' and earlyOxkintok collections have in common redware basinswith subrim flange (fig. 2, c, 11-9, 23 26, e; fig. 11,d-O which are markedly similar, though the Cobaflanges are usually thumbed while the early Oxkintokspecimens are not. Regional Coarse Redware jarsfrom Yaxuna (fig. 8, a-O are suggestive in form ofthe early Oxkintok Cinnamon Buffware jars (fig. 11,a-d), though here the form is more generalized andthe resemblance thus less certainly significant. Closersimilarities with Acanceh jars have been mentionedabove. Bowl forms at Oxkintok are not as closely simi-lar to those of Yaxuna and Coba' as are the forms ofthe jars. The Oxkintok bowls are also markedly closerto Puuc slateware bowl forms than are the easternspecimens. These facts would alternatively suggestthat the early Oxkintok collection is later than theCoba' collection, or that the distinctive shape range ofthe Puuc bowls developed in the western part of thepeninsula and spread east at a later date. The latterhypothesis is made nearly certain by chronologicalevidence given above and by the predominantly westerndistribution of the Puuc architectural style, as well as

by the evidence of sculptured dates and Peten trade-ware. Thus the Oxkintok ceramic assemblage datesperhaps later than the lintel date of 9.2.0.0.0 and cer-

tainly earlier than the Coba' collections; somewhereearlier than 9.8.0.0.0.

The bottom cuts of trenches 3 and 4, dug in thecourt adjoining the initial series building (see map 6),contained collections predominantly of early wares

but with a little slateware intermixture. These cutsunderlie the floor of the courtyard. The architectureof the adjoining building which bears the Initial Serieslintel is of the so-called early type (Pollock, 1940;Shook, 1940) which disappeared before the buildingof the classic Puuc sites. There is evidence at Dzibil-chaltun and elsewhere in northern Yucata'n that thefirst appearance of slatewares antedates the full de-velopment there of Puuc style veneer masonry, al-though there is also evidence at Dzibilchaltun that

14


slatewares either superseded the earlier slipped waresgradually, or overlapped them in time but were re-stricted regionally.

The occurrence of slateware sherds under the court-yard floor at Oxkintok indicates that this floor consid-erably postdates the Initial Series lintel. The buildingcontaining the lintel may also antedate the courtyardfloor, though this is not certain at present on the basisof the evidence given above.15 The possibility that thisfloor is later than the building is strengthened by thefact that it lies to the rear, abutting the building plat-form on which the Initial Series building rests. Thus itmay be seen that although the presence of late potteryin the bottoms of trenches near the Initial Series build-ing at Oxkintok does not agree with the early date ofthe lintel, it also does not present valid evidenceagainst the dating of either the building containing thelintel or of Early Oxkintok pottery at about 9.2.0.0.0.With present evidence it seems justifiable to assumethis dating, or one not far from it, for the early potteryas semblage.

Mercer (1896) recovered a collection of ceramicsfrom the labyrinth, or Sat un Sat, at Oxkintok (struc-ture 3, b, I on map 6). I have examined this collection,which consists almost exclusively of figurine incensariofragments dating from shortly before the Conquest,another evidence of Maya religious pilgrimage to an-cient sites (Brainerd, 1953). In the course of architec-tural exploration, Pollock discovered a single sherd ofplumbate pottery in Puuc-style fallen building debrisat Oxkintok. This fragment, of Early Mexican substage,need indicate no more than a casual visitor during thisperiod.

The architecture, like the ceramics, indicates twophases of occupation at the site. Whether they werecontinuous is not certain from the pottery but it islikely, since Pollock and Shook note evidence of arealexpansion in the architectural remains. The dating ofthe earlier phase has been discussed; the ceramics andarchitecture place the later as contemporaneous withthe Puuc sites. The documented time range thus runsfrom 9.2.0.0.0 probably to at least 10.2.0.0.0, or pos-sibly as late as 10.8.0.0.0 (about 500-1000 A.D., by the11.16.0.0.0 correlation). The ceramics, as noted else-where, are Puuc period closest to those of Uxmal,which is the nearest large Puuc site to Oxkintok. Shooknotes the possibility that a Maya causeway may haveconnected the two sites.

The sculptures found at Oxkintok are varied and im-portant. Shook (1940) notes their association with theearlier, central plazas of the site. Proskouriakoff,who has recently tackled the intricate problem of clas-sifying the strikingly varied Yucatan Maya sculpture,assumes (1950, p. 160) that the Oxkintok sculpturesdate 'much later than 9.11.0.0.0." If Proskouriakoffis correct, a continued occupation is indicated for Ox-kintok, with the stelae far postdating the dated lintelas well as the early ceramic assemblage, but perhapssomewhat antedating the Puuc-style architecture andceramics. On architectural-ceramic grounds, the mosteconomical hypothesis at present favors an earlierdate than does Proskouriakoff for these monuments,but the data are far from complete.

DZIBILCHALTUN

A pressing problem in the formulation of a Yucatanceramic chronology was the connecting of the sequenceinto historic times. Prof. Alfredo Barrera Vasquez,who gave generously of his time and knowledge in thisas in other matters, showed me a site situated about

1 kilometer south of the hacienda of Dzibilchaltun,which lies about 5 kilometers east of the 15 kilome-ter marker on the road from Me'rida to Progreso.This site, known locally as Xlaca (see map 7), con-sists of several ruin groups connected by two cause-ways. Near the center of the site is a deep, clear ce-note or aguada. surrounded on three sides by a flatbedrock ledge only slightly higher than the water lev-el. This is an ideal water source and has probably beenso for a long time.

East of the aguada stands a ruined church of thecapilla abierta type with arched stone apse and sacris-ty standing, as well as a detached single-room build-ing nearby to the north, possibly used as the curate'sresidence (Rl, R2, R3, on map 7). The nave of thechurch must have been of perishable material, sinceexcavations for its corners yielded no results. TheseSpanish buildings were constructed largely of stonefrom the ruins, including several fragments withcarved decorations. The stones are large and squared;there are few of the tapered edged "veneer stones"used during Florescent and Early Mexican times. Thecarved stones seen at the site were concentrated aboutthe Maya plaza in which the church was found, savefor one in the bottom of the aguada (see Andrews,1942, for architectural discussion). On the altar wallof the church, fragments of an elaborate fresco werestill in place. Drawings were made and will be pub-lished elsewhere. There is some stylistic evidencein these frescoes for a 16th century dating. A test ex-cavation at the altar end of the church showed severalsuperimposed altar platforms and the presence of nu-merous burials under the floor, those seen being poor-ly preserved. Excavation of the environs of the detachedbuilding showed a courtyard with low surrounding wall,and a loose stone found near the north door bearing abadly eroded but probable date of 1593. The figuresare in rounded relief against a flat background. Thelast figure in the date is more uncertain than the firstthree.

Several of our workmen came from the village ofChablekal about 1 kilometer to the east of the ruin.Local legends that the "virgin" in the Chablekal churchcame from the aguada, and a general tendency amongthe inhabitants to accept the ruined church as a fore-runner of their own, prompted a visit to the Chablekalchurch. To the right of the church doorway an unusual-ly large stone showed faint traces of an inscription.This stone, as well as the surrounding wall, had beenwhitewashed many times. Upon removal of a quarterinch of paint, an inscription became visible in the Mayalanguage in Roman script, dating 1612 or 1617. Thisclear inscription (fig. C) as well as another, a re-worked corner stone set in the wall on the west sideof the corral at the hacienda of Dzibilchaltun and bear-ing an incised date of 1662 or 1663, bolster the readingof the early inscription date at the doorway of the de-tached building.

In an effort to confirm this dating from historicalsources, Mr. France Scholes was consulted and verykindly contributed the following references and com-ments:

Chablekal was probably the Pech town of Chablefor which we have sixteenth and seventeenth centuryreferences. The story of Chable, insofar as I canpatch it together from the data at present in myfiles, is as follows:

The pueblo of Conkal was granted in encomiendato Don Francisco de Montejo, the son. This townhad various suietos. i.e., other villages whichformed part of the jurisdiction of the cabecera of

15


Conkal. Among these suietos were Chulul, Sitpach,and probably Chicxulub. Montejo also held othertowns in encomienda, such as Dzilam, Itzimna, andDzibical. These towns were inherited by his son,

and during the last third of the sixteenth centuryvarious changes took place. In 1605 there was liti-gation concerning the encomienda holdings (A. G. I.,Mexico 242) from which we learn that at some timeprior to this date the town of Chable, undoubtedlythe Chablekal of later times, was separated fromChulul, one of the suietos of Conkal. The Spanishreads: "Cable que es cierta parcialidad de indiosque se aparto de Chulul." What does this mean? Itmay mean that Chable was founded as a separatesettlement made up of Indians from Chulul; or itmay mean that there were small settlements atChable and nearby, such as Xlaca, regarded as su-,ietos of Chulul, which were brought together andconcentrated at Chable, resulting in the setting up

of a separate pueblo. Personally, I lean to this inter-pretation. In any case, where we first find referenceonly to Chulul, we now find separate pueblos men-

tioned in 1605. When did the separation take place?On this point all that is possible is to set a periodduring which Chable was set up as a separate pueb-lo. Obviously it took place some time prior to 1605,when the litigation was in progress. The 1582 list,published in DocumentosPara la historia de Yuca-

t

p. 56, mentions Chulul but not Chable as one ofthe pueblos served by the convento at Conkal. Con-sequently, it may be inferred that the establishmentof Chable as a separate village took place after thattime. Thus you can fix the period 1582-1605 as thetime when the change occurred. Throughout theseventeenth century the encomienda of Conkal in-cluded both Chulul and Chable. Moreover, it appears

that both Chulul and Chable remained attached toConkal for religious purposes because as late as

1688 these two towns occur in a kind of census listprepared by the priest of Conkal.

Now, what about Xlaca? Here all I can do is tospeculate. Possibly this settlement was one of theoutlying suietos of Chulul and if your 1593 date iscorrect, it may indicate that the friars began to setup some kind of mission establishment there aboutthat time. It is unlikely, however, that the friarswould have maintained two chur ches a mile and a

half apart at both Xlaca and Chable. Lack of refer-ences to Xlaca in a subsequent period may mean

therefore that this village was administered as partof the Chable mission. Indeed, it may be that thepeople there were congregated or concentrated atChable as part of the movement which resulted inChable's becoming a separate pueblo. If this specu-lation has any validity, and taking into account thedates, it is possible that the setting up of the Chablemission occurred between 1593 and 1605 when thelitigation shows Chable as a separate place. I wouldnot try to be very definite about these dates and Idon't imagine that you have to be. The developmentin the Chable area, which included Xlaca, apparentlytook place after 1582 due to the fact that there is no

mention of either Xlaca or Chable in the list of thatyear or prior to 1605. If your mention of the churchat Xlaca as a capilla abierta means that it was nev-

er roofed over in a permanent manner, this may beevidence that it was not permanent and that with thesetting up of Chable as a separate pueblo the friarsdecided to move their church to the latter pueblo.This again would explain the later date on the Chablechurch. In short, it seems to me you have a periodof some twenty-five to thirty-five years when these

developments were taking place in the Xlaca-Chablearea, which ought to be good enough for your pur-

poses.Incidentally, the name Chablekal appears in a

census of about 1794 recently published by RubioMane in the first volume of his Archivo de la His-toria de Yucata'n, Campeche y Tabasco. This listdoes not include Xlaca or Dzibilchaltun. This ref-erence appears in the census data for the coastalarea north and northeast or Merida. In this area

there were 256 estancias, among which may havebeen Xlaca and Dzibilchaltun.

Thus it appears clear that Chablekal (earlier Chable)became a part of the Yucatan religious organizationnot much earlier than the date on the church frontthere. The Xlaca church was larger than the churchat Chablekal, although it was never completed in per-

manent materials. It seems likely that it was abandonedafter a few years' use, possibly because of its incon-venient distance from the settlement at Chablekal, andthe church at Chabekal built to replace it. (For plansand description of the Chablekal church, see VegaBo-lanos et al., 1945, vol. 1, pp. 419-420.)

The only pottery of the early Colonial period foundat Dzibilchaltun consisted of a few redware jar bodyfragments, a small fragment of a cream paste, whiteglazed plate, and fragments of a deep redware caldronrestored in fig. 220, a. This pottery all came from theimmediate environs of the dwelling. The lack of frag-ments of Coarse Redware pottery, which characterizesthe period from well before the Conquest to the pres-

ent day, argues that the Maya site surrounding thechurch did not support a village during the period ofthe church, or at most was settled for a very shorttime. No surface fragments of this ware were foundon the site, although a considerable number of earlierfragments were collected during the exploration andmapping of the area.

The site was surveyed using an oil suspension pris-matic compass. Most distances were measured bytape, others were paced. The map is believed to befree from any major inaccuracies of orientation andplacement. The western sacbe measures 1.32 kilome-ters between platforms. It consists of a slightly raisedmasonry pavement edged by a roughly faced line ofstone. Bordering the causeway, at an average distanceof 3 meters on each side, are rows of massive, roughlimestone slabs set on edge. The eastern sacbe isshorter, being only 402 meters. This causeway is lo-cally known as the Izamal Road, and the villagers as-

sured me that it continues to Izamal. The direction ofthe section of the causeway which was mapped is some

50 north of the bearing of Izamal. Group 7 at its east-ern end lies on the outskirts of Chablekal. The area

of the village was unsuccessfully checked for a con-

tinuation of the "Izamal Road," though the possibilityof such an extension farther east remains and was

not checked. The connecting stretch may have beendestroyed in construction of the village.

We discovered twenty-two stelae on the site, whichis the northernmost on which a similar concentrationof monuments has been observed. Only two carvedstelae are known from farther north, from Dzilam on

the coast northeast of Dzibilchaltun. Of these twen-ty-two stelae, six were carved, all badly eroded, andthe other sixteen were plain. Of the six carved stelae,two, in fragmentary condition, were built into theSpanish buildings. Andrews, who was shown the carvedstelae, states that none bear Initial Series inscriptionsand he was unable to read any dates from them (An-drews, 1942, p. 259). Two standing vaulted Maya build-

16


ings were discovered (in groups 2 and 5, map 7). In aneffort to utilize the developing ceramic sequences forthe chronological placement of Yucatan architecture,Andrews visited the site and drew plans of group 2from which, during his stay, small subfloor ceramicsamples were taken. He also discovered a nearly bur-ied range of vaulted rooms in group 4 (see ibid., p. 259for summary description).

The pottery from Dzibilchaltun, aside from the Co-lonial wares mentioned above, may be divided intothree discrete periods, all separated in our chronologyby intervening time intervals. The earliest period isthe Late Formative substage. Only 33 sherds of thishorizon were found, and most of these occurred assingle sherds in large collections. In trench 2, how-ever, they occur concentrated in the lowest level, C,and account for 16 out of a collection of 36 sherds. Thistrench, strangely enough, lies beside the Spanish church;thus the settlement seems to have ended for the time inthe location where it began. In trench 2 there is a traceof the Yaxuna Cenote period (early Regional) overlyingthe prestela wares, 4 sherds which seem to be CobaGroup B period, and above this lies the main period ofthe site. These trench 2 collections are so small thattheir stratigraphy, proved correct by other excavations,could certainly not be proved from this trench alone.

The major period of the site is one undoubtedly closeto or overlapping with that of the Puuc ruins. MediumSlateware is a major constituent of the collections andits form and decorative repertory is generally similarto that of the Puuc sites, save for various of the moreelaborately decorated forms which are missing here.Of the rest of the Puuc ceramic repertory, Thin Slate-ware is prominent, Thin Redware nearly lacking; un-slipped wares including incensarios are present, thoughnot very similar to the corresponding Puuc types. Afine orangeware is present, similar but by no meansidentical to Z Fine Orange; Fine Grayware and ThinBlackware, rare in the Puuc, are much more abundanthere. In addition a ware virtually absent in Puuc collec-tions is very prominent at Dzibilchaltun. This is Redon Thin Gray. Several of the forms of this ware areidentical to those of the accompanying Slateware, al-though there is no intergrading of the two in ware,paste, or slip.

The proportions of the major wares encountered on

the site are given in the following tabulation of the to-tal collection of sherds excavated.

Total number of sherds from Dzibilchaltun: 10,010.Unslipped 5,082 50.4 per cent

Fine Slate 115 1.1Medium Black 41 .4

Fine Orange 43 .4

Medium Slate 2,097 20.87Fine Gray 271 2.7Red on Thin Gray 1,619 16.0Coarse Slate 1,160 11.3

It will be noted that Red on Thin Gray nearly equalsSlateware in frequency, and that unslipped ware occurs

with higher frequency than in most collections.The presence of Medium Slateware in quantity in

these samples places the period as being near the Flo-rescent stage or Early Mexican substage, in both of

which Medium Slateware is the major slipped pottery.The style of nearly all the slatewares, as well as that

of the accompanying wares, further restricts the col-

lections as closer to Florescent. Does the period then

precede or follow the Florescent stage? Early Mexicanceramics seem immediately to follow Florescent at

Chich6n Itza, thus leaving no room for this period and

thereby leaving it earlier than Florescent. Much morepositive evidence is in the form as well as other char-acteristics of the Slate and Red on Thin Gray wares,which place them stylistically between Early Oxkintokand the Puuc period, and also between the Coba GroupB and the Slateware period at Yaxuna. The fine wares,Fine Gray, Fine Orange, and Thin Blackware, alsoplace this material stylistically as an early or pro-to-Puuc occupation. This occupation seems to havecovered a relatively short time span. The large collec-tions from trench 14 show (graph 6) Fine Graywareand Thin Blackware being replaced by DzibilchaltunFine Orangeware, which is a close relative of Puuc orZ Fine Orangeware.

Robert Smith, in checking this collection in 1942,found no Peten-like wares at all. When considered inconnection with the fact that at Yaxuna and elsewherein Yucata'n, as will be seen, Peten pottery resemblancesdecline markedly after Tepeu 1, this lack of Peten ce-ramic resemblances suggests a date of after 9.13.0.0.0for this deposit, although there is a possibility, lesslikely to my mind, that the northern position of the sitewould account for lack of Peten contact. Similaritieswith Usumacintla ceramics are noted, although thereis but little published comparative material from thatarea. The most marked of these similarities is be-tween the Fine Grayware beakers and certain coarsergray-brown ware beakers bearing incised monkeys,found in some of the latest deposits (9.16.0.0.0 or later)at Piedras Negras (specimens at University Museum,Philadelphia). A vessel found at Guaytan in the MotaguaValley also shows resemblances (cf. fig. 28, e.-; 36,b-i, i with Smith and Kidder, 1943, fig. 27, a).

Evidences of a third later period occur near thesurface of some of the Dzibilchaltun deposits as col-lections strong in Coarse Slateware. This ware is thediagnostic of the Middle Mexican substage. One frag-ment of Medium Slateware (fig. 15, d, §j is definitelyof the Early Mexican substage. The Coarse Slate (fig.20, a, j.) is very close to its counterpart at ChichenItza' in all details noted. It is found only in the upperlevels of trenches 3, 7, 11, 14, and 16, all of which, itwill be noted, lie along the north range of group 3.Sculptured stones of this period are mentioned by An-drews (1942, p. 259). I am not certain which fragmentshe attributes to this period, but two fragments of largelimestone slabs were noted bearing relief decorationssomewhat similar to those on Mexican Chich,6n build-ings. One was visible on the bottom of the aguada,which the natives claim is part of a subterranean tem-ple; the other (plate IIIb) was found in front of the de-tached building north of the west side of Group 5.

Although Dzibilchaltun failed to produce any quan-tity of the Spanish contact ceramics hoped for, itsmain phase sheds much-needed light on the develop-ment of the Puuc ceramic complex. The Coarse Slate-ware gave small but nearly pure samples of a period(Middle Mexican substage) found at Chiche'n Itza' onlyin mixed collections. The few early sherds discovereddemonstrated that here, as in nearly every site sam-

pled, a Formative stage settlement existed, the evi-dence for which would indubitably be much strongerif the later heavy occupation had not nearly destroyedit. In this case, the aguada may have been the reason

for the long occupation of the site, but here as else-where, at Yaxuna for example, there is evidence ofeven closer localization; the Maya used the same build-ing platforms over long, though often discontinuous,periods. And here, as in so many other sites, theSpaniards chose ruin areas for their churches, utiliz-ing the stone blocks in their buildings and to burn forlime.

17


The site is worthy of further study. The stelae, al-though badly eroded, merit description, which Andrewswill perhaps supply from his notes. Proskouriakoffhas commented that their composition is similar tothat of central area Maya monuments (Proskouriakoff,1950, p. 164), a similarity paralleling such features asthe closely spaced plan of the site, the Peten-like ma-sonry, and architectural details. The stucco faiadediscovered and reported by Andrews (1942) may be ofmajor importance. Its excavation should not be under-taken without governmental arrangements for its care.

The exact dating of the major Dzibilchaltun occupa-tion cannot be fixed certainly without more excavationsin its region. In general terms its position seems clear.It must date from a period slightly prior to the fluoritof most of the sites of the Puuc area, but almost cer-tainly after the beginnings of some of them. During theoccupation of Dzibilchaltun, it seems likely that slate-wares were being manufactured in some parts of Yuca-tan, and Regional monochromes in other parts; thissituation can only be checked by careful analysis ofceramics from a series of sites spaced over the areaconcerned. The time period of this occupation canroughly be estimated as within the range 9.13.0.0.0 to9.18.0.0.0.

ACANCEH

A series of tests was made at Acanceh. The site waschosen in the hope of dating the famous stucco fa,adeas well as a pyramid on the north side of the town pla-za (see map 8, b) which bears large stucco faces sug-gestive of those on the Chicanel-phase temple E-VIIsub at Uaxactun (Marquina, 1951, pl. 242). Andrewsspent several days during the ceramic excavations inthe determination of architectural sequence on theacropolis (Andrews, 1942, pp. 257-259). During thistime we excavated trenches 17, 18, and 19 (see map 8,a) in areas calculated to aid him in dating the architeo-tural sequences.

The ruin groups in the town have been robbed exten-sively for local construction; irregular, rounded areasshown in the town plan consist of masonry fill so badlydestroyed as to afford little indication of form. Archi-tectural stone elements of Florescent-stage form inplace on the acropolis surface attest an occupationlater than that evidenced by the standing buildings (ibid.,p. 258).

Nineteen trenches were dug, spread over the recog-nizable ruin areas, yielding a total of 9,441 sherds. Thetrenches north of the plaza near the early style pyramidyielded altogether only 36 sherds; all trenches encoun-tered rock fill with no recognizable floors. This areais at present nearly formless in contour, making choiceof trench location difficult. Trench 20 (depth 1.50-2.50meters) yielded a deposit containing one Incised Di-chrome sherd with 14 other Flaky Redware sherds al-so assignable to the Yaxuna Cenote horizon. Higherlevels show later wares. Relationships of this collec-tion to the stucco-faced pyramid are not clear; thereis only a suggestion that it may be Early Regional indate.

The trenches around the edges of the acropolisshowed considerable depth of deposit, and several pro-duced large stratified collections. These trenches weremapped by plane table, using permanent benchmarks,and levels of all cuts, floors, and other structural ele-ments encountered are on record; these were takenwith a tripod telescopic level and all measurementsare within 5 cm. accuracy. This survey is on file, andif consulted should assure freedom from errors due

to disturbance by our trenches if this group is laterexcavated. It should also allow a correlation of anyarchitectural features discovered with our ceramicscale.

With the pottery of Yaxuna, Oxkintok, and Dzibil-chaltun as a guide, wares were relatively easy of de-finition at Acanceh. This was fortunate, as nearlyevery trench showed a stratigraphic sequence runningthrough several periods. As in most of the materialheretofore described, the deposits show considerablemixing between successive periods, which are recog-nizable only by predominance of characteristic waresin successive cuts.

Quantities of Flaky Redware appear in the earliestcuts. This ware is very similar in appearance to Ya-xuna Cenote redware, but forms are of Late Formativeas well as of the later Yaxuna Cenote (Early Regional)period (fig. 17, e-_h fig. 18, a, 1-7). A few sherds ofFlaky Dichrome occur (fig. 17, g, 1-0), but none withincising. Formative Monochrome cream ware bowls(fig. 17, L 3-34), preslip striated redware jar frag-ments, and negative painted monochrome (fig. 17, e,l, , -20_;Q, 3 3,), all of Late Formative dating,occur in smaller quantities. The Late FormativeDzibilchaltun sherds have a flaky slip and also showclose form resemblances to the Acanceh specimens(fig. 5, fand 17, J).

The variation among the Acanceh and Dzibilchaltunwares shows the ceramic situation in the Late Forma-tive substage in Yucatan to be complex. Although someof this variation in slip character and minor waresmay in part be regional, chronological change is alsolikely to be involved. More material is required be-fore these wares can be plotted through space andtime.

Above the Flaky Redware in sequence come depos-its containing Coarse and Medium Redwares, manywith mottled slip similar to those of the Coba'-YaxunaIII wares. These collections may be classed as ofmid-Regional stage. Shapes and decoration in this pe-riod resemble similar wares from both Early Oxkin-tok and Coba Group B and probably periods inte:me-diate between them in time, as well as earlier andpossibly slightly later wares. The pottery of the Ox-kintok-Coba time span (138 years between the Oxkin-tok lintel and the earliest Coba' date, and 69 additionalyears for the Coba span of dated stelae, plus or minuswhatever discrepancies exist between the date of oursamples and of the inscriptions) has been difficult todifferentiate into periods in the Acanceh deposits,which contain rather small, heterogeneous collectionsof the wares of this time range. Thus our proof thatthe Coba Group B period dates later than Early Oxkinr-tok must be primarily based on the Peten pottery cor-relations, supported by the stela dates of the two sites.Stylistically, Oxkintok Early wares, slipped bowls forexample, are closer in form to slateware than arethe Coba Group B-Yaxuna wares; the latter, althoughthey must date later and thus closer to the Puuc hori-zon, were made far to the east of the Puuc area inwhich nearly all of our Florescent sites are concen-trated. There thus seems to be evidence for develop-ment of Yucata'n Florescent-style pottery in the west,while the eastern peoples clung closely to the Petenceramic traditions. This same regional dichotomy isdocumented in the concentration of Initial Series ste-lae and Peten style architecture at Coba, contrastedwith a near absence of carved monuments and the dis-tinctive Puuc style architecture in western Yucatan.

A ware of the Regional stage at Acanceh which wasnot demonstrated to occur in either the Coba or Yaxu-na collections is Red on Thin Gray, which has the same

18


or closely similar ware characteristics to that of Dzi-bilchaltun. This ware occurs in stratified sequenceabove Coarse and Medium Redwares, and intergradessomewhat with them in paste, slip, and color. It alsoshows form similarities in jars and basins (cf. figs.18, c, 2 4-11 with fig. 2, c, ; fig. 8, e-g). The jarshapes of both wares are suggestive of Early OxkintokMonochromes. Many pieces of Coarse Redware varytoward cinnamon-buff in slip color; a few fragmentsare definitely classifiable as of Early Oxkintok wares.Acanceh Red on Thin Gray is indistinguishable to theeye from Dzibilchaltun Red on Thin Gray, and occurscharacteristically with slatewares. Thus there is strat-igraphic evidence at Acanceh for placing the Dzibilchal-tun collections as following the Coba' Group B-Yaxunacollections and the Early Oxkintok collections in time,and preceding the Florescent collection.

Supporting our establishment at Dzibilchaltun of atime period characterized by the use of Red on ThinGray in nearly equal quantities with Puuc-style slate-wares, the same association is apparent in severalcollections from Acanceh. However, the later Acancehcollections show a greater proportion of slateware andof Puuc Thin Slate and Redwares and a smaller propor-tion of Fine Gray and Thin Blackwares than at Dzibil-chaltun, coupled with certain Puuc-like shapes in theRed on Thin Grayware. This places them later in alocal sequence which probably culminated in the domi-nantly slateware collections which characterize thePuuc sites. The stylistic connections between EarlyRed on Thin Gray and the cinnamon-buff and red waresof early Oxkintok, coupled with the evidence of a grad-ual supplanting of Red on Thin Grayware by slatewares,documents a general trend for northwest Yucatan.

There are no sizeable Acanceh collections contain-ing typical Puuc pottery assemblages free of Red on

Thin Gray. Trench 4-A (see map 8a) is the nearestapproach to one. A few Medium Slateware sherds ofEarly Mexican substage shape and slip range (fig. 21,a, , i, for example) attest a very minor occupation ofthis period.

The lack of typical Puuc ceramic assemblages inthe sites we explored in northwest Yucatan is possiblynot significant; the area was very lightly sampled, andfew of the sites tested showed the short occupationalsequences most likely to yield pure samples if an oc-

cupation did exist. An easily sustainable argument can

be presented that Puuc-style buildings once existed innumbers in northwest Yucatan, but have been robbedof masonry so extensively for churches and otherbuildings later built upon and near them in Mexican andpost-Conquest times as to leave little trace of eitherPuuc architecture or ceramics. The ceramics in theMe'rida collections are predominantly Puuc in period,and their proveniences, in the rare cases in which theyare known, are widely distributed over the country.Surface ruins are certainly not a reliable indication ofculture distributions in Yucatan, as witness the beliefheld until recently that Yucatan supported only a primi-tive agricultural population until after the dissolutionof the Old Empire and the moving north of its survi-vors. In this case, the temples of the earlier peopleshave proved in many instances to lie under the archi-tecture of the later. Perhaps the Spaniards did a more

thorough job of demolishing earlier structures thandid the Maya before them.

Conversely, present evidence suggests a markedconcentration of religious centers in the Puuc area

during the Florescent stage. Furthermore, the ceram-

ics from the Puuc sites suggest that these were reli-gious centers on new locations, a break in the nearlyuniversal Maya custom of utilizing time-honored sites.

The possibility exists that the Medium Slatewarecollections from Dzibilchaltun and Acanceh, whichshow closer similarities to early collections than dothose of the Puuc, must be examined to determine iftheir variation from the Florescent Puuc collectionsmay not be due to regional rather than temporal dif-ferences. It must first be said that the Puuc collectionshave unfortunately yielded stratigraphic evidence ofonly slight stylistic change, despite careful analysisdirected toward that end. Thus, unless samples are

large, we do not have the criteria from the Puuc thatshould clearly attach the other sites either to the be-ginning or the end of the Puuc chronology. Puuc pot.tery shows more stylistic elaboration and technicalvirtuosity than the slateware-containing collectionsfrom the sites to the north and east of it, but such dif-ferences may be explained, at least to some extent, bythe indubitably peripheral position of any other sitescontemporaneous with the Puuc concentration. TheRed on Thin Graywares of Dzibilchaltun and Acanceh,which form a considerable part of each sample con-

taining Medium Slateware, give definite chronologicalclues because of their stylistic similarities to EarlyOxkintok and Coba' Group B wares, the dates of whichare fixed in the Peten chronology and the positions ofwhich are fixed in the Yucata'n sequence.

Unfortunately for the solution of this problem, we

do not at present have sufficient ceramic samples toassure us that the total ceramic sequence through theRegional and Florescent phases is represented forboth the Puuc and northern areas. If we had, it shouldbe easy to determine a series of synchronisms whichwould allow the sequences from the two areas to becompared for regional variation. Was there a timewhen a ceramic assemblage containing slatewares as

the only slipped ceramics existed in north Yucatan?And conversely, was there a period of occupation inthe Puuc during which Regional monochromes were

the only slipped wares (as at Oxkintok and Acanceh),or were used simultaneously with slatewares (as atDzibilchaltun and Acanceh)? And what are the chrono-logic relations of these assemblages?

At Acanceh we can document with some certaintythe chronological change from Regional monochromeassemblages to an even mixture of monochrome andslatewares. In the Puuc, despite more thorough sam-

pling, we find only slateware assemblages, and thissituation has been further supported by similar resultsin the Chenes and Rio Bec areas in 1949.

The disparity between the ceramic sequences ex-

cavated in the Puuc and northern Yucatan sites can

best be explained as a function of a combination ofregional and chronological variation. In the analysis,the problem is the evaluation of the relative effectsof these two factors; and of these, only the chronologi-cal factor can be isolated since its effects, by defini-tion, should be observable in collections from all thesites sampled. Regional variation can be recognizedonly after the chronological variation has been allowedfor, and should appear as the variations between anytwo occupations identifiable as contemporaneous.

In the determination of chronological markers,which must precede any determination of regionaldifferences, there are various measures which maybe applied. The slatewares from the northern sitesshow various similarities to the slatewares fromMani, the Puuc, and the Chenes sites, as well as tothose from 'Maya" or Florescent Chichen Itza', andall of these show certain traits rare or absent in Puucsite collections and lack certain others common inPuuc site collections. Also, there are certain indica-tions that these differentiating criteria place the other

19

20 ANTHROPOLOGICAL RECORDS

collections at the early end of the Puuc ceramic se-quence. This evidence is given a degree of confirma-tion by the 1949 Chenes collections, as yet incomplete-ly analyzed.A further chronological yardstick may be found in

the finely made, obviously imported ceramics in thecollections of the northern and Puuc sites. At Dzibil-chaltun, Thin Blackware, Fine Grayware, and Dzibil-chaltun Fine Orangeware are all found in some quan-tity (see figs. 53, i, 28, g, and 59,g, , respectively,for these wares). Thin Black and Fine Graywares oc-cur only as traces in the large Puuc collections. ZFine Orange, which occurs in the Puuc (fig. 59,b--,,is stylistically distinct from the Dzibilchaltun variety,and quite likely represents a later development of thesame region which was earlier the source of Dzibil-chaltun Fine Orange. It should be added that certainstylistic similarities between the ceramics of the Puucand northern sites and those of the central Maya areabear out the evidence given above for the placing ofthese northern sites as somewhat earlier than the Puucfluorit, but probably contemporaneous with the earlierpart of the occupation of some or all of the Puuc sites.

If it be hypothesized that there were marked region-al differences in the kinds of utilitarian ceramics man-ufactured in Yucata'n during certain periods, it shouldpay to attempt a reconstruction of the conditions ofceramic manufacture and trade which might producethe mixtures found in the excavated collections. To be-gin such a reconstruction, some cultural interpretationsof ceramic typology are prerequisite. The ware is ourmost important ceramic category in the determinationof locale of manufacture. A ware is defined as a massof pottery which in general has similar characteristicsof paste, slip, and finish. These characteristics dependupon the sorts of materials selected from those geo-graphically available, and upon the techniques of manu-facture. Even when as broadly drawn as they are inthis report, they mirror categories of pottery whichmust have been apparent to their original users, andthey must also reflect cultural traditions of manufac-ture, limited by local accessibility of suitable rawmaterials.

Pottery manufacture in Yucatan today, and indeedgenerally in Mesoamerica, wherever wares of indige-nous tradition are made, is carried on only in certaintowns. Although the industry is in large part a familyaffair, the families who make pottery tend to clusterby town or even in one quarter of a town. The samesituation exists in the less elaborately organized mod-

ern pueblos of the Southwest. Coupled with craft spe-cialization by community in Mesoamerica is the customof trade between communities, at present carried onin town markets which are held in a succession oftowns according to a weekly cycle. There seems to beevery reason to assume that this pattern is an aborigi-nal survival. If so, the collections from Dzibilchaltunand Acanceh containing Red on Thin Grayware andMedium Slateware can be explained by the assumptionof a southern (Puuc) origin for slatewares and generaltrade over northern Yucatan. Various other ceramicassemblages, such as those in Late Formative andPuuc deposits, are also easily susceptible to such ex-planation; it seems quite possible that such a custommay stretch back in time to the Formative stage-and,from the Old World distribution of similar trading pat-terns, it may even be suggested as a general phenome-non in folk culture.

The distribution of fine and elaborately decoratedwares is harder to explain by historic archaeology;the religious hierarchy for which such wares wereprobably produced is gone and the historical continuity

thus interrupted. The Kamninaljuyu Esperanza tombs(Kidder, Jennings, and Shook, 1946), as well as numer-ous other less spectacular finds, attest the great ex-tent and volume of such luxury trade in Classic-stageMesoamerica. There is some suggestion in the largequantity of fine orange pottery from Veracruz foundatChichenItza' that luxury trade was further extendedin post-Classic times, perhaps coincident with an in-crease in coastal sea traffic.

If the analysis proposed above is acceptable, ourarchaeological data show a marked regional variationin the ceramic industry of northern Yucatan at a hori-zon which can be placed at sometime near 9. 16.0.0.0(about 750 A.D.). At this time the peoples of northernYucata'n were using monochrome pottery in thetraii-tion of the earlier Regional stage pottery of that area,but they were also using slatewares, which seem tohave been the local ware of the Puuc area during theRegional stage and probably were developed within thePuuc-Chenes-Rio Bec area. We know that at sometime before theMexican stage, slatewares were thesole native slipped pottery of the area of Chich6nItza',and we suspect from reconnaissance that this samesituation existed over all of northern Yucata n. Thusthe evidence points to the gradual replacement of lo-cal monochromes in northern Yucatan by slatewareswhich are known to be the native wares of the area tothe south.

Further and broader reconstructions which dependon these findings are discussed in the conclusions.

A more fully represented Middle Mexican substageoccupation is next in stratigraphic sequence (cf. trench3, a-d, chart 10), topped by a slight deposit of Late

Mexican Coarse Redware. There are cuts showingabout 80 per cent Coarse Slateware period wares,with 20 per cent of wares characterizing other pe-riods. The Coarse Redware occurs only as minorpercentages in collections consisting mainly of waresof earlier periods.

Advantage was taken of Andrews' discovery thatthe building facing the stucco fagade was a later con-struction, on a northward extension of the pyramid.Portions of the fill within this extension and betweenthe two buildings were removed (see map 8, jb,trenches17, 18, 19) and a small but workable collection re-sulted. From Andrews' stratigraphic architecturalevidence (Andrews, 1942, p. 258), materials in thisfill must all predate the northern building and post-date the stucco faiade building. Both these buildingspredate a Florescent occupation, posited on the pres-ence of Puuc architectural members scattered on thesurface of the acropolis, and fragmentary architectur-al remains in place above the level of the stucco fagadebuilding. The pottery deposits of this horizon musthave largely disappeared as the masonry of the top ofthe mound was robbed.

The ceramic collection from the trenches whichpostdate the stucco fagade building (fig. 19, d-.g) con-sists of a mixture, the earliest identifiable sherdsprobably being of Formative date; the latest, whichconstitute the majority of the samples, show clearsimilarities to Early Oxkintok specimens. A basalflange polychrome fragment of Tzakol style is in-cluded (fig. 19, g, 3). The Formative pottery must becontained by chance in earth used much later.

If the Early Oxkintok collections are contempora-neous with the lintel there, the stucco fasade templedates not later than about 9.2.0.0.0. Although these col-lections may date as late as 9.8.0.0.0, it must also beremembered that the deposit under discussion postdatesthe stucco faiade temple by an unknown time span. Itthus seems safe to date the fasade as early ninth cycle.

20


An excavation of the acropolis, with its evidencesof long occupation, should yield much superimposedarchitectural evidence. The trenches showed a use ofoccupation earth fills rather than the loose stone fillscharacteristic of the Florescent and Mexican periods.The pottery contained in these fills should allow adating of the architecture, thus making this site ahighly informative one in an area where architecturalremains of this age are scarcely known.

MAYAPAN16

Mayapan is a most important link in the joining ofthe Chichen Itza ceramic chronology to the Christiancalendar, since Mayapan is known through early Span-ish documentation to have been abandoned only shortlybefore the Conquest. Vaillant found some ceramic sub-stantiation of this late dating in a small sample fromthe site (Vaillant, 1927, pp. 353-354, 366).

Approximately two weeks were spent at Mayapan inthe collection of 32,413 sherds from 13 trenches. To-ward the close of these excavations, Andrews beganarchitectural studies at the site (Andrews, 1942).

During this trip two trenches, numbers 11 and 12,were sunk in the cenote of the nearby town of Telcha-quillo (map 9). The large collection (2,644 sherds)obtained all seemed post-Conquest, but lack of Euro-pean tradeware makes its dating difficult. It seems nodifferent, however, from the better-dated Mani Colo-nial collection, save that unslipped wares were com-pletely lacking at Telchaquillo; all the fragments be-longed to water jars, which seem always to have beenslipped during this period. This difference is likely tobe due to the circumstances of deposition; the Manideposits were kitchen refuse, the Telchaquillo sherdscame from a water source.

Fortunately, a survey of Mayapan had been madeby R. T. Patton in 1938 (Carnegie Institution Yearbookno. 37, pp. 5-6), resulting in the map of the centralgroup which we have here used (map 10). The trencheswere concentrated in this group, with only threetrenches, 4, 8, and 10, being dug in outlying areas.Trench 8 yielded a small stratified collection, all ofwhich, except for one sherd, is Middle or Late Mexi-can in date. For a description of the masonry of thegroup in which trench 8 was dug, see Lawrence Roys,1941. 19,016 sherds of the collection, including nearlyall sherds of the earlier horizons, came from trench-ing in Cenote Chun Mul (see map 11), which shows ce-ramic evidence of almost continuous use from Forma-tive to Conquest times. This cenote collection, then,may be taken as a fair indicator of the span of humanoccupation of its immediate area, which is the Mayapanmain ceremonial group.

The slipped wares of the cenote collections may besummarized as follows:

Coarse Redware (Late Mexican 84.3Period)

Coarse Slateware (Middle Mexican 7.2Period)

Medium Slateware (Classic and Early 2.2Mexican)

Tzakol- style polychrome (may include .3some late Mexican polychrome)

Flaky Redware (made only during Form- .4ative and Early Regional Periods)

Formative monochromes .394.7

per cent

per cent

Other miscellaneous pottery made up the remainder.

Despite the wide chronological range defined strat-igraphically here (chart 17), this is one of the mosthomogeneous collections recovered from excavationsin Yucata'n. Late Mexican Coarse Redware is heavilypreponderant. Next, but a very poor second, is theCoarse Slate characteristic of the preceding MiddleMexican period. This points to a short intensive occu-pation of a hitherto lightly inhabited location. Thisoccupation follows with little or no overlap, and per-haps even with a disjunctive interval, the last Chich6nItza' building period.

Trenches 1 and 2 in Cenote Chun Mul (map 10)show the earliest occupational evidence, consistingof 67 sherds of Formative monochrome in mixed col-lections from their lowest levels; in trench 1 thisdepth is 1.50 meters, in trench 2, much less. In gen-erally ascending order in these trenches occur asmall number of sherds of most of the major chrono-logical periods known from the Peninsula, thus check-ing the stratigraphic order arrived at elsewhere, andadding evidence that no major ceramic groups havebeen missed in the survey. The proportions of Tza-kol-style polychrome sherds to other sorts of contem-poraneous pottery are far larger than in deposits fromother sites. In fact, very few undecorated sherds attri-butable to the same horizon were found in these poly-chrome-containing collections. In addition, a poly-chrome bowl (fig. 9, f, 8) with badly eroded surfacehas been found in this cenote, containing human bones.Polychrome bowls may have been preferred to theless spectacular wares for secondary burials, and itis thus conceivable that this concentration of poly-chrome sherds may have resulted from the use of thecenote as a burial place. R. E. Smith did not see thesepolychrome fragments (fig. 9, f); they all seem tocome from basal flange bowls and are doubtless clas-sifiable as Tzakol style. The situation in the cenoteis therefore what one finds in most water sources oflong standing in Yucatan, save that the late Regionalstage seems to be nearly or completely absent, andthat the intensive late occupation in this vicinity hasmarkedly enlarged the deposits attributable to thattime span.

All trenches at the site yielded small percentagesof Medium Slateware, usually most strongly repre-sented in their lower levels. Only 3 sherds out of the219 of this ware found (including fig. 15, p, 6, 2) showform characteristics of the "Mexican" part of thechronological span of the ware. These three charac-teristic sherds must represent a larger frequencythan their number indicates, since more indeterminatesherds probably come from this period. However,they are still scarce enough to indicate that the Me-dium Slateware occupation must have been restrictedto a minor occupation belonging to the Florescentstage, after which the site was abandoned, or verynearly abandoned, until Middle Mexican times.

The Coarse Slateware at Mayapan occurs in everytrench and in nearly every cut, tending to show moreprominently at the bottom of the trenches. Its propor-tion in regard to the later redware is, however, smallin every collection, although larger than that of Me-dium Slateware. No collection shows a predominanceof Coarse Slateware over Coarse Redware; its largestfrequencies occur in trench 7 which adjoins but doesnot pierce the pyramid slope of the serpent columntemple, which was cleared and drawn by Andrewssubsequent to the excavation of the ceramic trench.A floor covered deeply by ashes was found at a depthof 1.20 m. from the lowest surface point at the trench,under level e (chart 14). The trench continued to bearpottery including Coarse Redware until it reached

21


bedrock .70 m. below the floor. Two building periodsare thus suggested for this temple, during the earlierof which Coarse Redware had already come into use.There are no data to tell us whether the first buildersof Mayapan had begun to make Coarse Redware in ad-dition to Coarse Slateware at the time of their arrival,or whether more excavation on the site would show apure Coarse Slateware horizon such as was found atDzibilchaltun.

The samples are enlightening with regard to thechronologic placing of the attached figurine incensariosfound over so much of Yucatan and adjacent areas.These have long been known as late in the archaeologi-cal record because they characteristically occur onor near the surface, often overlying collapsed Mayastructures. There is thus considerable evidence oftheir placement on ruins by postoccupational visitors.This custom is still followed by the Lacondone Indians,who make pilgrimages and place incensarios in ruins.

The Mayapan collections show clear evidence fromseveral trenches (see charts 17, 16, 15, trenches 2, 5,6) of increasing proportions of figurine incensariostoward the surface of deposits, beginning in the lowerstrata with predominantly Coarse Redware collectionscontaining few fragments of figurine incensarios; intrenches 1, 6, and 8 there are no fragments of them.The manufacture of figurine incensarios therefore be-gan later than that of Coarse Redware. Mayapan isrevered as an ancestral capital by the Maya even now,and this type of incensario may well date until afterthe Conquest, since there are many sixteenth centuryrecords of the difficulty experienced by the Spanishfriars in stopping the manufacture of idols.

The determination of the dates of occupation ofMayapan is of particular importance in the placing ofall earlier periods in Yucatan. As we have stated above,the Mayapan ceramic sequence gives evidence that thesite was settled at the time the period of architecturalactivity ended at Chiche6n Itza'. The earliest phase ofarchitectural activity recognized thus far at Chiche6nItza' is contemporaneous with the Puuc ruins, and thereis reason to believe that the Puuc fluorit began about9.18.0.0.0. This sequence of occupation is the mostfully documented one which covers the time that inter-vened between the end of the Initial Series period andthe Spanish Conquest in the Maya area.

A major handicap inherent in ceramic dating, andfor that matter in all direct archaeological dating, isthe lack of criteria for estimating the amount of elapsedtime. The only two methods normally useable are depthof deposit and amount of stylistic change, both highlyvariable in rate through time and therefore unreliable.

The short-form dates used in the post-ConquestMaya chronicles and in certain inscriptions, althoughbased on a definite day count which is inherently capa-ble of furnishing accurate information on elapsed time,are of a cyclical nature which does not allow their cer-tain placing in sequence. In pottery studies the sequenceis accurately portrayed, though time duration is uncer-tain. With these two crutches we may walk with greatersteadiness. Mayapan is known to have been a late site.Its abandonment in 1441 or 1446 is possibly the bestsubstantiated of the pre-Conquest dates; Landa statesthat it was occupied for 500 years before that time(see Tozzer, 1941, pp. 33, 37, 38, for discussions).Three inscriptions bearing Ahau glyphs with coeffi-cients have been found at Mayapan. Pollock (1936) ac-

cepted these as katun-ending dates written during themain occupation; Thompson (1945, p. 19) dates themas probably of the Puuc period; E. W. Andrews (ver-bal information given in 1947) professes an open mindon the subject but believes the style of the inscriptions

close to those of Coba' (9.6.0.0.0-9.12.0.0.0); Pros-kouriakoff (1950, p. 164) assumes that they are katunendings, but does not commnit herself on their dating.The ceramic evidence leaves little likelihood of anearlier occupation and is notably free of ceramics ofCoba' Group B date. If these inscriptions are acceptedas coming from the main occupation of the site, theyshould probably be read as katun endings, the type ofdating still in use at the Spanish Conquest. If so as-sumed, the following Christian dates may be read atthe 256 year cycle intervals.

Stela 1Stela 5Stela 6

10 Ahau4 Ahau13 Ahau

Possible Christian Datings1441 A.D. 1185 A.D. 928 A.D.1244 A.D. 988 A.D.1283 A.D. 977 A.D.

Examination of the above table shows that, if thedates were contemporaneously written and actuallybelong to the period documented by an overwhelmingmajority of the pottery, Mayapan must have been oc-cupied from at least 1244 on, the latest date which maylogically be read for Stela 5, since the next 4 Ahaucame at 1500 A.D., at which time the site had beenabandoned. This evidence points to an occupation ofat least 200 years for Mayapan.

Andrews (1942, pp. 261-262) discusses certainfinely cut stones, reused in the roughly built Mayapanstructures, which he suggests may be assigned to theearlier ceramic period there. He also mentions evi-dence of a temple underlying the temple atop the largepyramid at Mayapan which may possibly also belongto my earlier period, although its masonry type isstill unknown. This earlier ceramic period is the Mid-dle Mexican substage of which the characteristicslipped potteries are Coarse Slateware. I know of no

evidence unearthed till now that Coarse Slateware andfinely cut stone masonry are contemporaneous at Chi-che6n Itza', where this ceramic period is best repre-sented. From the Casa Redonda at Chiche'n Itza', H. B.Roberts (Pollock, 1936a, pp. 151-152) tabulates but4.5 per cent of red pottery in a collection of 488 sherdsfrom general postconstructional debris and no red-ware in a collection of 186 sherds from a pit, the con-tents of which must certainly date from before thecollapse of the building (ibid., p. 141). The masonryof the Casa Redonda is composed of rough stone blocks,either unworked or but slightly worked and set in mudmortar, and this type of construction is believed ingeneral to have followed the fine-cut masonry of theMexican architectural period at Chichen Itza' (theEarly Mexican substage). Thus the Casa Redonda, a

rough-block masonry building, dates from either theMedium Slateware period or the Coarse Slatewareperiod; Roberts did not distinguish in his tabulationsbetween the two characteristic wares. This does not,of course, prove that no finely cut masonry was usedin the Coarse Slateware period, but there is as yet no

evidence of its use in that period. Another reasonableassumption is that these worked stones date from theFlorescent stage, which is so strongly represented inYucatan. Unfortunately, the ceramic r epr e sentationof this stage is very light at Mayapan.

The presence of an underlying temple in the Maya-pan Castillo pyramid and the presence of a heavy pot-tery concentration under a floor in ceramic trench 7suggest at least some rebuilding during the occupationof Mayapan. However, the relative scarcity of highpyramids and the crudeness of the buildings on thesite may be taken to suggest a relatively short occu-pation. As has been stated above, lack of stylisticchange in the Mayapan redware deposits also suggests

22


a relatively short occupation-the shortest of any sitewe have sampled on the Peninsula. Therefore, thechoice of a 200- or 250-year duration of occupationseems indicated decidedly over the next choice of a500-year occupation. Although this goes contrary toLanda's statement, the archaeological evidence of thelack of a long overlap between the Chiche'n Itza' andMayapan occupations also controverts the historicalsources. Political histories have notably failed inother instances to agree with the archaeological evi-dence of occupation of sites.

Following back our chronological sequence, thismeans that the end of the intensive building period atChiche'n Itza dates from not later than 1200 to 1250A.D. Working back from that date, we must place theCoarse Slateware period, or a major part of it, thenthe Medium Slateware (Early Mexican substage) pe-riod which bears plumbate and X Fine Orange pottery,and finally the Florescent period, to reach the Mayadate of about 9.16.0.0.0. The time period to be coveredis roughly 500 years if the 11.16.0.0.0 correlation isaccepted, 750 years for the 12.9.0.0.0 correlation, 250years for the 11.3.0.0.0 correlation. The Early Mexi-can and Florescent (Puuc) periods are each charac-terized by the development of highly integrated andelaborate architectural styles, and by the erection oftremendous masses of carefully executed carved stone-work. Compared with the building of Mayapan, theseachievements loom tremendously. 250 years seemstoo short a time for this four period sequence. As tothe choice between 500 and 750 years, this researchconsidered alone yields little evidence (see chart 21for a more extensive discussion).

TICUL

The long prehistoric sequence already attained inYucatan ceramics had been markedly strengthened,checked, and filled in by the excavations at Yaxuna,Dzibilchaltun, Acanceh, and Mayapan. The lack of know-ledge of early Colonial ceramics made still furtherwork on that period advisable in order that the histori-cal anchor of the sequence might be more firmly at-tached. In addition, regional differences between north-ern Yucatan and the Puuc region were suspected, antdI was anxious to determine the regional boundariesalong the north face of the Puuc hills. A trip to Ticulwas first made, on which Dr. Alfredo Barrera Va'squezwas kind enough to accompany and guide us. Methodsof modern pottery making were studied there, and theenvirons of the Ticul monastery explored for possibili-ties of excavation. Such work proved unfeasible becauseof lack of cooperation in the inhabitants of the churchand monastery.

We explored the small site of Chanpuuc about 3 km.south of the Ticul church, on top of the first low rangeof the Puuc hills. There is a group of minor ruinswith a chultun. A trench was sunk in the bottom of thechultun. A collection of pottery (641 sherds, fig. 15,f-n) was unearthed, consisting in the main of typicalFlorescent stage wares but also containing two prob-able Late Mexican substage unslipped jar fragments(fig. 15, f 5, 6), and two atypical incensario fragmentswhich seem closer to Puuc forms than to any other(fig. 15, n, 1, 2). These collections were of some in-terest since they represented our first Florescent col-lection on the north side of the hill range.

DZAN

Following the excavation of the Chanpuuc chultun,

the town of Dzan was visited and considerable evidenceof pre-Conquest occupation noted in the form of re-used cut stones. Nine collections of pottery from Dzanand environs were excavated, all small and all datingfrom the Puuc period and later. No map of the townwas made. Excavations will be described relative tostreet intersections.

Just north of the northwest corner of the main pla-za lies the center and north side of a considerablemound showing part of a standing structure with ve-neer type walls, a medial molding, and a cornice andvault, with plaster roof in place, upon which stands awall seemingly of Spanish construction. The vaultfacing was missing in the area observed. This build-ing must have been encased in a later structure nowlargely gone. Two trenches were dug; trench 1 was25 meters north of the north side of the standing build-ing, and trench 2 was 10 meters north of it. Bothpassed through large stone fill to a thin layer ofsherd-containing earth resting on bedrock. This bot-tom layer contained Puuc-style Medium Slatewareand no earlier wares.

At the northeast corner of the intersection of Ca-lles 19 and 14 is a large sunken area approximately50 by 50 meters and about 3 meters deep, partiallylined by a battered rough-masonry wall. Each sideof the area bears a staircase about 2 meters wide.All masonry seems of Spanish style. This sunkenarea is called a rejoyada (seed bed) by the townspeo-ple, and is used for a banana plantation. No informa-tion as to its original purpose could be obtained. Atleast part of this reioyada is plaster floored. In thecenter seems to have been a masonry-lined depres-sion, probably a pool. The whole depression is re-ported to hold water in the rainy season. I have noevidence as to whether this area was originally ashallow cenote into which embankments have beenbuilt for the streets, or whether it was originally ex-cavated. In either case it represents much human la-bor. Trench 3 was sunk against the north side of thewest staircase in hopes of obtaining a pottery colleo-tion. The trench revealed a retaining wall of heavilyspalled post-Conque st masonry containing occasionalreused Maya cut stone, and a plaster floor resting onbedrock. No pottery was encountered.

In grounds at the southeast corner of Calles 19 and16, a chultun was located and a trench excavated onits floor (trench 4). Thirty meters south of the chultunis a ruin mound showing vestiges of worked stone.About 10 meters west of this mound is another chultun,now filled with modern refuse, which we did not dig.

Thirty meters northeast of the corner of Calles 26and 25 is a building ruin bearing three columns 40centimeters in diameter centered in the building frontand spaced 1.5 meters apart. Vestiges of a sculpturedmolding belonging to this building were found. Largercolumn drums were observed nearby. Thirty metersnorth of the building, a cache of three copper celtsand three copper rings was found by the owner in amortar pit, unassociated with other artifacts. Fivemeters north of this building we excavated a chultun(trench 5) containing a Medium Slateware deposit.The presence of the three columns and the metal sug-gest that this area bore a Mexican stage occupationbut this and the other chultunes are probably of Flo-rescent date. Trench 6, which produced a shallow butrich Medium Slateware deposit, was dug from a chul-tun in Calle 26, 25 meters north of Calle 23.

Trench 7 was dug in a chultun 15 meters south ofthe Dzan-Mani road, about one-half kilometer east ofDzan. The pottery contained a Late Mexican-style ani-mal-head bowl leg, and Late Mexican Coarse Redware.

23


Two pottery collections from the northeast of Dzanand a Puuc-style Medium Slateware molded monkeyhead were brought in by interested land owners. Onthe whole, the individual collections from Dzan wererelatively small and mixed, although together they con-stitute a sizeable sample. The general impression isthat of the usual concentration of ruins within the town,further documenting the establishment of Spanish townson ruin sites. Confirmation is given by the large pro-portion of Maya cut stone incorporated in the walls ofthe large church and of the masonry houses about theplaza. As far as could be told by the pottery, the ruinshad all been Florescent of Puuc type, and Mexican.

After the sampling of the Dzan deposits, which aidedmore in the chronologic placement of the town than inadding to our information on Yucatan ceramic sequen-ces, we established ourselves in Mani, where excava-tions proved more enlightening.

MANI

At Mani, eight trenches were dug in the town cenoteto the southwest of the main plaza. This cenote (map12), still an importance source of water, has been mademore accessible by the construction, in post-Conquesttimes, of a masonry ramp with flights of stairs at topand bottom. This ramp, described by Stephens (1843,vol. 2, p. 259), is now little used in favor of a verticalshaft which was later cut directly over the pool.

The deposit to the sides of the ramp bears mainlypost-Conquest ceramics, but underlying them is a de-posit showing traces of wares of the Mayapan periodin one trench; then, in sequence, evidence of the ear-lier Mexican, the Florescent horizon, a little Regional,and Late Formative, of which a large quantity occurs(see table below).

Below the Late Formative stratum occurred a size-able deposit of a ceramic horizon hitherto undiscov,.ered in Yucatan (see fig. 30, c), and which I have calledEarly Formative on the basis of results from my 1949excavations (Brainerd, 1951). One fragment similar tothis ware has since been recognized in the sample froma Formative deposit at Holactun. A whole vessel, whichfrom its shape may possibly be of a related ware, hasbeen photographed and published by Andrews (Andrews,1943, pl. 28e) but cannot be attributed without examina-tion and study at present unfeasible.17 This specimenwas photographed at Tiradero, north of Ri:o San Jos6Martlr, Campeche. Andrews notes the presence oflarge deposits near there, and I have examined hissamples from these. The deposits are referable to theFormative stage on stylistic grounds; the wares showthe orange paste characteristic of that region, and ap-pear very similar to Formative wares illustrated byRuz (n.d.) from excavations along the Campeche coast.

Totals of diagnostic wares in the cenote trench col-lections are here arranged in their stratigraphic order:

Wares SherdsColonial Coarse Redware 12,068Late Mexican Coarse Redware 812Middle Mexican Coarse Slateware 272Early Mexican Medium Slateware 252Florescent Medium and Thin 1,755

Slatewar e s

Red on Thin Grayware 2Tzakol-style Regional Polychrome 287Late Formative monochromes 776Early Formative Burnished 4,203

Design WareTotal of Diagnostic Wares 20,427Grand Sherd Total 21,310

Per Cent58.94.01.31.28.7

.01.43.9

20.6

100.0

Notable in the tabulation is the strength of the firstand last phases and the virtual absence of the LateRegional wares and Red on Thin Grayware, of whichwe have found large samples elsewhere. The Flores-cent Medium Slateware shapes are distinct enoughfrom those of the Puuc sites to suggest either a re-gional or time difference, and there is enough stylis-tic evidence in their similarity to the Dzebtun collec-tion (fig. 35) in both color and form to believe themearlier than that of the Puuc sites.

The lack of Regional wares of the Early Oxkintokperiod and the presence of Tzakol-style polychromehas been remarked at Mayapan. The same situationexists here, more clearly apparent because of thelarger collections obtained. Trenches 4, 6, and 12(map 12) are the richest in this polychrome. Intrenches 6 and 13, Mexican-period pottery extendsto the lowest levels, thus evidencing some late mix-ing. In trench 4, however, cuts f and . show samplesof the ware with little slateware, and in 4, e, no slate-ware is intermixed (see charts 18-20).

The stylistic ties linking Tzakol-style polychrometo Incised and Punctate Dichrome, and the early strat-igraphic placement of the Mani Orangeware, suggestthat this horizon falls early in the Regional stage. Ifthe Early Oxkintok wares actually date about 9.2.0.0.0,the Mani orangeware may with some reason be as-signed somewhat earlier because of the relative prox-imity of the two sites and the marked lack of similari-ties between their pottery. An alternative explanationof perhaps equal likelihood is that of a marked region-al boundary between the two sites.

Excavations were also carried on in the groundsabout the Colonial church and monastery (map 13).Plans of these buildings and a valuable discussion oftheir dating and use are given by Vega Bolanos et al.,1945, vol. 1, pp. 299-306. Trenches 13, 15, 16, and17 were dug to the north of the northwest corner ofthe monastery, where a well and a room formerlyused as a kitchen gave promise of refuse beds. Trench15 immediately penetrated large stone fill. It was con-tinued to a depth of 2.25 meters in hopes of earlierdeposits, and then discontinued for lack of space andbecause of mounting danger from the loose fill of thewalls. The remaining three trenches yielded exclu-sively Colonial ceramics and trade objects; trench16 contained only scanty collections, the others muchfuller. The trade material has not been intensivelystudied but there is no evidence that any of it dateslater than early in the 19th century; also, Stephensdescribes the monastery as already in ruins in 1842.In the upper levels European china is plentiful, but itis absent from the lower levels. In nearly the lowestlevels are found fragments of majolica, which persistthrough the chinaware period but always in small quan-tity. The majolica bears close resemblance to Pueblawares, but we have no proof as to its origin. Glassfragments in small quantity appear contemporaneouslywith the china. A brown pottery bearing underglazemanganese paint also appears with china, and laststhrough the remainder of the occupation. Pottery ofthis description is made in various parts of Mexiconow; Mercer (1896, pp. 16 1-166) describes the makingof glazed pottery in Merida; the clay was imported,generally from Orizaba, according to his description.Scattered fragments of metal also occur, the lowestof them in prechina strata. The chinaware stratummust begin in the latter years of the 18th century.The Puebla majolica industry was started in the mid-dle of the 16th century, but did not attain much impor-tance or size until the middle of the 17th century (Cer-vantes, 1939, vol. 1, pp. 17-20). Thus we may divide

24


these deposits into two periods, or possibly three: (1)Upper levels dating from late 18th through early 19thcenturies, with chinaware as the diagnostic; (2) Lowerlevels, lacking china but showing majolica, rnid-17thto late 18th centuries; (3) A possibly earlier level intrench 13, i, j and 17, , which lacks majolica but inwhich the collections are too small to prove this lacksignificant. It is also possible, of course, that the ma-jolica in part or all of the collections of group 2 isearlier than mid 17th century in date. It may be addedthat the single plain white glazed fragment of majolicafrom the Dzibilchaltun deposit, which has a probableend date of 1593, is thinner and finer than the Manispecimens; it may be of European importation.A clue to the dating of the local ceramics is given

by Roys (1943, p. 48) who quotes, presumably from theVienna dictionary which he dates from the late 16th orearly 17th century, a reference to a bowl "painted withsmall white pebbles. " The calderon sherds shown infig. 34, b, I 2, meet this description. Painting has beendone with an opaque white pigment on the red slip. Thepigment has been fired and is moderately permanent,though with a tendency to scale off.

The slip of most of the Colonial Coarse Redware isindistinguishable from that of the Mayapan ware. Thejar form is also similar but somewhat slimmer, andnew base and rim forms have been introduced. Theother vessel forms are quite distinct and new. Also,two ware changes are notable; a considerable part ofthe collection bears slip showing a sparkle like that ofspecular hematite, and there is another group of ves-sels with finer paste and a smooth slip showing a morepurplish red than the common ware. In some modernpottery, imported slip pigment is mixed with local redclay (see Mercer, 1896, p. 164), and this practicedoubtless started early; in fact, specular hematitepaint was used during the Florescent and Early Mexi-can stages in Yucatan. Modern Yucatan pottery makingis described elsewhere. It can be said that the 600years beginning with the making of Mayapan potteryand including the present has been the most conserva-tive span we know in Yucatan ceramic history.

The central part of Mani was searched for evidenceof pre-Conquest occupation. A considerable number ofcarved stones from ruins were observed and photo-graphed in house walls on and near the central plaza.The inhabitants were cooperative in reporting aborigi-nal material, including several chultunes; the latterwere explored.A chultun was trenched (trench 7) at Calle 27, no.

228. On the rear of these grounds is a Colonial mason-ry ruin, said by the natives to have been the hospital.Another chultun on the same grounds had recently beenexcavated by the owner. The earth was screened anda pottery collection made (trench 9). Another chultunat the extreme west end of the town was dug, but avery small sample resulted. These chultun collectionsconsisted mainly of early Florescent pottery with lateradmixtur e.

Trench 8 was dug at the northeast corner of themonastery just south of the large corner buttress (forplan, see Vega Bolanos, 1945, p. 299). A small depositwas found, all Colonial of uncertain date, and remainsof a Colonial wall running north and south. Trench 10was dug at the southeast corner of the present church,outside the northwest corner of a cement dance plat-form reputed to have been constructed about 1900 A.D.A few redware sherds resulted.

The rancho of San Jose', also called X'cabenchen,about 2 kilometers south on the street adjoining theMani cenote, was visted. An actun or cave there issaid to lead downward to a pool, but was not explored.

Just inside the cave mouth a trench was dug, yieldinga small sample. A number of sculptured stones werephotographed, including a jaguar head. To the northof the present ranch headquarters, which is a thatchhouse, there is a large sunken area like that describedat Dzan. This area is said to flood during heavy rains,and to drain in three or four days.

The pottery from these deposits, none deep orstratified, consists of mixed collections dating fromperiods varying from Florescent to modern. Thisagain suggests, as at Dzan, an intensive, relativelylate occupation, outlasting the abandonment of the Puucarea to the south and perhaps accounting for the largepopulation which may have drifted or have been movednorthward at the end of the Florescent stage. Unfor-tunately, the ruins which must have stood at thesetowns are nearly all gone into lime-kilns, churches,and masonry houses, and with them must have gonemost or all of the stratified ceramic deposits adjoin-ing the pyramids.

Post-Mayapan-pre-Conquest deposits are to be ex-pected of Mani from the post-Conquest sources whichrelate that the Xius ruled in Mani after the fall of Ma-yapan. This period has not been recognized ceramical-ly. If ceramics of this period exist at Mani, they mustbe among the 'Colonial" collections, which show cer-tain distinctive shape differences from the Mayapanperiod ceramics.

"MOUND NEAR MERIDA" CERAMICS

A collection of twelve potsherds (collection C-20,77) in the Peabody Museum, Harvard, is cataloguedas coming from a 'Mound near Mdrida." The collec-tor was E. H. Thompson. They are illustrated and dis-cussed here because of their interest as evidence ofthe evolution of polychrome pottery painting in Yucata'n.

An unsuccessful attempt was made to locate thismound through a check of Thompson's notes and pa.pers at the Peabody Museum, and through questioningthe late Mr. Thompson's acquaintances in Merida.

There is no external evidence that these twelvespecimens (fig. 63, a, 1-12) come from the same timeperiod. However, the appearance of paste and slip onall of them is so similar that they would normally beclassed as the same ware. There is also overlappingamong them in decorative techniques and form whichlink them in such a manner that it seems justifiableto consider them as a group made during a relativelyshort time span in the same area.

Allowing this assumption, we have clear evidencethat Incised and Punctate Dichrome (see figs. 1, g, i,k- 6; 64; 66, a) was made at about the same time aswere Tzakol-style polychrome bowls bearing geome-tric designs (fig. 63, A, 6. , 9, and12), which conformwith reasonable closeness to certain specimens fromHolmul I (cf. Merwin and Vaillant, 1932, plates 18A,19c) and with San Jose II (cf. Thompson, 1939, fig.4:4).8 The designs on A^, I and _, on the other hand,correspond to those on other Flaky Dichrome bowlsfrom Yucatan sites (cf. figs. 1, j, ;k 6, d, j .2 Ag 'and the designs and forms of the jar rims (fig. 63, a,I 10) are also close to those on Yucatan Flaky Di-chrome (cf. fig. 64). The forms of all the bowls alsoshow Peten similarities, and it is of some interestthat the "basal angle" as described by Smith (1936,plates 2, 16, 17; 4, 13, 16), which begins in Chicaneltimes at Uaxactun and lasts into Tzakol, is used inthese specimens rather than the "basal flange" whichis limited to the later Tzakol phase. Smith also tellsme that the use of rows of black dots as design fillers

25


(see fig. 63, a i, 6, .2) is a characteristic of his ear-liest Tzakol polychromes.

The design repertory of Flaky Dichrome shows sim-ilarities to that of Tzakol-style polychrome'9 closeenough, in my estimate, to make cultural connectionsbetween the two extremely likely. These similaritiesinclude use of heavy stepped elements, rectangularvolutes, and combinations of the two (step-fret ele-ments) as major constituents of the two masses of de-sign. Rows of punctations, used like the painted dotsmentioned above as an early Tzakol characteristic,occur on Flaky Dichrome (see fig. 6, a, 18. 20:

_

O, adding a minor but close correspondence. There isalso marked similarity between the slips of the twowares in Yucatan, and between the interior striationscharacteristic of jars of both wares (cf. fig. 63, bwith fig. 64).

In summary, evidence from the collection underdiscussion, supported by evidence from other Yucatanceramics, makes it very likely that Flaky Dichromeand Yucatecan Tzakol-style polychrome were made bypeoples closely related culturally, and may even havebeen made for a while by a single people contempora-neously. Since most of the Yucatan Regional Poly-chrome is so similar to Uaxactun Polychrome that itis distinguishable from it only with difficulty, it seemssafe to assume synchronism between these two groups.There is evidence from similarities, associations, andoverlappings in ware characteristics and in form, thatFlaky Dichrome closely followed the Late Formativesubphase in Yucatan, which was synchronous with theChicanel phase in the central Maya area. It is alsolikely that the polychrome bowls of the Thompson col-lection are synchronous with early Tzakol for reasonsgiven above.

The hypothesis which best fits this series of prob-abilities is that Incised Dichrome, with its associatedFlaky Redwares, fits between the Late Formative ce-ramics and the Regional polychromes in the Yucatanchronologic sequence. Since both the latter Yucatecanwares can be placed in the central Maya sequence, theposition of Flaky Dichrome, if located between them,is secure. These sequences will then run:

YucatanRegional PolychromeRegional Flaky DichromeFormative Flaky RedwareLate Formative Monochrome

Uaxactun (Peten)Tzakol 1 phaseDi sjunction

Chicanel phase

Since the Tzakol phase at Uaxactun has been esti-mated as having begun at about 8.12.0.0.0 (A. L. Smith,1950, p. 86), somewhat earlier than the earliest Mayastela date, it follows that the Yucatan Flaky Dichromeantedates in its entirety the central Maya Classicstage which is conventionally considered to have be-gun with the earliest dates. I have begun the YucatanRegional stage early enough to include Flaky Dichromeand its associated wares (Yaxuna II) because I feelthat there is a sharper stylistic break before than aft-er this phase, and major time divisions are easier touse if sharp typological distinctions are available fortheir recognition.

The implications of the reconstruction attemptedhere are wide. I have summarized elsewhere (Brain-erd, 1948a) the evidence that Incised Dichrome deco-

rated potteries are widespread in nuclear America,and seem characteristically to constitute a transitionbetween earlier monochromes and later polychromes,which in several regions consist of broad line designsoutlined by narrow black or white painted lines whichseem to have been substituted for the earlier incised

outlinings. Realizing that evidence for the develop-ment of the first polychromes is lacking in the centralMaya area, Thompson, with his usual clear sense ofproblem, surveyed the evidence for their place of ori-gin and, on the basis of information then available,eliminated Yucatin, an area striking in its lack ofpolychrome pottery (Thompson, 1939, pp. 239-243).The new evidence discussed above demonstrates thatour sampling of Yucatan ceramics, as well as of Petenceramics of these time spans, is still so light as togive little grounds even for intelligent speculation onpolychrome origins, save to suggest that they werenot limited to a single area.

THE PUUC SITES

In the years 1932-1936 Henry B. Roberts and H. E.D. Pollock collaborated in a series of field trips, acombined architectural and ceramic survey of Yuca-tin.20 During these field trips, ceramic collectionswere excavated in the Puuc region sites of Sabacche,Muluch Seca, Labna, Huntichmul, Sayil, Kabah, andUxmal.

Of these seven sites, four produced sizeable col-lections of vessel rims; Roberts did not save bodysherds from his excavations. The following data fromthese sites have been drawn from Roberts' notes andchecked against the author's later tabulations whenpossible. All are believed essentially correct; an as-terisk following the number indicates that it was ob-tained by approximation from incomplete notes.

TotalSite Trenches SherdsKabah 21 48,466Labna 34 29,060Sayil 29 55,000*Uxmal 24 53.500*Total Puuc 186,026

Rims2,6702,680*4,1797,51417,047.

RimFr equencv

.0551

.0922

.0721

.1405

Three of these four collections of rim sherds werefound virtually complete in storage; of the Labna col-lection, only 349 certainly labeled sherds were found.Although the Uxmal rim collection was found to be ap-parently complete in bags as collected, only abouttwo-thirds of the bags were labeled. The unlabeledlots of sherds have been treated as separate collec-tions, exact provenience unknown, and analyzed withthe rest. The locations of pottery trenches in the foursites have been plotted from Roberts' notes, with rea-sonable accuracy I believe, on maps redrawn frompublished sources (see maps 14-17), and from the thusfar unpublished Carnegie Institution Yucatan architec-tural survey.

The Puuc site collections, in comparison with Yu-catan sites of other regions, are characterized by aremarkable homogeneity. No more than a few dozensherds from the total of several thousand rims belongto wares not of the Florescent assemblage.21 No othersite of the nine from other regions which are sampledand reported upon in this volume presents a situationapproaching this; all the other sites show at leasttwo well-represented phases marked by separate con-geries of wares. It may be remarked here that al-though the architecture of the Puuc sites has not yetbeen published, it is believed to show a picture ofcomparable homogeneity.

Two possible causes for this situation presentthemselves: either the occupation of the Puuc sitescovers only a relatively short time span, or the oc-cupation has been characterized by a very conserva-

26


tive ceramic industry (and probably also by an equallyconservative architectural style) during its occupation.Until 1949, when sites in the Chenes and Rfo Bec areaswere sampled, I had suspected that the total occupa-tion of the Puuc sites postdated the major occupationof such sites as Yaxuna, Dzibilchaltun, and Acanceh,where wares of the Regional assemblages occurred inquantity with admixture of Florescent assemblageslatewares. Although a close statistical analysis ofthe Puuc area pottery assemblages will not be pre..sented here,22 it should be stated that it seems likelythat the slatewares which characterize the Florescentstage, as characterized from the pottery assemblageof the Puuc sites, originated at a time during whichRegional monochromes were still the preponderantwares in other parts of Yucatan. The Regional andFlorescent ceramic assemblages probably representchronologically overlapping, localized developments,rather than a purely chronological sequence. In fact,it is not certain from the analysis to date that Flores-cent wares did not directly follow Formative wares inthe Chenes and Rlo Bec areas, although it seems cer-tain that the slatewares of the Florescent stage dis-placed Regional wares throughout all of Yucatan wellbefore the beginning of the Mexican stage.

The estimated dating of the Puuc pottery collectionstreated here is discussed elsewhere in the report;there is no good evidence so far for beginnings earlierthan 9.14.0.0.0 for the major occupation of any of them,nor for occupation as late as the Mexican stage in any.These four sites then seem to document a period rang-ing from a time within the range 700 A.D. to 1000 A.D.according to the 11. 16.0.0.0 Maya-Christian calendarcorrelation. Data on other sites of the Puuc area donot lengthen this time range.23 The tremendous size,concentration, and architectural elaboration of thesites, coupled with a relatively short time span, attesta large human population and a remarkable degree ofsocial organization, as well as high levels of crafts-manship. The documented development of such a cul-tural peak in a short time in an area which was butlightly inhabited both earlier and later presents an

archaeological problem of unusual interest.The ceramics.-The unusual uniformity of Puuc ce-

ramics through time has been remarked upon. Statisti-cal studies, as yet incomplete, have suggested thatminor chronological subdivision of the collections isfeasible. Its elements are described and illustrated invarious sections of the report; I shall here character-ize it as a whole.

The crudest pottery is unslipped ware, consistingin the main of large, nearly spherical, round-bottomedjars of a capacity averaging about 6 gallons. Exteriorsurface of these jars is combed or striated, rims are

usually thickened and elaborated, and of exceedinglyeven circular conformation. Minor unslipped shapesinclude hourglass incensarios, incense ladles, andrare bowls. The tradition for pottery of this generalkind seems to extend to modern times from Formativestage beginnings.

Slateware is the predominant type of pottery. Thisware is characteristically slipped with a tightly ad-herent clay of faintly glossy, soapy, translucent ap-pearance. The paste ranges from an off-white to a

medium brown in color, occasionally ranging to darkgray; the slip corresponds closely with paste color,presumably because of translucence. This slip is con-monly decorated by paint which seems to have beenapplied as a somewhat viscous liquid by a crude,broad implement. This paint never shows relief andusually does not change the surface luster of the slip.Its color varies widely, usually on single vessels, and

notably between edges and centers of lines. Colorsometimes fades to invisibility from the center of theline; sometimes a sharp, strongly colored outline sur-rounds a pale center; occasionally "clotted" or reticu-lated areas of color appear as though the slip had beenstained during firing by a paint which had previouslycrazed from shrinkage during drying or heating. Thecolor range of the paint is that of the slip itself, rang-ing from black to a pale gray or tan. Contrasts betweenslip and paint color, although most commonly betweena lighter slip and darker paint, are occasionally re-versed to lighter painted areas on a darker slip. Thiscolor reversal is sometimes apparent on parts of asingle vessel and sometimes appears even in a singlepaint stroke, which as a result may appear dark, fad-ing to a light edge on slip of medium tone.

Variability within this range in paint color seemscharacteristic of many ceramics in North and MiddleAmerica, and may be suspected to be a characteristicof organic paints-paints made of organic extracts-which, when painted on clays of certain characteristics,are carbonized in firing.24 Depending on the atmospher-ic conditions during the heating periods, these paintsmay show carbon black against a light clay color, mayfade to invisibility, or may act as a resist to leave a

lighter slip color in the painted area to contrast witha carbon-darkened background of unpainted slip. Thislast condition, not uncommon in the Southwest,25 hasbeen called rresist smudged by Mera (1945).

No evidence was discovered suggesting regionaland time differences in the paint of Puuc slatewares.However, the paste, and perhaps the slip, do showsuch differences, while form and design also varywidely. These variations among the attributes of Puucslateware seem to have no simply ordered interrela-tionship, and thus have hampered its division into con-

sistent subtype s.Thin Slateware, the only slateware of the Puuc area

which has been separately designated, has been sortedusing vessel shape as the primary criterion. Twoshapes, the beaker and the hemispheroid bowl withtaper or bead lip, are diagnostics (see fig. 50). Potteryof these shapes in the Puuc area characteristicallycan be distinguished from Medium Slateware by thefollowing additional attributes. It is thin-walled (under5 mm. thickness); the slip is light gray in color; thepaste, which corresponds in color, is usually temperedwith fine crystalline calcite. Variants show darkerslip and paste color s, often with other temper types;hemispheroid, taper-lip bowls may be heavily walled.In Medium Slateware, light gray paste and slip colorand fine crystalline calcite temper (Thin Slatewarecharacteristics) appear in some collections on basalbreak bowls. These variants are definitely localizedto certain sites. Basal break bowl variants of ThinSlateware, for example, are particularly common atUxmal, taper-lip hemispheroid bowl variants are com-

monest in the Chenes sites. Thus Medium and ThinSlatewares, here defined primarily by form, inter-grade to a variable degree in slip and paste at thesite s.

The redware of the Puuc sites occurs in lower fre-quencies than does slateware. The slate and redwareshave in common the closeness of color between slipand underlying paste, and the soapy, tightly adherent,translucent appearance of the slip. The redware isseldom painted, and the paint is nearly invisible on

the few sherds showing it. Jar forms are completelymissing in redware, and basins are extremely rare.

Other forms are identical to those in slateware, andshow the same correlations between wall thicknessand vessel shape as those mentioned above for slate-

27


ware. It seems reasonable to consider the redware ofthe Puuc sites as a color variant of the concurrentslateware, used most commonly for the smaller, finervessels, less well suited to painted decoration andowing its color to a difference in the composition ofthe paste, possibly enhanced by a difference in firingprocedure. Supporting this theory is the fact that thethin-walled redware vessels do not show the nearlyexclusive use of calcite temper characteristic of theirlight-gray-colored slateware counterparts. Much ofthe Thin Redware is tempered with volcanic ash.

The remaining pottery of the Puuc collection canwith reasonable certainty be called tradeware. Z FineOrange (Brainerd, 1941) predominates in this group.Pottery from the Peten (central Maya) area is veryrare, but suggestions of trade from the Motagua valleyand from the Campeche-Chiapas area to the west arepresent (see figs. 9, 60). The rarity of tradeware andof recognizable foreign traits may reflect the leadingcultural position of the Puuc during its time period,at least in the Maya area; the distribution of trade-wares may reflect the distribution of important Mayacenters during the time horizon of the Puuc sites.26

The place of pottery in this, the final flowering ofindigenous Maya civilization, is important in the asses-ment of the culture. The pottery is of excellent tech-nical quality, hard, smooth, serviceable, and nearlyfree of flaws. It is markedly uniform and unvaried byodd shapes and designs. Such ornamentation as doesoccur is limited to surface treatments other than poly.chrome painting; the pottery is starkly simple in com-parison with the notable architectural elaboration ofthe ruins in which it is found. Evidence that time-savingmechanical devices were used in the production of thispottery is discussed elsewhere. The conservatism inthe pottery craft of the Puuc area, as well as the lackof regional variation in the pottery of Yucatan duringthis period, have increased the difficulties in ceramicdocumentation. They argue for a stable, long-continuingsocial niche for the potter, as well as for his minorstatus in the society; only thus can we explain suchconservative competence in a period of architecturalbrilliance. Pottery making in the Florescent stage inYucatan was indeed a minor art in contrast to its posi-tion in the polychrome-making phases of the Maya cul-ture of the central area. But the technical competenceof the Puuc potters in general exceeded those of thecentral area, whose wares are soft, with a tendencytoward spalling and weathering off of the slip. The po-sition of pottery among the arts and crafts of the Mayamay be compared to that which it held among the earlycultures of Egypt, Mesopotamia, and the Indus, wherethe elaborate decoration characteristic of the potteryof the Neolithic and Chalcolithic periods was super-seded in favor of utility and speed of production asmen became organized into theocratically governedlarger aggregates. The potter's craft, here as there,is characterized by well-made but somewhat uninspiredproducts.27

There can be but little question that the Puuc pot-ters were at least parttime specialists. This situationis generally true in modern Mesoamerican groups.The men may well have made the vessels; in modernYucata'n only men work at the craft in some towns,both sexes in others. The variability in ware and tem-per in each of the Puuc sites suggests that there werevarious manufacturing centers and considerable trade.This situation exists now in Yucatan and includestrade in materials, as for example between Maxcanuand Ticul. If materials are, as seems reasonable, thebest clue to manufacturing centers, the widely vary-ing proportions of various tempering materials at dif-

ferent Puuc sites suggests considerable trade. Thegeneral distribution of the varieties of shape and de-coration among all temper types suggests that formsand decorations were extensively copied by craftsmenin the various centers.

The uses to which the various types of vessels wereput can in some cases be intelligently guessed. Thestriated jars disappeared before the Colonial periodand direct knowledge of their use disappeared withthem, although their use for cooking is made quitecertain by comparisons of their shape and finish withother American pottery used for this purpose. Modernunslipped ware is used for cooking, and has been sinceColonial times, but it does not bear exterior striationsand is never in the form of jars. Striated jars do notcharacteristically have sooty bottoms when found, butneither do other vessel forms; carbon must have large-ly disappeared from the deposits through time, or char-coal fires may have been used then as now. The usesof unslipped incensarios and incense dippers are wellunderstood from Maya representational art, Colonialdescriptions, and modern Lacandon survivals.

Slateware jars in the main form two types, a largevessel probably slung on the back, and a much smallertype (see fig. 32). The large jar must have been usedfor carrying water on the back for a distance; thesmaller type, from the high proportion of sherdsfound in chultunes as well as from size and form,must have been used for drawing water from thesecisterns. Although the lack of jar body sherds fromPuuc ruins makes the exact proportions of the twotypes hard to determine, the Puuc collections definite-ly have higher proportions of the small jar than dothe Chenes sites or sites of this general horizon innorthern Yucatan. The finding of occasional potterylids for the small jars, as well as of jar rims withinterior ledges and tie holes (see fig. 40,dX, 2364 e, 19 30), suggests the additional use of this ves-sel form for transporting goods, for example, thehoney and wax for which this area is still famous asit was at the Conquest (see Tozzer, 1941, index underexports).

The basins would seem well suited for receivingand carryingmaadz. the wet-ground corn meal whichis the modern Maya staple. The vessel most common-ly used at present is an enameled washbasin. Absenceof clay griddles (comales) in this and all other hori-zons in Yucatan argues against the baking of tortillas,as does the fact that Landa does not mention them(see Tozzer, 1941, pp. 89-91 for preparation of cornin Yucatan). Basins may also have been used for thestorage of maize and other foods in the house.

Two types of bowls, basal break and hemispheroid,are distinct through the period. The hemispheroidbowl would make a satisfactory drinking cup, and mayhave been thus used. The hemispheroid gourd cup nowused dates at least from Landa's time (Tozzer, 1941,pp. 196-197), and evidence of such vessels has comefrom the Sacred Cenote at Chichen Itza (see Ekholm,1942, p. 91 for similar vessels from Guasave, Sina-loa). Both gourds and hemispheroid pottery bowls arenow used in Yucata'n for drinking. Basal break bowlsof varying proportion and decoration have a long his-tory and wide distribution in the Maya area (figs. 107,108). They are the commonest type of vessel shownin Maya art for sacrifices as well as for utilitarianpurposes. They have also been found holding humanbones in subfloor deposits of Maya ruins in Yucatan,and were sometimes used as covers for cached jars.They most probably were used as service plates forsolid foods, and perhaps for offerings as well.

Thin Slateware, Thin Redware, and Fine Orange

28


vessels occur as small hemispheroid bowls and beak-ers, some of which reach vase height. The markedfluctuations in frequency of these wares in the depositssuggests a specialized use, or at least that they wereused predominantly only by certain members of thepopulation. They may have been used as cups; theirfinish and delicacy suggests a luxury or ceremonialware. The very occasional manufacture of redwarebasal break bowls and small basins, as well as thefact that these forms also occur with the paste finishand color of Thin Slateware, suggests that no hard andfast line was drawn in the employment of the finerclays and slips.

Statistical work as yet incomplete shows a gradualincrease of smaller, finer vessel forms, as well asof finer wares, during the occupation of Kabah, andsimilar indications are given by other sites (see Brain-erd, 1951, fig. 92). There is an increase in basal breakslateware bowls, and in Thin Redware and Thin Slate-ware, at the expense of unslipped ware, slateware jars,and slateware basins. The meaning of this trend canbe sought most fruitfully in a gradual social speciali-zation in the sites, which may mirror an increasingwealth and number of priestly occupants of the reli-gious centers. It seems dubious that these ceramicsmirror a gradual change in the economy or food-pre-paring habits of the people. It would seem certain,however, that this pottery sequence cannot be inter-preted as showing a decline in the wealth of the cul-ture; the Puuc sites seem to have been abandoned atthe height of their development.

Settlement pattern.-The intensive short-term de-velopment of the Puuc region, and the marked contrastbetween its former intensive occupation and its pres-ent nearly unpopulated condition, prompt an attemptto determine the causes of such a striking phenomenon.

The Puuc region is sharply marked off physiograph-ically from the areas to the north and to the west of it.The hills which characterize the area rise abruptlyfrom the level northern Yucata'n plain.28 The watertable is far enough below surface to preclude the suc-cess of aboriginal well digging, cenotes are absent,and water-bearing caves are rare and difficult of usebecause of their depth. The scarcity of water duringthe six-month dry season is mirrored in the modernpopulation pattern of Yucatan. The Puuc is lightly set-tled and contains no large villages; although consid-erable milpa farming and bee raising is carried on inthe Puuc, most of this work is seasonal in characterand performed by people who live to the north.

The agricultural possibilities of the area are good.Soil is generally deeper than on the northern plain,and rainfall during the growing season is considerablygreater than that of the more heavily populated areato the north. The present light population of the Puucarea thus may be laid to the inability of the presentinhabitants to obtain during the dry season even therelatively small amount of drinking water to sustainMaya farmers; domestic animals are neither numer-ous nor essential to the Maya milpero.29

The Puuc sites could not have been built withoutlarge groups of laborers. These laborers must havebeen conscripted by a group or groups of men in au-thority, who, from the nature of the buildings and fromour knowledge of Mesoamerican religion, are likelyto have been priests who gained their power throughthe assumed control of agriculture by the round ofMaya religious ritual. There is archaeological evi-dence at the Puuc sites and elsewhere during thepre-Mexican stages in Yucatan to support Thompson'sthesis that Maya sites were not urban centers (Thomp-son, 1945, p. 15). Although a certain number of small

building groups which were probably of a residentialnature are found at and near the sites, it seems likelythat the bulk of the Maya population lived scatteredand near their milpas, gathering at the ceremonialcenters for labor and for religious and political ob-servance s.

The only known source of water in the Puuc areaduring the dry season is from the cisterns or chul-tunes which profusely dot the sites.30 Nearly all thebuildings of the Puuc ruins face upon leveled plazaswhich were paved with carefully laid lime-mortarfloors, and seldom if ever does such a plaza floor ina Puuc ruin not drain into the mouth of one or severalcisterns. The cisterns, many of large size, are bot-tle-shaped excavations into the soft native limestoneor into the plaza fills, lined with mortar and oftencapped at the mouth with a cut stone slab. If buildingin the Puuc sites, with its requirements for plentifulwater, was at least partially restricted to the rainyseason, it seems easily possible that the chultuneswould provide sufficient water for the priestly resi-dents as well as for the constant pilgrims to the ruins.

In addition to the chultunes which occur so exten-sively below the ceremonial plazas, there are eviden-ces of smaller, probably domestic, architectural as-semblages throughout the Puuc area, each consistingof building foundations bordering a plaster-flooredterrace which drains into one or more chultunes. Thedescriptiori and data on distribution of these chultunplatforms must await Pollock's full account of Puucarchitecture. It may be said, however, from Roberts'notes, as well as from reconnaissance by the writerin the area immediately north of the Puuc range, thatchultunes and platforms are by no means confined tothe near vicinity of the ceremonial groups; the settle-ment pattern does not seem to be one of urban, or evenvillage, concentration. In regard to water storage inthe Puuc area, Pollock should be quoted (1935, p. 125):

Reviewing the results of the survey up to thepresent time, it is felt that certain facts dealingwith the environment of the Puuc region, or moreprecisely the effect of the environment upon theancient culture, are beginning to be understood.It is, for example, interesting to note that a regionwhich probably harbors more known remains thanany other area in Yucata'n appears to be the mostfertile agricultural region in the peninsula; alsothat due to the almost complete absence of naturalwater supply, the inhabitants resorted to the arti-ficial storage of rain water in underground cis-terns, known as chultunes, and that great numbersof specialized, platform-like structures were builtfor this purpose.

If our data are correct in general outline, it maybe said that the exploitation of the rich agriculturalarea of the Puuc by the Maya farmer is limited onlyby the necessity of obtaining water for human (notagricultural) use. During the wet season this problemis solved by the presence of frequent rain which fillslocal pools or aguadas. The Maya who lived at thePuuc ruins solved the problem of water for the dryseason by digging chultunes, and this custom seemsto have been limited quite closely in the northern Yu-cata'n peninsula to peoples of the Puuc area and tothose of the related and contiguous Chenes culture tothe south. A group of computations and estimates havebeen made in an effort to determine the limitationson size and structuring of population imposed by thechultun system of water storage. Several factors areknown, or may be estimated.

29


Roberts' observation of 2 gallons per day for aMaya family may perhaps be taken as minimal. Duringthe 1949 field season at Xtampak, we hauled water forcooking, drinking, and bathing 6 men at hard labor inho; weather. A 50-gallon drum lasted about 5 days;this averages 1.7 gallons per man-day.

I compute the capacity of chultunes at Labna in thePuuc area at an average of 7,500 gallons each. E. H.Thompson (1897) gives plans and elevations of 30 chul-tunes. These chultunes are quite regularly hemisphe-roid with flat bottoms. Thompson's drawings werescaled and volumes estimated by the following formu-la: (average radius in feet)2 x height x 3.1416 x 2/3 =volume in cubic feet x 7.47 gallons = capacity in gal-lons. Judged by what chultunes I have seen, the Labnaexamples are typical. The chultunes show a surprisingregularity in size and shape. Although this regularitymay be the result of such physical factors as ease ofexcavation, repair, and cleaning, it is also possiblethat it may furnish a clue to the size of the householdgroup.

The collecting surface necessary to fill the averagechultun during a year of average rainfall is estimatedat approximately 40 square yards. Annual rainfall atTekax, Yucatan, is 1001 millimeters and at Halacho,988 millimeters; both of these figures are a 7-yearaverage (Enciclopedia Yucatanense, vol. 1). Theseweather stations are respectively on the east and westborders of the Puuc. One cubic meter holds 264 gal.lons; assuming 1000 millimeters of rain per year, 28square meters of drainage will fill an average chultun.Forty square yards allows a generous percentage forlo s se s including evapor ation, which at time s must behigh on a warm paved surface after a short tropicalshower. The rainfall divides sharply into a 6-monthseason, May through October, with high monthly aver-ages, and a 6-month season with rain for every monthless than half of monthly average for the year. If thecollecting surface for the chultun were enlarged 50per cent to 60 square yards (540 square feet), it shouldbe full at the end of the rainy season despite constantuse through that period, and thus would be availablein entirety for the dry season of about 190 days.

The area included within the plazas and buildingroofs of the three largest complexes at Uxmal, theGovernor's, the Monjas, and the Palomas group, totalsabout 580,000 square feet, or enough collecting surfaceto fill more than a thousand average chultunes. It isquite certain that not all of the collecting surface wasutilized; to do so would have necessitated a subterra-nean chultun floor area of more than one-quarter thearea of the overlying collecting surface, and there isno evidence that chultunes were this closely spaced.3

A single 7,500 gallon chultun, if full at the end ofthe rainy season (beginning of November), would pro-vide 39.50 gallons per day during a 190-day dry sea-son, assuming no increment of water from rains. Tothis quantity would be added an average of 2,400 gal-lons from the dry season rains off a 60 square yarddrainage area, adding an average of 12.6 gallons perday. Thus 50 gallons per day per chultun is a reason-able yield figure. This means that a maximum of 25families might be kept in minimal drinking and cook-ing water from a single average chultun. Since require-ments of 2 gallons a day per family are minimal, andaverage rather than minimal rainfall figures havebeen used in computing water supply, an estimate often families, or 50 people, per chultun would seemmore realistic. The chultunes are thus of a size whichwould make them sufficient for an "extended family"when found in isolated habitation ruins.32

E. H. Thompson describes 30 chultunes from Lab-

na, several of which were found with stone covers ce-mented in place. It seems probable that at least 30more chultunes remained undiscovered at Labna; chul-tun mouths are not normally noticeable in the bush un-less approached to within 10 or 15 feet, and if coveredare found only by meticulous clearing. Sixty chultuneswould give water for 3,000 permanent inhabitants ifthe above estimate for a single chultun is acceptable.It is our theory that full habitation of the Puuc religiouscenters was only occasional. It seems obvious that areligious center of the Puuc area with functioningchultunes would not need to fear water shortage. It isquite possible that even such water-using occupationsas the laying of mortar masonry and the making ofpottery might be engaged in with moderation duringthe dry season. The drinking water supply of isolatedfamily groups of farmers should provide even lessdifficulty. A single chultun per group would be ade-quate. The building of cisterns in modern times inthis and other areas might again redistribute and al-low a great increase in the population of the YucatanPeninsula. There have been few changes in the farm-ers' mode of life; archaeological evidence may stillbe serviceable here.

The whole problem of agriculture in the Puuc areais worth investigation. Unlike conditions farther tothe south, the sites are not uniformly restricted toraised land surfaces. The landscape, in turn, doesnot show the vegetation division into hill and bajocharacteristic of the Peten and as far north as partsof the Chenes. A careful study of modern agriculturein the area might provide clues to ancient land utiliza-tion.

Chultunes were not found at the ruins sampled inthe northern Yucatan plain with the exception of thosealong the northern base of the Puuc range, at Ticul,Dzan, and Mani. The Mani chultune s were not neces-sary because of the plentiful local water sources avail-able within a few hundred yards of them. They seemto have been all of Florescent date on evidence oftheir ceramic contents. The presence of chultunes atMani thus may be with some reason considered an in-dication of cultural rather than physical requirements.Chultunes were found to be very numerous at SantaRosa Xtampak in the Chenes region, whereas at near-by Dzibilnocac, where the water table is accessiblesave in very dry years, they are replaced by wellswhich were excavated in a manner very similar tothe procedure for chultunes.

This marshalling of data suggests that the concen-trated population demonstrated by the Puuc ceremo-nial centers was made possible by the digging of chul-tunes. In Yucatan, this custom seems to have beenrestricted to the Florescent stage and, although abit less closely, to the Puuc area.

The two large known Mexican-stage sites of Chi-ch6n Itza and Mayapan are located in areas conspic-uous for their concentrations of water-bearing cenotes,and do not contain chultunes associated with theirMexican-stage assemblages. This suggests that thecustom of chultun digging had been dropped at the be-ginning of the Mexican stage. If Thompson's hypothe-sis (1943) of the Mexican Toltec introduction of theurban pattern is sound, it is possible that this innova-tion of the Early Mexican substage was coupled withthe moving of populations into concentrated settle-ments near cenotes on the northern Yucatan plain,quite possibly in order to make political control easi-er. Whatever the causes, the art of chultun digging,so valuable to survival in the Puuc region, was lostand has never been regained.

Dating of the occupations.-The homogeneity of the

30


collections from the Puuc sites has already been em-phasized. There is no evidence in these sites for quali-tative changes in the predominant wares; MediumSlateware, Thin Slateware, and Medium and Thin Red-wares form the bulk of all collections. Z Fine Orange-ware forms a small but constant constituent of thesamples, while the Fine Grayware and distinctive fineorangewares of the earlier horizon at Dzibilchaltun,as well as the X Fine Orange and plumbate of the EarlyMexican substage at Chich6n Itza, are all either com-pletely absent or very scarce and sporadic in Puuccollections. The early Oxkintok assemblage is alsoabsent from these sites, as is all evidence of trade-wares from the Island of Jaina. Peten tradewares, aswell as close similarities to Peten shapes and decora-tion, are also rare.

These evidences, although negative, seem well-enough documented to provide dating limits for the be-ginning of the major Puuc occupation as certainly fol-lowing that of Early Oxkintok (date 9.2.0.0.0), and inthe main following that of Jaina (date 9.11.0.0.0). Itpossibly overlapped, but at its height followed that ofDzibilchaltun, which in turn shows, through the inter-mediate link of Acanceh, a stylistic sequence stemmingfrom Early Oxkintok. In turn Acanceh, as well as Ear-ly Oxkintok, shows links with Yaxuna and Coba GroupB, which bear tradewares and close form and decora-tive similarities to the Uaxactun and San Jose' sequen-ces. This mass of interrelated finds, impossible toassess at present with any show of objective judgmentbut nevertheless impressive to the author, argues thatthe bulk of the Puuc site occupation follows 9.14.0.0.0,as stated earlier in this chapter. The end date of theperiod of major occupation of the Puuc sites is fixedat Chiche'n Itza' by Toltec ceramic and architecturalinnovations, which are discussed later in detail.

The ceramics characteristic of the postconstruc-tional deposits in the so-called Mexican period build-ings at Chiche'n Itza' uniformly show marked admixturesof, and stylistic influence from, X Fine Orange pottery,which is indubitably associated with plumbate ware;neither of these wares, nor influence from them, showsin the Puuc collections. Also, the presence of cachevessels exclusively of Florescent age in the Caracolsubstructures, and concentration of Florescent potteryin excavated collections from 'Maya style" buildingsdefinitely link these two strikingly different ceramicassemblages to the two equally distinctive architectur-al styles. Although no certain stratigraphic evidencefor the sequence from Florescent to early Mexicanceramics is yet available, architectural stratigraphyis definite between the two styles in the Caracol andMonjas complexes at Chiche6n Itza', and is accompaniedby documenting ceramics. In addition, the stylisticand stratigraphic sequences both in architecture andceramics between the Early Mexican and later Mexi-can substages are certain, and thus fix Early Mexicanas a connected part of a chronological sequence.

The evidence for the dating in Maya calendrics ofthe beginning of the Puuc occupation, and the evidencefor its preceding the Early Mexican substage, are

summarized above and detailed at other points in thetext. The dating of the start of the Early Mexican sub-stage in the Christian calendar is estimated as duringthe tenth century A.D. If the period lengths here de-tailed at the latter end of the chronological sequencebe held, the estimation of the length of the major Puucoccupation is dependent on the Maya-Christian corre-

lation. The 11. 16.0.0.0 correlation has been chosenas the most likely; it leaves a time span of somewhatless than three centuries for the Florescent-stagedevelopments of the Puuc area. This may seem a

short time for the building and use of one of the mostspectacular regions in New World archaeology. Sincethere seems to be evidence from material collectedin the Chenes-Rio Bec region in 1949 for considerableantecedents to the cultural developments shown by thePuuc sites, perhaps this time period is not too short.If accepted, the 12.9.0.0.0 correlation would add 256years to this span (see chart 21).

Regional and chronological relationships.-The re-gional variations among the ceramics of the Puuc sitesare potentially of considerable interest. On detailedanalysis, the collections from the four sites revealdefinite quantitative distinctions in ware frequencies,and more marked quantitative distinctions in such de-tails as vessel forms, rim details, and such ware dis-tinctions as slip and temper. They also show that varia-tion among the sites in kinds of temper and slip are

correlated with minor variations in vessel form. Sta-tistical seriations as yet not completely studied sug-gest that much of this variation among collections ischronological, but that there is a considerable residu-um which is likely to be regional. My impression, as

yet unverified objectively, is that the Puuc potterycollections consist of the ceramic products of severalcommunities, each with its own sources of materialsand with its own specialties in wares and shapes. Ifthis assumption is correct, the utilitarian ceramicsof these sites must have been freely traded, the users

of each ritual center relying upon the other groupsfor specific items in their normal household repertoryof pottery vessels. This sort of interrelation is nor-

mal in the folk culture of various Mesoamerican andother areas at present, and it seems reasonable toassume it for the Puuc area in the past.

The study of ceramic trade relationships withinthe Puuc and with the neighboring areas would requiretechnological studies, correlated with precise studieson the minor form variations in the local pottery.More extensive excavated collections than are now

available would be needed. An unusually detailed re-

construction could emerge from the correlation ofsuch ceramic studies with the rich mass of architec-tural data from this region. There is perhaps nowhereelse in the Americas where the archaeological dataon such an advanced and integrated region is as wellpreserved as in the Puuc.

The archaeological chronology within the time pe-riod of the Puuc sites is as yet unknown, save for cer-

tain clues in architectural superposition and othersgained in my still incompletely studied statisticalseriations. At Uxmal, two buildings which Pollock be-lieves show Chenes area architectural features were

partly incorporated within the west side of the plat-form of the Palace of the Governors and thus antedateit (Pollock, 1936, pp. 123-124). In 1947 Alberto Ruzexcavated at one of those buildings, and has verykindly sent me photographs and drawings of a barrel-shaped fine orange vessel recovered with a slatewaredrum in a context which he believes represents a se-condary burial excavated into the building platform(and thus presumably dates from the period of use

of the building). Although it does not exactly matchmy sherd materials, this vessel seems closer toDzibilchaltun Fine Orange than to Puuc Fine Orangein style of engraving and in vessel shape. If this at-tribution is correct, the evidence of placement ofChenes architectural style as earlier than Puuc isthereby strengthened. Further ceramic support forthe placement of Puuc remains later than Cheneswas given by the ceramic survey of Chenes sites in1949. In general, the scanty and somewhat dubiouslyread Maya dates in the two areas also support this

31


ordering, but both the ceramic and epigraphic evidenc-eare sufficiently complex to be to some degree suspectof circular reasoning. However, the evidence is strongfor at least the partial contemporaneity of Puuc andChenes styles of architecture at the several sites ineach of the two areas where both styles occur.

The chronological relationships of the Puuc siteswith neighboring areas are important in the reconstruo-tion of Yucatan culture history. It is unfortunate thatthe evidence at hand is so equivocal. There is evidence,resting mainly on the vessel forms of the slatewarerepertory of Dzibilchaltun, Mani, Dzebtun, and of theChenes sites, for a widespread slateware horizon ante-dating at least the major part of the occupation of thePuuc sites. This ceramic horizon is dated at least asearly as 9.16.0.0.0 at Santa Rosa Xtampak (Brainerd,1949). Among the tradewares of this horizon is FineGrayware which comes in quantity from Dzibilchaltunand is represented by a whole vessel at Dzebtun; itoccurs sparingly at Holactun and at Jaina but, surpris-ingly enough, does not occur at all in the large SantaRosa Xtampak collections, and it is almost absentfrom the Puuc. Holactun Slateware, present in quanti-ty at Uxnal and at Etzna, Campeche, and in smallerfrequencies at the other Puuc sites, is also absent atSanta Rosa Xtampak, a situation even more puzzlingthan the sporadic distribution of Fine Grayware.

My present inclination is to place major occupationsof Dzibilchaltun, Holactun, Santa Rosa Xtampak, prob-ably Etzna, Acanceh, Mani, and the Dzebtun collectionsearlier than the Puuc fluorit, but contemporaneouswith occupational beginnings of certain of the Puucsites, and to support this hypothesis with the evidenceof Chenes architectural influence on certain buildingsin the Puuc. This chronological horizon, according tothe stage nomenclature used here, is early Florescentin the Chenes, Puuc, Holactun, and Mani regions, andpossibly at Chiche'n Itza, where there are suggestionsthat it occurs. At Dzibilchaltun and Acanceh, the samehorizon must be called Regional-Florescent. Thisplacement is given further support by the lack of ZFine Orange in the earlier horizon, and the dating ofthe horizon at 9. 16.0.0.0 by Initial Series inscriptionsat Santa Rosa Xtampak and at Holactun. To fit thishypothesis to various blanks in ware distributionsmentioned in the last paragraph, it must further beassumed that trade in ceramics was sharply limitedby political or ethnic boundaries, or that some wares

and sites covered very short time spans. The moststriking discrepancy to be explained is the lack ofboth Fine Grayware and Holactun Slateware in thelarge samples from Santa Rosa Xtampak. Since con-

temporaneous carved Maya dates come from the twosites and they are only about 70 kilometers apart, a

trade barrier seems the easiest explanation. It shouldagain be cautioned, however, that our information isincomplete, and reconstructions are consequentlyshaky.

HOLACT UN (XCALUMKIN)

The ruins of Holactun lie on the savannah of Xca-lumkin in the State of Campeche, about midway on an

east-west line between the main Puuc sites and theIsland of Jaina off the Campeche west coast. The sitehas been characterized by Pollock (1935, p. 126; 1936,pp. 122-124) as showing architecture suggestive of a

single period directly related to that of the Puuc area,and also to the site of Etzna on evidence of the singlebuilding then visible there. Built into a well-preservedbuilding at Holactun is a Maya date of about 9.16.0.0.0.

Pollock was inclined at the time of his field work togroup the Puuc area, Xcalumkin, and Etzna as belong-ing to an architectural complex distinct from eitherthe Chenes or the Peten areas, and probably not de-rived from them.

Ruz (1945) has since added evidence of an earlieroccupation at Etzna showing ceramic similarities tolate Tzakol and early Tepeu at Uaxactun, and stuc-co-faced block-masonry buildings in general similarto buildings of corresponding date in the Peten and ofthe Regional stage in Yucatan. Inscriptions from Etznarange 9.12.0.0.0, 9.13.0.0.0, 9.15.0.0.0, 9.18.0.0.0,9. 19.0.0. 0. Ruz suggests an additional date of 9. 10. 0. 0.0from the collapsed hieroglyphic stairway; under thestelae dating 9.11.0.0.0 and 9.12.0.0.0 were foundpolychrome ceramics corresponding to late Tzakoland early Tepeu. According to Ruz, the ceramics onthe floors of plazas and rooms are typically Puuc,with many thickened, T-shaped basin rims of Holac-tun Slateware type. A few fragments of X Fine Orangewere found superficially.

The Holactun ceramic collection here discussedwas made by Roberts in 1935 during the Architectur-al-Ceramic Survey. He records (1935 Diary) that theflora of this area is quite distinct from that of the re-mainder of Yucatan, and is characterized by a sharpdivision into savannah areas which are covered heavi-ly with grass and hill tops which are inclined to bewooded. All ruins are on the tops or slopes of emi-nences above the savannah. All modern milpas arelikewise on the high areas, presumably because ofdifficulty in clearing grass, since the natives claimthe savannah soil is richer. This settlement patternshows suggestive similarities with that of the Peten,where the ceremonial centers and house mounds arefound on hills, while the 'bajo s," which are not fardifferent from the Holactun savannahs, were bare ofoccupation.33 Small paved terraces occur on the hill-ocks. According to Roberts, all these were drainedinto chultunes, or more commonly into caves or cre-vasses which thus seem to have been used as cisterns,as were the chultunes of the Puuc area. There are al-so two cenotes near the ruins; the larger of these,described by Maler (1902), does not show in map 18.Roberts reports that this cenote holds much waterduring and after the rainy season, and that the bottomwas moist when he visited the cenote on April 20,1935. The other cenote, although dry, showed thou-sands of potsherds about its brink, suggesting aborig-inal use. It is not certain but seems likely that thiscenote is the depression shown at trench 5, map 18.The trench 5 collections are rich in Formative pot-tery and a surface collection labeled simply Holactun,which may come partially or entirely from the bankof this cenote, contains 60 per cent of Formative pot-tery. A surprisingly large number of Formativesherds, about one-eighth of the total sample from thesite, came from the clearing of the Initial Series Tem-ple. This demonstrates that this plaza location was

established in Formative times, as were so manyothers tested in Yucatan. The general distribution ofFormative monochrome in all collections of any size,as well as its high frequency, suggests that Holactunmust have seen a very large Formative occupation.A single chultun was excavated and yielded sherds ofthe later Puuc-like occupation.

The total pottery excavated at Holactun recordedby Roberts (record incomplete) was 22,152 sherds;rims were saved. The rims tabulated by the writertotal 2,282. Several collections were lost before tabu-lation; the ratio of rims to total sherds on collectionswith both figures available is 0.1307.

32


The position of the Holactun collections in relationto those of surrounding areas is at present unclear,since it was the only site sampled in that area. Theabsence of a deep, stratified series of collections andalso the fact that body sherds were discarded, haveadded to difficulties of analysis. Another series of ex-cavations at Holactun or at another site of the areashould be rewarding.

The Formative collections are unusually rich invariety of decoration and form, and do not matchexactly any other Late Formative collections fromthe Peninsula. Unfortunately, the Formative colleo-tions from Yucatan have never been examined side byside. However, some comparisons can be made fromnotes. Closest resemblances are with the Yaxuna col-lections which, like these, are nearly free of tricklepaint. Collections from Yaxuna trenches 29 and 30 offthe sides of Group 8 are completely free of tricklepaint. On the other hand, the Mani Formative Mono-chrome pottery characteristically shows this paint.Since the Middle Formative collections found in theChenes in 1949 do not show trickle paint, and sincethis type of paint seems to have continued as a tradi-tion to as late as the Middle Mexican substage, it maybe logically hypothesized that trickle paint began tobe used in the later part of the Late Formative sub-stage. The Usulutan type of decoration (Lothrop, 1933,pp. 47-51) as represented by a fragment at Mani (fig.31, c 6) has elsewhere been placed late in the Forma-tive or pre-Classic stage (Kidder, Jennings, andShook, 1946, pp. 182, 242-245), and the more general-ly distributed trait of "negative painting" is placed inan equivalent phase by Wauchope (1950, pp. 233-234and fig. 3). The "negative" character of trickle paintis evident in the Yucata'n samples, adds support to itsplacement in Yucatan Late Formative, and suggeststhat the Holactun and Yaxuna collections may some-what antedate other Late Formative collections.A single fragment of a narrow jar mouth, quite sim-

ilar to those of Early Formative Pattern BurnishedWare and identical with Middle Formative fragmentsfound in the Chenes in 1949, came from the Holactuncollections, and suggests again that the occupationthere may have been earlier than that shown by theMani Late Formative deposits. A further argumentfor dating the Holactun deposits earlier than the ManiLate Formative is the similarity of certain ManiFormative Monochrome bowl rims to Late FormativeFlaky Redware rims (see caption, fig. 17). Althoughall of these comparisons seem promising for a LateFormative chronological subdivision, there must bereservations because of the wide geographic distribu-tion of sites and the sparse sampling.

The Holactun Flor escent stage deposits containlarge quantities of Holactun Slateware. This ware con-stitutes between 30 and 65 per cent of all slipped waresin the sherd lots which are large enough to be reliablesamples. This large percentage is not closely ap-proached in any Puuc site. Uxmal contains 5 to 25 percent, Kabah 0 to 7 per cent with definite evidence thatit is early in the sequence there (see Brainerd, 1951,fig. 92), and Sayil 0 to 3 per cent, all counted with to-tal slipped ware rims as the base. Further occurren-ces of this ware add little light. None was recognizedat Dzibilchaltun, and two fragments come from thusfar undateable strata in the Hacienda Cenote, ChichenItza. To my surprise, the 1949 Chenes excavationsproduced almost none. No Holactun Slateware wasfound in the Oxkintok collections, although certain ofthe Medium Slateware basin rim forms are quite closeto those of Holactun Slateware (see caption, fig. 43).The exclusive or nearly exclusive use of sherd temper

in Holactun Slateware demonstrated by Miss Shephard'sexamination of these sherds, linked to its presence inHolactun Formative Monochromes and its concentra-tion in pre-Classic horizons at San Jose (Thompson,1939, fig. 99) and Uaxactun (R. E. Smith, 1940), addsmore weight to the likelihood that Holactun Slatewaremay range somewhat earlier than do the MediumSlatewares of the Puuc. The absence of this ware atDzibilchaltun and the Chenes sites, however, is noteasy to fit to this hypothesis, particularly since wehave collections from near inscriptions of nearly iden-tical date at Holactun and Santa Rosa Xtampak. TheFine Grayware found at Holactun is paralleled quiteclosely by Fine Grayware at Dzibilchaltun, suggestingcontemporaneity. This ware was certainly widelytraded and has stylistic ties with the Usumacintla andwith the Southeastern Maya area. The near absenceof Fine Grayware in the Puuc suggests that Holactunand Dzibilchaltun are earlier.

The Jaina stratigraphic collections of several thou-sand sherds from the trenches excavated by PavonAbreu in 1942 do not clarify this picture. Sr. Pavonwas kind enough to allow me to examine these in theCampeche Museum in 1942. I did not record and tabu-late the individual collections, but noted no markedstratigraphic changes in them. The Yucata'n waresrecognized were scarce, well under one per cent onrough estimate. They included one rim of HolactunSlateware and a half-dozen sherds of Florescent Me-dium Slateware, two of which, coming from a jar likethat shown on fig. 32, a, were found at a bottom level(1 meter depth, trench III). In the same lot was noteda polychrome fragment showing near identity with abowl from the Merida Museum illustrated as fig. 30,b polychrome fragments were rare in the collections.One sherd of a Fine Grayware beaker was noted, andtwo sherds of X Type Fine Orange. Much of the localpottery was of a medium-textured, orange-coloredpaste; a few fragments of seemingly untemperedorangeware were noted. I found no figurine fragmentsin these collections. Ruz, in analyzing these collec-tions with comparative material from Campeche andelsewhere (1945, pp. 71-72), finds two main periods,the first similar to a period found by him at the cityof Campeche, at Etzna, and at Oxkintok,34 and characterized by painted pottery which he believes contem-poraneous with late Tzakol and early Tepeu at Uaxac-tun. Ruz places the hand-modeled Jaina figurines inthis period. His second period shows the disappear-ance of painted pottery, the presence of vases deco-rated with bands of hieroglyphics, slateware, and a

type of fine orange ware not corresponding either tothe Puuc or Chiche'n Itza types. The mold-made Jainafigurines are placed by Ruz in this period. Followingthis, Ruz describes a poorly represented Mexicanphase marked by fine orangeware similar to the Chi-chen Itza type and by the common wares of Yucatanat this period.

Ruz' information on periods adds light to the pio-ture and also poses problems. It looks as though slate-ware may occur in the horizon which Ruz equates withlate Tzakol and early Tepeu as well as in the laterdeposits, although the deposits from Jaina are likelysomewhat mixed and therefore unreliable for closechronology. The placement of the Jaina mold-madefigurines in the Puuc period invites the question ofwhy not a single fragment possibly attributable tosuch a figurine has come from documented Puuc col-lections. The Puuc figurine heads are of slatewarepaste and in a distinctive style which has similaritiesto figurines as far west as Cerro de Las Mesas. Fig-urines generally similar to the Jaina molded figurines,

33


both in style and in their fine-grained orange-coloredpaste, are also found as far west as Veracruz. Theabsence of Jaina pottery and figurines in our Chenescollections as well as in those from the Puuc suggeststhat surprisingly little trade, rather than a chronologi-cal difference, may cause this lack of ceramics com-mon to Jaina and the Puuc area. The relative scarcityof ruins on the coastal plain inland from Jaina, asnoted by Pollock (1940, p. 267), suggests that this lackof culture contact may be due directly to sparse popu-.lation in the intervening area. The political situationduring the Yucatan Regional stage must have been ofa peculiar sort; the presence of monuments with thesame elaborate calendrics at Jaina, Holactun, theChenes, and to a lesser degree in the Puuc area ar-gues for a common priesthood, but the peoples usingpottery in these areas (the common folk) seem to havetraded with each other in a peculiarly selective man-ner. We have evidence of considerable ceramic tradeand influence between the Puuc and Chenes; amongHolactun, the Puuc (particularly Uxmal), and Etzna;and between Jaina and the City of Campeche area. Thedistribution of Fine Grayware, which links Holactun,Dzibilchaltun, and Santa Rosa Xtampak, is a furtherfactor to be considered in what is at present far froma clear situation.

It seems obvious that any reconstruction of the ori-gins of the culture of northern Yucatan will not be oncertain ground until the archaeology of the westerncoast and adjacent regions of Campeche and Tabascois known. From the sporadic finds thus far recorded,it seems safe to assume that this was an importantarea from Formative times on. The early Oxkintokfinds suggest influences of this area on the Puuc, andthe Holactun-Etzna similarities suggest a later, iso-lated province which brought western influences to thePuuc region. The area is definitely in need of furtherwork. The publication of Ruz' coastal reconnaissance(Ruz, n.d.) would add considerable detail to generalknowledge of this area.

CHICHEN ITZA

The Carnegie Institution pottery collections of Chi-chen Itza were accumulated over a long period and bymany persons during excavations made under the di-rection of Dr. S. G. Morley. This report is in largepart the result of their labor; I have made no excava-tions at Chichen Itza. Beyond a brief report by Robertson the pottery from the Casa Redonda (in Pollock,1937, pp. 151-152), Roberts' brief progress reportsin Carnegie Institution Yearbooks (Roberts, 1931,1932, 1933, 1934, 1935), Vaillant's preliminary dis-cussion in his doctoral thesis (Vaillant, 1927, pp. 335-367), and two short discussions of his excavations inthe Initial Series group, no formal reports on the ce-

ramics of Chichen Itza are available. A few illustra-tions of whole pieces of pottery have been publishedin Carnegie Institution reports on the excavation ofthe various buildings, but no attempts at ceramicanalysis were made in any of these.

Vaillant (1927), later confirmed by Roberts, notedthe occurrence of three sequent ceramic periods:First is a Maya, or pre-Mexican, or Puuc period inwhich the ceramics show similarities to those fromthe Puuc area. Following this is a horizon which hasvariously been called a period of Mexican, Nahua, or

Toltec influence, during which Isla de Sacrificios FineOrange pottery and plumbate pottery were importedto Chichen Itza and numerous other evidences of Mexi-can mainland influence are apparent in architecture,

sculpture, painting, and chronicled history. Finally,a period of Maya Resurgence was described duringwhich Mexican influence was less strong, and the ce-ramics resemble the earlier Maya wares (Vaillant,1927, 1933; Roberts, 1934).

The major slipped ware of the first two periodswas called slateware by Vaillant, and that of the lastperiod named red lacquer ware. Two of these periods,the Maya and the Mexican, agree in general with thestratigraphic and architectural periods defined forthe site by Morley (Morley, 1931, pp. 106-107). Thelater Maya Resurgence period has not been recognizedarchitecturally save perhaps in the Casa Redonda(Pollock, 1937, p. 152), which, however, from its ce-ramics should definitely be classified into Vaillant'sMexican period, and perhaps toward the close of theperiod.35 Vaillant's ceramic evidence for a Maya re-surgence seems to have rested on his belief that redlacquer ware is Maya in inspiration, but the hypothe-sis of a Maya resurgence probably rests more ondocumentary than upon archaeological evidence. Pres-ent ceramic evidence indicates little if any drop offin foreign influence during the later part of the Mexi-can stage.

As the result of a Carnegie Institution conferenceat Chichen Itza in 1931, Morley (1931, p. 106) describedfour periods of occupation, supported by a combinationof architectural, epigraphic, and documentary evidence.Period I was based upon the date of the Initial Serieslintel with a long-count date of 10.2.10.0.0 which Mor-ley believed to antedate the period of the Puuc ruins.According to the dating scheme used in this report,this lintel belongs in Morley's period II along withthe architecture of the pre-Mexican period. Morley'speriods II, IIU, and IV correspond with those used byVaillant and Roberts.

Morley's period II belongs in my Florescent stage.My reanalysis of the Chich6n Itza ceramics has estab-lished an additional period characterized by a changein the major slipped ware, which allows a subdivisionof the Toltec or Mexican period (Morley's III) into myEarly and Middle Mexican substages, and Morley's IV(Vaillant's Maya Resurgence) chronologically fits myLate Mexican substage. Although there is a consider-able amount of Vaillant's red lacquer ware in the Chi-che'n Itza collections and it predominates in the slippedware of E. H. Thompson's cenote collection, there isat present no evidence that any of the buildings at Chi-ch6n Itza' were built during the red lacquer-ware hori-zon, and evidence of but little construction during mynewly established Middle Mexican substage which pre-ceded it. Thus we have a sequence of four major ce-ramic periods at Chichen Itza, which I classify as

(1) Florescent stage, (2) Early, (3) Middle, and (4)Late Mexican substages. Of these, architecture isknown certainly only for the first two. This sequenceplaces the close of the architectural period at ChichenItza early in the 13th century, if the Mayapan occupa-tion (almost completely Late Mexican) can be assumedto have lasted 200 years. The Mexican-influencedoccupation at Chiche'n Itza saw the erection of a largenumber of massive buildings and a distinct change inthe major slipped-pottery ware of the site. If the es-

timated length of the Mayapan occupation is taken as

a yardstick, an admittedly dangerous procedure, theMexican occupation may be estimated on archaeologi-cal grounds as beginning sometime during the 10thcentury.36

The ceramic and architectural periods at ChichenItza are equated in the table on the following page forpurposes of clarity in reading the account which fol-lows.

34


Architectural Ceramicnone Late Mexicannone ? Middle MexicanMexican or Toltec Early Mexican(sometimes subdivided)

Puuc or Maya Florescent

It would seem that the abundant ceramic materialfrom Chiche6n Itza should furnish a legible, detailed,and accurate account of the culture history of the city.However, the widely varying methods of collectingsamples and the secondary position to which ceramicswere relegated during much of the excavation havecomplicated the task of their study. More serious hasbeen the deterioration of labeling during the long pe-riod of time between my study and their excavation,and the fact that their initial study was interrupted inmid-career; Henry Roberts' ceramic analyses madefrom 1932 through 1935 were terminated by ill health.It was unfortunately necessary to begin them afreshrather than to continue where they were left off. In aneffort to date certain of the buildings at Chiche'n Itza,Roberts' tabulations have been used in cases wherethe collections have since disappeared.

The extensive and detailed architectural studies ofthe Carnegie Institution at Chich6n Itza' are now nearlyall published.37 The following account relies to a ma-jor degree upon information contained in these reports.I have compiled maps from material which was unpub-lished, or published in less usable form, at the timeof writing to aid in the understanding of the ceramics.A close inspection and analysis of the collections,

which included complete tabulation into some 200 ce-ramic types, and the graphing and sorting of the fre-quencies of wares and many vessel forms has thusfar failed to produce a detailed sequence of ceramicdevelopment through the architectural constructionperiods of the site, although the major periods, asnamed above, are clearly shown by evidence fromChichen Itza' and elsewhere. A close sequence wouldunquestionably be of value if only to allow the check-ing of historical reconstructions based upon such cri-teria as stylistic variations in sculpture and painting.The history of Chiche6n Itza through these times repre-sents one of the most radical and best-documentedarchaeological examples of acculturation available.Its detailed study would be of great interest. A statis-tical seriational analysis of the ceramic tabulations,begun in 1949 under a University of California grantbut postponed to allow the completion of this report,shows signs of giving at least parts of the desirableclose ceramic sequence (see Brainerd, 1951). Timeand funds permitting, it will be completed later. Fur-ther ceramic excavations would unquestionably yieldimportant results. Some of the major lacks in ceram-ic sampling to date include the absence of the follow-ing: (a) demonstrably pure Florescent samples, (b)pure Middle Mexican substage samples, (c) subfloor(construction period) samples from certain structures,the dating of which is important.

In the descriptive section which follows, ceramicdating by period is discussed for various of the archi-tectural groups. The architecture from which we havesamples dates almost exclusively from Florescentand Early Mexican times. As may be seen from theware descriptions, the primary determinants betweenthe native ceramics of these two periods are shapeand decoration. The distinctions in paste, slip, andtemper are of only limited and secondary value. Ithas not been possible in sherd collections to separatecompletely the ceramics of the two periods, since theknown stylistic determinants are recognizable only

on certain of the sherds. The stylistic criteria for theEarly Mexican substage Chiche'n Itza' Medium Slate-wares and Medium Redwares were selected from thosenot found on the same wares in the Puuc collections.

The validity of this procedure, the best availableat the time of sorting in 1940 and 1942, depends uponthe assumption that the traits selected as period de-terminants are actually such, rather than regionalvariants. This assumption has since been given sup-port by the results of later excavation, but the Flores-cent collections from Chichdn Itza' or immediate vi-cinity which would allow the final chronological proofare not yet available. Since 1942, two vessel typeswhich were used as chronological determinants, thepedestal (see fig. 71, e-i), and the grater bowl (seefig. 74, h-ij), have been found in pre-Mexican associa-tions in the Chenes and Rio Bec areas and are therebydisqualified as exclusively Mexican determinants. Themajor changes in local ceramics which we considerto be diagnostic and the changes in imported ceramicsfor these two periods are listed below to aid in thereading of the following text. They are detailed in thefollowing section on ceramic description and summa-rized in the section on results.

Florescent(Puuc sites)

Medium Slateware:Jars:

Basins:

Basal break bowls:

Rounded bowls:

Thin Slateware:

Medium Redware:

Fine Redware:

Plumbate Ware(import):Fine Or an-ewar e

(import):

Early Mexican(Toltec Chiche'nItza)

"Chultun jar" Concaveforms; perfo- neck; grrated lug neck.handle s.Thick angled Oval crbolster rims. tion bolSharp, thick- Curved,ened basal ened babreak; slab spheroilegs. feet.Hemi spheroid Shallowform, exterior no rim.faceted rim; no ringring base.Present in dis- Absent.tinctive formrepertory.Forms like FormsFlorescent ration fMedium Slate- Fine Orware. pertory

Present in dis-tinctive formrepertory.Absent.

Z (Puuc) FineOrangewar e.

e cylindricrooved

oss sec-.ster rims.unthick-

sal break,.d rattle

forms;facets;bases.

and deco-rom Xrange re-m,much

incised decora-tion.Absent.

Effigy (Tohil)Plumbate.X (Chichen) FineOr angewar e.

Akab'dzib Cenote. -Five trenches were sunk byHenry Roberts in 1932 in the depression east of andabout 50 meters behind the Akab'dzib (see map 19,4D1). The Akab'dzib is classified architecturally as

of Florescent date, and bears a lintel dated by Thomp-son (1937) at 10.2.1.0.0. I know of no Mexican-periodarchitectural evidence from its vicinity. The trenchesranged in depth from 1 to 2 meters. Sherd yields were

small (332 sherds total) and no stratigraphy was noted,with the possible exception of trench 1 which suggestsFlorescent-Mexican change, but samples are toosmall for certainty. Medium Slateware was prepon-

35


derant in the collections, and no Coarse Slateware orCoarse Redware occurred below the top levels. Twosherds of Z Fine Orange come from trench 2, b (fig.89, e, i), but no X Fine Orange was found. Three sherdsof plumbate occurred, all thin, weathered, and of dullbrownish color. Thus the deposit is limited to Flores-cent and Early Mexican date by its local wares, withindications of both periods in its tradewares.

Analysis of the forms of the native wares placesthem as largely Florescent in date. One Puuc-styleslateware thickened jar rim appeared; this is an ex-ceedingly rare type at Chichen Itza', as not more thana dozen came from the large Monjas collections. Of64 slateware bowl rims, 57 were of typical Puuc form,with only 7 of Early Mexican style. Of 48 slatewarebasin rims, 32 were Puuc and 16 of Early Mexicanform. 31 grater bowl fragments (Early Mexican) areincluded. Medium Redware vessel forms differ fromthe Puuc style as shown by the presence of 7 jarsherds; jars of this ware are very rare in the Puucbut common in the Early Mexican substage, where al-so appear rattle tripod and pedestal base vases whichare absent in the Akab'dzib Cenote collections. Sever-al sherds of the distinctively Florescent Thin Slate-ware are induded.

These collections show less Early Mexican potterythan most other groups from Chich6n Itza, and thushave an interest out of proportion to their small size.The following hypotheses are suggested by their study,but unfortunately, due to the small size of the samples,are not proved. (1) Plumbate may have preceded thefirst importation of X Fine Orange at Chiche'n Itza,and been contemporaneous with the later end of the ZFine Orange period.38 (2) Medium Slateware Basinforms may have changed earlier than did bowl forms(which hypothesis is borne out by the Monjas collec-tions), and the grater bowl, one of the original sortingcriteria for the Mexican stage, appeared at ChichenItza' during Florescent times (also borne out to somedegree in the Monjas collections, and by the presenceof rare grater bowl fragments at Uxnal and in theChenes and Rfo Bec collection s).

Southwest Group (5B12-21).-A series of trencheswas dug by Henry Roberts during 1933 in the SouthwestGroup, off the outer walls of terraces (see map 19,area 5B). Although some of these collections weremisplaced before my study, and the labeling does notcorrespond with Roberts' field notes, the collectionsremaining seem intact and total 406 sherds. The dis-crepancies between labeling and field notes cast some

doubt on the location from which the separate colleo-tions come, but I believe that they are all from theSouthwest Group.

The collections are approximately uniform in ware

percentages; the faulty labeling does not allow the in-vestigation of stratigraphy. The preponderant slippedware is Medium Slate. Accompanying the collectionare seven sherds of X Fine Orange, three sherds ofZ Fine Orange, and four sherds of plumbate. As in theAkab'dzib trenches, evidence points either toward theFlorescent-Early Mexican transition or toward mixeddeposits of the two horizons. No Coarse Slateware,and only two sherds of Coarse Redware occurred.Medium Redware sherds in Early Mexican forms oc-

cur, which forms are absent from the Akab'dzib Ce-note collections. The proportions between Puuc andMexican shapes in Medium Slateware also point to a

later dating than that of the Akab'dzib Cenote trenches.There are 37 Mexican-stage basin rims to 17 Puuc,but 34 Mexican-stage bowl rims to only two Puuc. Thebody of the collection certainly dates later than thosefrom the Akab'dzib, and the three Z Fine Orange

sherds should not be regarded too seriously in theface of the evidence of lateness in other criteria.

Small group iust east of the Sculptured Jambs(5B4. 5?).-Two collections of 175 and 240 sherdslabeled with this location were found, representingtwo levels of a single trench. Roberts' field notesfurther specify that the trench was dug on the northside of the platform near the northwest corner. Themap of Chiche6n Itza' shows two groups meeting thesespecifications, one bearing buildings 4B 2 and 3 and5B 3, the other bearing buildings 5B 4 and 5. The lat-ter seems the more likely location as it is near theTemple of the Sculptured Jambs. These collectionsare very similar to those of the Southwest Group. OneZ Fine Orange sherd was present. Medium Slatewareand Redware are largely Early Mexican in form; therewere 56 Mexican basin forms to 6 Puuc and 23 Mexi-can bowl forms to 5 Puuc. Middle and Late Mexicanperiod pottery are completely lacking in these collec-tions.

Caracol (3C15).-The Caracol at Chiche6n Itza' (map20) was investigated and restored by the Carnegie In-stitution during the years 1925-1931 (Ruppert, 1935).From architectural criteria it is generally believedto date from the transitional period ushering in theso-called Mexican occupation of Chichen Itza (Morley,1931, p. 107; Pollock, 1936a, pp. 103-104).

The sherd collections from the Caracol total 6,523sherds in 34 lots. Since the work at the Caracol wasdone with the investigation and reconstruction of itsarchitecture as the prime aim, the deposits were ex-cavated only where this aim would be served. The ce-ramics thus obtained consist almost entirely of frag-ments accumulated during occupation of the buildingand in debris left by later visitors. The great majorityof the collections contain material characteristic ofat least three of the periods which we have definedelsewhere from less mixed deposits.

No collections completely lack Coarse Redware,but none contain much Coarse Slateware. A group ofcollections of nearly unmixed Early Mexican materialcome from the area east of the Bench House, includingits South Annex. The intentional razing of these build-ings (Ruppert, 1935, p. 270) is thus dated as in theEarly Mexican period.

Medium Slateware constitutes over half of theslipped ware in all collections analyzed, save fortwo general surface collections and two other collec-tions coming from the south side of the West Annex.This fact limits the major occupation to Florescentand Early Mexican times.

Evidences of an earlier occupation include a fewFormative and more early Regional Flaky Redwaresherds. These come from the base of the Lower Plat-form Stairway and from the West Annex. Similar ma-

terial comes from some of the Monjas excavations.Ruppert found occupational levels well below theground surface in an exploratory trench at the baseof the Lower Platform Stairway (Ruppert, 1935, fig.60) and it is possible, though unproved, that the earlysherds were deposited originally at those levels.

Fortunately for ceramic dating of the main struc-ture of the Caracol, several vessels were foundcached during its construction (see map 20 for loca-tions). In the absence of sherds which definitely ante-date parts of the construction, these pieces are ofprime dating value. The vessels were as follows:Field catalogue numbers are given; page and illustra-tion references in parentheses are to Ruppert, 1935.

25 cm. below the lower platform floor on northside of upper platform, a striated unslipped jar

36


and a Medium Slateware bowl (cat. nos. 1048, 1144;pp. 38, 42, figs. 46, 47), see fig. 67, a.

In cyst in lower platform at center of first cir-cular upper platform, sealed in place by construc-tion of this platform, an unslipped, striated jar (cat.no. 1035; p. 86, fig. 102, a), see fig. 68, e.

In upper rectangular platform 20 cm. below floorunslipped striated jar (cat. no. 1049; p. 119, fig.144, a, b).

In upper rectangular platform, base 48 cm. un-der floor, unslipped striated jar (cat. no. 1103),Medium Slateware rounded bowl (cat. no. 1137; pp.119, 120, figs. 145, 146), see fig. 74, f.

Between outer circular platform and masonryblock dividing upper portion of stairway, associatedwith burials, "sherds of a plain incense burner ofPorous Gray Ware" (cat. no. 1073; p. 123).

In a cyst within upper rectangular platform, un-slipped striated jar (cat. no. 1194; pp. 124-126) con-taining a cremation burial.

Under the upper rectangular platform on floorof stylobate in occupational deposit 'sherds ofsub-Plumbate Ware, Porous Gray Ware and a GrayWare with raised knob decoration. " This deposit'definitely antedates the rectangular platform, asthe stairway and niche of the latter rest upon it"(p. 150).

Under the floor of the outer chamber of the Cara-col Tower, an unslipped turtle effigy of Coarse Gray-ware decorated in unfired bluish-green, yellow, andblack paint (cat. 1093; p. 219, fig. 278, ,), see fig.93, e; an unslipped miniature Coarse Graywareconical jar painted red on white and black (cat. no.1100; p. 220, fig. 279, a); a Coarse Redware jar (cat.no. 599, p. 220, fig. 128, b), see fig. 94, c; two smallunslipped jars (cat. nos. 658, 739; fig. 279, ;, c);and an unslipped figurine of a jaguar painted white(p. 221, fig. 128, c), see fig. 93, d.

The above ceramics from the Caracol divide intothree chronological groups. The slab-legged, MediumSlateware basal break bowl is very close to the mate-rial from Yaxuna, Dzebtun, and Mani, and dates eitherEarly Florescent or is a regional variant contempora-neous with, though typologically earlier, than the Puucpottery. The rounded Medium Slateware bowl found inthe upper rectangular platform is probably an EarlyMexican form. The Coarse Redware jar found underthe floor inside the west doorway of the Caracol super-structure is of Late Mexican period, as are probablythe turtle and the miniature jars. The other specimenslisted are less surely diagnostic, but in general theysubstantiate the above chronology. There is no positiveproof that the three slipped-ware specimens describedabove did not considerably postdate the structures inwhich they were found, although, since they are de-scribed from under floors, the supposition is that theydate from either the period of construction or at leasta period during which floors were still being replaced:a period of occupation rather than one of visitation bytransients. A building as spectacular as the Caracolhas certainly seen a constant stream of visitors eversince its construction.A number of other whole or restorable vessels have

been found at the Caracol, the majority of them datingfrom the Late Mexican period and being chiefly smallCoarse Redware bowls such as were found containingcopal in the Sacred Cenote. The surface of the site wasstrewn with such ware and with fragments of figurineincensarios. The people bearing these were probablypilgrims to Chiche'n Itza shortly before the Conquest(see Tozzer, 1941, pp. 109-110).

A puzzling find is that of an unbroken plumbate jar,fig. 91, f. in the talus slope of the tower just under thefive-member cornice on the west side. We have no evi-dence of plumbate ware in Yucata'n in deposits laterthan Early Mexican. It does not seem likely that thetower collapsed as early as this period, particularlysince we have evidence of floors being laid in the outerrooms of the tower in Late Mexican times. Either thejar was an heirloom of long standing at the time it wasleft in the talus, or it had been deposited in the upperpart of the tower and slumped with the talus, remainingunbroken in the fall. The latter explanation, althoughextremely unlikely at first thought, is supported by thefinding of a large section of the tower slumped unbrokento the platform on the east side. Perhaps the jar wasprotected by such a mass, which has since disintegratedthrough time.

If the above explanation is accepted, the construo-tion of the tower as well as the upper platform datesfrom the Early Mexican substage, and reflooring ofthe interior for the last time probably occurred in theLate Mexican substage. The lower platform dates fromthe Florescent stage. The Middle Mexican substage isbarely represented in the ceramic samples, while therelatively abundant ceramics of the late period comefrom pilgrimages to the site.

Although this outline is not documented very fullyby the small number of whole specimens found asso-ciated with the structural sequence, it is supported tosome extent by the gross counts of sherds of the vari-ous periods in the collections. It should be consideredas confirmatory to architectural dating, and estab-lishes several new facts and closer datings. The sig-nificance of the whole pottery in the position in whichit was found should be understood. The Florescentbasal break bowl must date with or after the lowerplatform, and the rounded, probably Early Mexicanbowl with or after the upper platform. In both cases,the pottery is probably contemporaneous rather thanpostconstructional. The plumbate jar, if our preferredhypothesis is accepted, is contemporaneous with theconstruction of the tower, while the Late Mexican jarand figurine precede or are contemporaneous with thelatest floor in the outer range of tower rooms.

Mercado (3D11).-The Mercado was excavated andrestored during 1932 under the direction of Karl Rup-pert, and the results have been published (Ruppert,1943). The building was seemingly erected as a singleunit upon the southern edge of the Court of a ThousandColumns, and has been dated in the Toltec period, ashave all buildings in its vicinity. It is similar in planto the Temple of the Hieroglyphic Jambs (see Ruppert,1943, p. 233 for discussion), from which a ceramiccollection is described. The accompanying map (map21) shows both the Mercado and the contiguous South-east Colonnade, also excavated by Ruppert, ceramicsfrom which will be discussed in the following sectionof this report.

Thirteen sherd collections from the Mercado havebeen tabulated totaling 622 sherds; one collection, con-taining only six sherds, is from within the dais. Theremaining collections, as far as can be judged, comefrom debris overlying the structure and resulted fromthe process of clearing it.

The collections all contain Medium Slateware vary-ing from 25 to 40 per cent of the total slipped ware,and most contain X Fine Orange which ranges fromabout 10 per cent to zero. The predominant slippedware in all but two collections is Coarse Slateware,the Middle Mexican diagnostic. Coarse Redware isabsent in several collections and in no case runs over10 per cent of slipped wares. The two collections

37


bearing small percentages of Coarse Slateware range66 and 63 per cent of Medium Slateware in samples of76 and 74 sherds respectively, and are thus predomi-nantly Early Mexican. Both are labeled 'Mercado gen-eral. " The exact provenience of these two "general"collections is not known, and it seems quite possiblethat they came from the Southeast Colonnade, dug atabout the same time as the Mercado proper. Theirearly dating as compared with the bulk of the samplesis disturbing, since collections resulting from the gen-eral clearing of a ruin normally date later than colleo-tions from the more specific and later work of clearingthe component parts of a structure.

Of the six sherds from within the dais, one is ofCoarse Slateware, and all the others are unslipped.Thus, subject to some uncertainty since Coarse Slate-ware had not been defined at time of excavation and thecollection may possibly have been mixed during or sub-sequent to excavation, the dais is datable as of the Mid-dle Mexican substage. The dais was a secondary con-struction, contemporaneous with the contiguous bench(Ruppert, 1943, pp. 245-250). As Ruppert shows, theform and certain of the elaborate decorative featuresof bench and dais link them to structures in the Templeof the Warriors, to certain Mexican elements knownfrom codices, and to the frescoes of Santa Rita. On thewhole, these similarities seem rather generalized (al-so see discussion of design placement in Andrews,1943, pp. 74-76).

The most likely placement of the construction of theMercado would then seem to be near the end of theEarly Mexican substage or at the beginning of theEarly-Middle Mexican transition period, with renova,-tion consisting of the building of the bench and dais ofthe gallery during the Middle Mexican substage. Itwould be interesting to know when the amazingly light,broad vault of the northern gallery fell. The collectionscoming from the area of this vault all show CoarseSlateware, but nothing later. The collections are small(only 63 sherds in all), but Coarse Redware should bepresent in these collections if it was in use while thevault stood; the vault must have fallen before the endof the Middle Mexican substage.

Southeast Colonnade (3D10).-The Southeast Colon-nade was excavated by Ruppert in connection with theexcavation of the Mercado. Although the results of theexcavation have not yet been published, Ruppert haskindly supplied the map (map 21) to aid in locating thepottery finds. He informs me that he believes that theMercado was built later than this colonnade. With theexception of the walls enclosing Room D of the South-east Colonnade, all the interior walls of the complexas well as several or all of the benches seem to besecondary to the original column arrangement. Thusa large colonnade has been remodeled into small com-partments by means of rather crude masonry construc-tion during a secondary occupation.

Six pottery collections, totaling 829 sherds, weretabulated from the Southeast Colonnade, 648 of whichcame from Room B. 448 of the Room B collection wereof unslipped ware; nearly all of these were of hour-glass form incensarios, suggesting the use of thisroom as a shrine. Two recognizable fragments of fig-urine incensarios were tabulated in this collection.Fourteen rims of Late Mexican style unslipped jarsare present, as compared with thirty-three jar rimsof earlier type. In the slipped ware of this collection,Medium Slateware runs about 16 per cent, MediumRedware another seven per cent, X Fine Orange 4.5per cent; Coarse Slateware composes 40.5 per centof the sample, Coarse Redware 30 per cent. Thus theRoom B collection points to a Middle Mexican and

early Late Mexican occupation, with a minor admix-ture of Early Mexican and scarcely a trace of Late orpost-Mayapan admixture. The Room B collection isdemonstrably later than the general run of Mercadocollections, as well as later than that of the Sweat Bathto the northeast of it.

The three collections from the Colonnade to the eastof columns 1-10 total 128 sherds; since the collectionsshow similar wares and percentages they will be dis-cussed as one. Sixty-one sherds of these collectionsare unslipped. Of this group, the jar rims are twelveLate Mexican substage form to two earlier. Incensariosinclude 32 pedestal bowl type to 10 figurine type. Thepredominating unslipped forms thus point to the Middleand Late Mexican substages. The slipped wares sug-gest a slightly earlier dating. Medium Slateware andMedium Redware together equal 47 per cent of totalslipped ware, Coarse Slateware 32 per cent and CoarseRedware 17 per cent. A small collection of 41 sherdsfrom Room C fits easily into this picture.A plausible reconstruction of the history of this

area is as follows: Initial construction of the SoutheastColonnade began in the fully developed Early Mexicansubstage. Occupation during that substage was followedby compartmentation of the large hall in Middle Mexi-can times. The occupation declined during the LateMexican substage. During the latter part of this lengthytime period (the whole Mayapan occupation span), char-acterized here by very light ceramic deposition, therewas probably no erection of new stone architecture.This area of Chiche'n Itza had descended from a cul-tural center to a sort of campsite. Only widespreadceramic sampling will show whether the rest of Chi-chen Itza was also unoccupied during the Late Mexicansubstage. The evidence at hand suggests that none ofthe large, well-made buildings thus far sampled waserected later than the Early Mexican substage.

Temple of the Three Lintels (7B3).-This templewas cleared and restored by Paul Martin during 1927and 1928 (Carnegie Institution of Washington Year Book26, pp. 233-236; Year Book 27, pp. 289 and 302-305).It has been described as the only known example atChiche'n Itza' of a pure Maya structure in the construc-tion style of the Puuc ruins using the monolithic-coredveneer masonry characteristic of that area. The build-ing showed no evidence of extensive remodeling; somestone had been robbed for small, crudely built habita-tions in the immediate vicinity. The indications thusseem to point toward an intensive occupation of rathershort duration contemporaneous with the temple, fol-lowed by a later occupation. Chichen Itza is rare inearly short-period occupation collections, and thisarea should thus be particularly worthy of study.

Two catalogued specimens are known by photographsonly. One (fig. 70, b) is an unslipped Early Mexicanincensario, the other the lower half of what appearsto be a Medium Slateware jar. The dark color suggestsa Florescent stage date but this, and even the wareclassification, is not certain.A sherd collection of 20 fragments was labeled as

from this temple and has been tabulated by me. It isprobable, though uncertain, that this is a part of thecollection excavated and tabulated by Roberts in 1933.Roberts dug two trenches, and obtained 578 rim sherds,from beside a well in the depressed area to the northof the temple. Martin (1927, p. 305) believed this wellto be contemporaneous with the temple. The masonrywell casing was flush with the ground; Roberts' exca-vations yielded what he believed to be an occupationaldeposit, built up from 1.27 meters below the rim ofthe well to the present level. The well originally hada high curb or else the casing was built higher during

38


time of use, to judge from Roberts' notes and drawings.Roberts' pottery tabulations show definite stratig-

raphy of a very interesting sort. His "lacquer wares,"which we can generally equate with our Coarse Red-ware at Chiche'n Itza, occur concentrated at the bottomof the trenches, and this apparent discrepancy is mark-edly contradicted by other internal evidences of a de-finite direct stratigraphy in the trenches. A check ofRoberts' shape analyses of these lacquer ware sherdsshows that they fit into the Yaxuna III and other Region-al stage series rather than into the Late Mexican sub-stage.

It seems safe to assume, even without opportunityto examine these sherds, that they must date from theearlier periods. This ware constitutes 7 per cent ofstratum C in the combined collection of the two trenches(185 sherds), 2 per cent of stratum B (201 sherds),and is absent in stratum A (215 sherds). Slatewarepercentages run constantly near 80 per cent but showsignificant variation in vessel form frequencies. Fineorangeware occurs in cuts A and B. No certain evidenceof types later than Early Mexican occur in the tabula-tions, and certainly nothing later than Middle Mexicanwas found (Roberts' tabulations do not distinguish thewares of these two subphases) nor the types of fineorange.

The changes in percentages of slateware vesselforms and in rim types in this deposit is of particularinterest for documenting the Florescent and EarlyMexican periods, since the majority of such materialfrom Chichen Itza' occurs in deposits so heterogeneousas to make chronological analysis difficult.

The percentages of Medium Slateware vessel formschange in the strata as follows:

ABC

Total slate

Basins14.422.5

28.3181

Bowls Jar s

3.3 82.48.3 69.3

15.2 56.6169 145

Jars thus increase in frequency at the expense of both

bowls and basins. The rim forms were tabulated byRoberts in reference to a standard set of drawingswhich he constructed. A check of these shapes shows

that in the basins, the group in which the demarkation

is most evident, cut A (the highest) contains no Puuc

forms; cut B, 68.4 per cent definitive Puuc; and cut

C, 87.7 per cent definitive Puuc. Cut B contains three

sherds, 7.9 per cent, of a rim which is not found in the

Puuc range of material but is common in Mexican Chi-

che'n deposits.The slateware bowls tell a similar story. A Thin

Slate bowl rim constitutes 4 sherds of 14 in cut B and

14 out of 22 sherds in cut C, but is absent in A. A sin-

gle sherd which probably came from a grater bowl oc-

curs in B. The slateware jar rims show little resem-

blance to Puuc forms, but there is evidence at Yaxuna

of regional variation from Puuc forms in the Flores-

cent period (fig. 10, a, b). Thirteen slateware jar rims

out of 82 in cut C are of a type very close to fig. 10,b, 29, and this type is not found in cuts A and B.

The redware listed by Roberts corresponds to our

Medium Redware and is distinguished by him from his

"lacquer ware, " a term coined by Vaillant who defines

it as follows:

The basal paste of Red Lacquer Ware is gray,coarsely kneaded, flaky, and porous. Over this

paste is laid a slip of clay more finely treated

with th\e grains adhering closely. The slip is thin,carefully smoothed, and burnished. So great a con-

trast exists between the basal clay and the slip infinish and texture, that the appearance of a lacqueris given" (Vaillant, 1927, p. 28).

Vaillant's description fits Yaxuna II Dichrome andthe Flaky Redwares very neatly, probably covers Ya-xuna III mottled Red on Coarse Paste ware, Red on

Thin Gray particularly of the Dzibilchaltun type, andVaillant used the term to cover Late Mexican CoarseRedware with which he was familiar. 'Lacquer ware"seems an unfortunate term, implying as it does thepostfiring application of a vegetal product, and it hasnot been retained in this report.

Two of the three shapes listed by Roberts in lacquerwares are included in the Yaxuna III Mottled Redwaresamples; his 'Orange Lacquer" dish rim approximatesfig. 8, g, 10, his Brown Lacquer jar form is near fig.10, d, 11, 12. The third shape of 'Red Lacquer" jar isnot found at Yaxuna but is a major form in Dzibilchal-tun Red on Thin Gray, and in the corresponding ware

at Acanceh, fig. 20, d, 6-_1 ;, 1 25-28. The colorranges of the wares described above, as well as therim forms, intermesh in such a way as to confirmthese assignments.

Of Roberts' three listed redware rim shapes, twoare of jars which are an extremely rare form in Me-dium Redware of the Puuc (Florescent stage) sites(see fig. 52, a, )j, and Roberts' listed jar rim formsare of kinds which do not occur in the Puuc collections.However, they are not markedly different either fromYaxuna IV (Florescent stage) Medium Slateware formssuch as those in fig. 10, b, 14-16, ,18 25, nor from Ya-xuna III (late Regional stage) Mottled Redware formssuch as fig. 8, d, 12 and may with reasonable cer-tainty be assigned to one of these two periods. Roberts'third shape, a redware bowl, can likewise be fittedeither to the Yaxuna III Mottled Redware series (fig.8, IL 1L, 1L k,k 8, 10) or to the Florescent MediumRedware series (fig. 52, m_,1 30 33.Since there are strong indications on the basis of bothware and form characteristics that Florescent MediumRedware and Slateware are later developments fittingthe same general ware series as Yaxuna III and Mot-tled Redware, it is not surprising that these few frag-ments cannot be placed more closely.

Four fine orange sherds occur in cut A, 5 in cut B.Since cut B seems to consist almost wholly of Flores-cent material, the 5 fragments would be expected tobe of Z type. This is not the case; three of them are

jars, a vessel form nearly absent in the Puuc FineOrange samples (fig. 59, e, 1-10), and the rim formlisted by Roberts does not appear in this series butis nearly identical to a conmon X Fine Orange form(fig. 79, ; fig. 77, jgj, ij; fig. 76, b). The bowl rimform also seems distinctly of X Fine Orange type,with strong curve and blunt lip like fig. 73, a, c, 1.All the fine orange from this collection must thus beconsidered of X type.

The classification of the deposit runs as follows:Cuts B and C nearly pure Florescent with traces, es-

pecially in C, of early Florescent and late Regional.Cut A consists of Early and possibly Middle Mexicanceramics, with very little mixture of earlier materialand no Late Mexican ceramics. The five X Fine Orangesherds in cut B suggest that this type appeared some-what earlier than the "Mexican Style" Medium Slate-ware basin rims. However, these later-style basinrims seem to appear somewhat later than the new bowlforms characterizing the Early Mexican period. Thelisted presence of four cascabel bowl legs in cut Cand of five in cut B suggests that these Medium Slate-ware bowls may date later than do the Yaxuna and

39


most other Chichen Florescent collections which uni-formly show slab legs. Pottery from the large Puucsites shows presence of cascabel legs, and the simi-larity in masonry construction between the Temple ofthe Three Lintels and the Puuc buildings might betaken to support the ceramic evidence for placing theoccupation of this group later than most of the Chich6nItza Florescent occupation. On the other hand, the pres-ence of X Fine Orange in these lower strata allows theplacing of these slateware bowls in the Early Mexicansubstage. These assignments are a problem of defini-tion. The virtual absence of X Fine Orange in the Puuccollections which are taken as the type for the Flores-cent stage would suggest the placing of cut B perhapsin "Florescent-Mexican transition," a convenient termfor a complex situation.

If cuts C and B date in major part from the occupa-tion of the Temple of the Three Lintels, as seemslikely, then cut A may date with the small secondarystructure on the temple terrace described by Martin(1927, p. 305), for the masonry of which the Temple ofthe Three Lintels was robbed. The robbing of stonesuggests a time lapse between the construction of thetwo buildings. Roberts suggests in his field notes thatthe interface between cuts B and C looks like a livingsurface and may be contemporaneous with the temple.It is unfortunate that we do not have data to allow thedefinition of Early from Middle Mexican deposits inRoberts' notes, as it is quite possible that cut A, andMartin's secondary structure, may be of Middle Mexi-can date. An important lead is suggested from the un-

certain but provocative evidence that the Florescent-Early Mexican transition at Chichen Itza was not asudden one.

Northeast Colonnade (3E1.-A sherd collection, tworestorable pottery vessels, and a pottery pestle re-

sulted from the clearing of the Northeast Colonnade(see Ruppert, 1952, fig. 40 for map). The sherd colleo-tion, totaling 257 pieces, was tabulated by me.

The Medium Redware jar, cat. no. 618, resemblesfig. 85, d,P1, save that the neck-shoulder junction ismore angular. The Coarse Redware bowl approximatesfig. 26, a, 9, but bears no grater pattern and no visibleslip.

The sherd collection is rich (55.2 per cent) in ped-estal-bowl incensario fragments. The remainder ofthe sample consists of Medium Slateware (73.1 percent of slipped wares), Coarse Slateware (15.4 percent), and Coarse Redware (11.5 per cent). Analysisof the Medium Slateware shows that out of 50 diagnos-tic sherds, only four basin sherds are of Puuc form.Two Coarse Redware bowl sherds bear incised designsthrough white slip band (cf. fig. 95, a). This style,stemming directly or indirectly through local wares

from X Fine Orange, does not appear in the Mayapansamples. It probably dates very early in the Late Mex-ican period, or may be contemporaneous with the latepart of Middle Mexican. Only three fragments of figu-rine incensarios were found; they probably postdatethe remainder of the collection.

Thus the major occupation of the Northeast Colon-nade was in Early Mexican time s, and probably in thelatter part of that substage. Occupational remains thendwindle to abandonment in Late Mexican times, prob-ably in the early part of the substage.

Temple adioinine the Northeast Colonnade to thesouth (3D6).-A small sherd collection from this area

(18 sherds) was tabulated by Roberts. All fragmentsseem to have been of Early or Middle Mexican type,with three fragments definitely limited to Early Mexi-can. One fragment of X Fine Orange is included in thecolle ction.

Temple on the northeast bank of the Xtoloc Cenote(3D13).-This temple, excavated by Ann Morris in 1925,yielded a collection of 13 rim sherds, known only fromRoberts' tabulations, and a sherd of plumbate ware

which was found on the floor of the middle room. NoFlorescent or Late Mexican wares are represented.An unmixed Early Mexican deposit is most likely,though the sample is so small as to provide only veryshaky grounds for placement.

Zumpulche (3E3).-This sweat bath (see Ruppert,1952, pp. 82-83, figs. 50, 51) was excavated under thedirection of Karl Ruppert during 1936. In ground planit is very similar to the T house, south of the Caracol(see map 20), which has been placed in the Toltec pe-riod by architectural studies, and in the Early Mexicansubstage by ceramic studies.

No whole or restorable ceramics were catalogued.I have tabulated three sherd collections, totaling 290fragments; one collection of 231 sherds is general fromthe excavations; one of 25 sherds is from the innerroom; and one of 34 sherds is from the water drain.No significant difference in ware percentages showsin the three collections, so they will all be treated as

one.The ware counts and percentages on a basis of total

slipped wares run as follows:

SubstageEarly MexicanEarly MexicanMiddle MexicanLate Mexican

Medium SlatewareMedium Redwar e

Coarse SlatewareCoarse Redware

X Fine OrangePlumbateRare types

Per centage32.111.432.610.3

12.01.1.5

100.0

Unslipped ware, including three Late Mexican typejars and no figurine incensarios, totals 17.5 per centof the total collection.

Thus it will be seen that the Early Mexican occupa-tion including X Fine Orange and plumbate, accountsfor over 50 per cent of the sample. Another third ofthe sample comes from a Middle Mexican occupation;the remainder is Late Mexican but probably does notinclude the latest part of that subphase, which is char-acterized at Mayapan by increasing quantities of figu-rine incensarios.

Analysis of the Medium Slateware vessel formsshows only five Puuc-style rims, all basins, out of49 significant sherds. No grater bowl fragments were

tabulated, but this is possibly a clerical error. Thesuggestion is that the occupational debris of this build-ing began during the well-developed Early Mexicansubstage.

Temple of the Hieronlvphic Jambs (6E3).-Thistemple, located in the southeast part of Chichen Itza,is discussed and a plan given by Ruppert (1943, p.233). It is one of eleven at Chich6n Itza having a groundplan similar to that of the Mercado.

Roberts dug a pottery trench in 1933, abutting theeast side of the terrace bearing this temple at the jogin the east terrace well. The collection was lost andthus not tabulated by me. Roberts notes two stratabut sees no difference in the collections from the two.He tabulates 51 rims, 25 of them slateware (Coarsenot distinguished from Medium), 7 sherds of X FineOrange, one plumbate, and the remainder unslippedjars. Three slateware basin sherds out of thirteenare of Puuc form. This, plus the presence of plum-bate and the large percentage of X Fine Orange, sug-

40


gests an Early Mexican dating, although the evidenceis incomplete because of the lack of paste classifica-tion in slateware.

Temple of the Interior Atlantean Columns (3C6) andHouse of the Grinding Stones (3C5).-This small groupstands on the south edge of the platform which bearsas its largest structure the Osario or High Priest'sGrave. It was cleared in 1927 by Paul H. Martin. Amanuscript report of this work is in the Carnegie In-stitution files. In this report Martin states his opinionthat these buildings are probably contemporaneouswith the High Priest's Grave. Plans and a descriptionof the two buildings are given by Ruppert (1953, pp.38-39, fig. 27).

To judge from the fallen stonework, the Temple ofthe Interior Atlantean Columns bore a vaulted roof,and the masonry contains many re-used stones. Thepottery from this building was found in the above-floordebris and bore at least two of the whole vessels whichwe shall discuss, although the catalogue does not spec-ify exact locations for individual pieces. The House ofthe Grinding Stones seems to have borne a flat tim-ber-beamed roof, and sherds were numerous at thefloor level.

Neither whole pieces of pottery nor sherd collec-tions remained available at the beginning of this study.Catalogue information and photos, and Roberts' tabula-tions are the only sources of information. Two vesselsof X Fine Orange (fig. 77, Z& and ji) and two designbearing sherds (fig. 79, m and 80, cc) were found. Apottery pestle (fig. 72, g, 4) and a small, hollow, un-slipped animal head, seemingly from the end of an in-cense blower handle, were found. A Medium Redwareflat-bottomed, sharp basal angle bowl with outturnedrim and three spherical cascabel feet came from thisgroup. This bowl has a dotted semicircle design inblack paint on the rim top, and a scroll in the centerof its floor (cf. fig. 87, a-t). Roberts' tabulation of the30 sherds suggests a typical Early Mexican samplewith no mixture of either Florescent or of the latersubstages. This placing is borne out by the whole ves-sels listed above.

Casa Redonda.-The published report on this struc-ture (Pollock, 1937) contains a description by H. B.Roberts of the ceramics from the building (ibid., pp.151-152). In the light of additional knowledge now avail-able, I shall discuss the placement of this building.

Roberts' tabulations and drawings of the Casa Re-donda pottery are available for the rims of what seemsto be a combined collection from the general clearingof the building and from the pit at floor level in thecenter of the building. 107 rims and other diagnosticsherds are included. Seventy-three incensario frag-ments are of the pedestal bowl type, some studded(see fig. 69); three are of the late figurine type. Nineunslipped jar rims are of the general Puuc-Early Mex-ican form (see fig. 68, a, b, 1-10). The slateware in-cludes eight basin rims of Early-Middle Mexican formand two of Puuc form (Medium and Coarse Slatewarenot distinguished). Bowls are all spheroid of Early toMiddle Mexican form; two of them are Red on Slate-ware, and thus can only be Early Mexican (see fig. 75).One of these sherds has been lost; the other seems tobe the ladle bowl sherd illustrated in fig. 75, e or onevery similar. Five redware bowl fragments must beEarly Mexican Medium Redware (late Mexican CoarseRedware was called red lacquer by Roberts). A black-ened, slipped, coarse-paste animal head leg (see fig.95, b) was also found, and this must be Coarse Red-ware.

This collection evidences a strong Early Mexicanoccupation, probably late in the period to judge from

the placement of Red on Slateware at Chichen Itza'. Thelack of plumbate, X Fine Orange, and Medium Slategrater bowls is not significant in such a small collec-tion (only 24 slipped ware rims). There is a smallscattering of Late Mexican material, three incensariosherds, and the slipped, coarse-paste animal head leg;this is not enough to constitute evidence of an occupa-tion.

Because of its architectural style and mud-mortaredblock masonry, it has been suggested that this buildingis one of the latest at Chichen Itza. If this stylisticarchitectural dating is sound, it supports the findingsgiven elsewhere in this report that architectural activ-ity ended not much later than the end of the Early Mex-ican substage.

Temple of the Wall Panels (3C16).-This temple,lying to the south of the Caracol, was excavated in1927 and has been reported upon by Karl Ruppert(1931). The resulting pottery is a collection which con-sisted of 172 sherds when tabulated, and of seven wholeor restorable pottery vessels.

The architecture of this temple is of the Toltec pe-riod, quite similar to that of the Temple of the War-riors, and to other structures of that period. Severalstages of reconstruction and addition are noted by Rup-pert, including considerable re-used stone; amongthese are drums of Atlantean columns, a form believedto date late in the period. The ceramics come fromthe clearing and reconstruction of the building, andprobably contain little if any material predating theoriginal structure.

The ceramic sample contains the following percent-ages of slipped wares: Medium Slateware 19 per cent,Medium Redware 7 per cent, Coarse Slateware 43 percent, Coarse Redware 23 per cent, X Fine Orange andplumbate 8 per cent. Thus the ceramic sample datesMiddle Mexican in major part, with smaller represen-tations of Early and Late Mexican wares.

Of the whole specimens recovered, only one is cer-tainly of Early Mexican times, the Medium Redwaretripod cascabel vase shown as fig. 86, d, which wasfound in a cyst on the lower terrace north of the latestairway (Ruppert, 1931, p. 137, plates 16, b, 17, c).This is also the only ceramic specimen which may beassociated with the construction of the building. Thelower terrace was the first element constructed, butthe cyst may have been added later, at any time up toabandonment. This vessel thus dates the structure notlater than the Early Mexican substage. The other sixwhole specimens are an incense ladle or "blower" ofunslipped coarse gray found in debris within the colon-nade, shown as fig. 68, f. 8 (Ruppert, 1931, plate 17,c); two Coarse Red pedestal vases (see fig. 95, h forone of these), found unbroken in debris in the southpassageway of the temple between the bench and westwall of the inner room of the temple; and three broken,thin, coarse unslipped ware direct rim jars (see fig.93, a, b, for two of these), found in the talus of the tem-ple pyramid.

Five of these six specimens may be identified withreasonable safety as Late Mexican from counterpartsfound in the Late Mexican deposits at Mayapan; theincense ladle probably dates earlier. The two pedestalvases are similar to specimens found cached besidea column in the North Colonnade, Temple of the War-riors (fig. 90, s, tj. These pedestal vessels, as wellas those from the Warriors, seem most likely to havebeen cached after the collapse of the buildings in whichthey were found, which thus must have occurred duringor before the Late Mexican period. The unbroken in-cense blower may well have been left under the sameconditions; if so, the fall of the building would be dated

41


as Middle Mexican. The unslipped jars are also likelyto be postoccupational.

The combination of the heavy proportion of CoarseSlateware, nearly absent in the collections of the Cara-col and from all but the East Building of the Monjas,and a Medium Redware jar in the subfloor cyst of theterrace, place at least the early part of the construc-tion during the Early Mexican period, the major occu-pation in Middle Mexican, and the collapse of Templeand Colonnade before the end of the Late Mexican pe-riod. The latter part of the Early Mexican period seemsindicated as the period of construction, explaining thesmall proportion of Medium Slateware sherds andtallying with the re-use of Atlantean column drums inthe colonnade. It would be of interest to know if any ofthe well-made veneer masonry which distinguishes thisbuilding dates from the Middle Mexican period, butdata on this is not obtainable from the pottery sample.

Initial Series Group (5C).-The Initial Series Groupwas investigated in 1926 by Vaillant (Vaillant, 1933;Ruppert, 1952, p. 157), who established the fact thatthe standing structures of the group (map 22) were be-gun with the construction of the main range of the Phal-li, which he believed to be early Mexican showing Mayaconstruction methods but Mexican style ornament.Vaillant (1933) distinguished three periods of architeoctural construction following this, the first of which,consisting of the colonnade south of the Phalli, heequates on stylistic architectural grounds with the Cas-tillo, the Chacmool Temple, and the Great Ball Court.The next period showed less ornamentation. It includedthe south patio of the Phalli and its bordering structuresto the south and east, which include Atlantean columns.Vaillant detects signs of plundering at the close of thisperiod in reconstructable ceramic fragments, whichinclude two X Fine Orange bowls (fig. 81, a and .),widely scattered over the South Court. Plumbate sherdswere also found in this debris. He suggests that thisevidence may document the resurgence of the Mayaafter the earlier Toltec conquest. Vaillant posits afourth and final Maya Resurgence period, including thecrude Temple of the Initial Series (into which the cer-

tainly earlier Initial Series lintel has been incorporatedin modern times). Vaillant also includes in this periodthe figurine incensario fragments which were foundscattered at or shortly under surface level.

I observed no significant ceramic differences in the18 small collections which, in addition to Vaillant'scollections, included the material from a trench dugby Roberts in 1933. The collections bear Medium Slate-ware as the preponderant ware. This ware character-izes the Flore scent and Early Mexican periods. A formanalysis done to separate Puuc-style vessels fromlater ware shows a high proportion of Florescent pot-tery. At the same time, collections made by Vaillantbetween successive floors of rooms of the earliest pe-riod in the group (north range of the Phalli) show sherdsof the Coarse Slateware of the Middle Mexican sub-phase. Probably these buildings continued to be usedduring a long period, but the collections have possiblybecome mixed since excavation. The ceramic evidenceplaces the group in the Florescent and Early Mexicanperiods, perhaps bridging the transition as Vaillantbelieved. The Initial Series lintel may not have beencarried far to its present secondary position since itdates in the Florescent stage by our scheme, as doesPuuc-style Medium Slateware. The absence of pureFlorescent collections beneath architectural construc-tions leaves unproved the possibility that some of theearliest construction dates from the Florescent stage.

Faint traces of the Coarse Slateware of the MiddleMexican period show in the collections, and slightly

larger amounts of the Coarse Redware of the Late Mex-ican substage. Ten fragments of figurine incensariosare listed from the 1,676 sherds examined from thebuilding group; some of these seemingly lay underfallen walls. Thus the late Mexican ceramics are notsufficient to argue an occupation of any size, but dosuggest that parts of the group were standing duringthat period. This is to be expected; some of the wallsstill stand.

Monias (4C1. 3. 4).-The Monjas group at Chiche'nItza' shows more stratigraphic complexity than anyother building group excavated, and should producecorrespondingly more information. The excavationswere conducted during the field seasons of 1932, 1933,1934, under the direction of J. S. Bolles.A total of 158 collections was taken during the ex-

cavation, and these contained, at the time of my tabula-tion, about 10,000 rims. The labeling of the collectionswas done descriptively, the terms used referring tothe parts of the building. Unfortunately, since the re-port of these excavations has not yet been published,the architectural stratigraphy of the building is notentirely clear to me, although I have had access toBolles' manuscript and many of his maps. The includedplans (map 23) have been compiled from photostats ofa series of partial plane table maps drawn during theexcavation, and must certainly contain inaccuraciesfor which I am responsible. It seemed better to com-

pile plans for locating the collections as far as possi-ble than to treat them blindly. I have not been able tolocate the field catalogue listing the provenience of thewhole pieces of pottery recovered, and most of thewhole vessels thus cannot be placed in the architecturalsequence.

The Monjas building proper as well as the SoutheastAnnex and the Iglesia have been generally classifiedas pre-Toltec (Morley, 1931, p. 107). The type of con-

struction and decoration shows general similaritiesto Puuc and Chenes buildings. Architectural placementsgiven here are provisional, pending the publication ofBolles' report. From map 23 it will be evident that theeast wing antedates platforms 3 and 4. Rooms 14 and15 are in Florescent architectural style and of course

postdate platform 4. The Ball Court is consideredToltec on the basis of the style of its bas-relief. Thenortheast annex, the east building, and the rooms facingthe southeast court on its south and east sides are inToltec-period architectural style and must be additionsmade to the group during that time. There is ceramicevidence, described below, of occupation of the eastbuilding and of the complex of rooms to the north ofit in the Middle Mexican substage. In general, thereis little pottery from the Monjas excavations datinglater than the Early Mexican substage, and little ear-

lier than that of the Florescent stage. Medium Slate-ware is the major slipped ware of all deposits, andthere is considerably less evidence of the later CoarseSlateware and Coarse Redware than in the Caracol col-lections. Coarse Slate and Redwares are largely con-

fined to surface collections in this series and, as re-marked above, constitute only minor parts of all sam-

ples. Since this area of Chiche'n Itza shows evidence ofmuch visitation of late peoples, it seems certain thatthe late wares are to be thus accounted for. As furthersupport for this interpretation, several sizeable col-lections composed almost entirely of figurine incen-sario fragments were found on the top of the pyramidand on its slopes.

On the basis of the ceramic changes from Flores-cent to Mexican, all Monjas pottery collections largeenough to be statistically significant as samples con-

tain some pottery definitely belonging to the Early

42


Mexican substage. The most surprising of these collec-tions includes a six-sherd sample (M28) from underthe floor of the Ball Court which, according to Bolles'field notes, definitely antedates the erection of platform.4. This in turn antedates rooms 14 and 15, classifiedas Florescent. Of this collection, four sherds are Flo-rescent, two definitely Early Mexican. Collection MHR51 is listed as coming from "in front of Northeast cor-ner platform from beneath Iglesia." It contained 81sherds, preponderantly Florescent but with severalEarly Mexican sherds, and one figurine incensariofragment.

A large collection (2,012 sherds) from the north endof the Ball Court, adjacent to and below the seat levelof the bench along the south side of platform 4, is over-whelmingly Early Mexican. It shows an unusually largepercentage of X Fine Orange, and is one of our besttype collections for the substage. Bolles' opinion, asgiven in his field notes, is that this pottery was refusefrom the buildings to the east of the Ball Court. It mostcertainly postdates the use of the Ball Court as wellas its construction, which in turn must have followedthe building of platform 4 of the Main Pyramid.

A collection of five sherds whichmay antedate plat-form 1 and certainly is not later than platform 2, showssherds of waxy orange slipped ware similar to certainsherds of the Coba Group B period (late Regional). Thisis faint evidence, but suggests that construction at theMonjas group may have been started at this period,which is strongly represented at Yaxuna (Yaxuna III),not far from Chiche'n Itza' to the south. It is significant,however, that few signs of this period occur in thelarge collections from elsewhere at Chichen Itza', whichobviously was not a site of importance at that time. Thewhole pieces from the Monjas which we have seen camemainly from caches and burials. They are not treatedhere since information on their characteristics andprominence is lacking. As far as we can judge, theyare all of the Early Mexican substage, save for thelarge Coarse Slateware jar (fig. 92, A) which was foundin the East building, and which must date Middle Mexi-can.

The lack of pure Florescent-stage collections mayseem strange in the excavation of buildings which havebeen placed in that period on architectural grounds,and it brings up the question as to the validity of (a)the architectural criteria, (b) the ceramic criteria,and (c) the method of collecting the ceramics.

The most definitive architectural criteria consistof judgment as to style of masonry and design on theelaborately ornamented east wing. H. E. D. Pollocktells me that in his opinion this wing seems to copythe style of Chenes buildings. The Chenes sites, onarchitectural and ceramic evidence as yet unpublishedin detail, probably either approximate the Puuc sitesin dating, or slightly precede them with a decided over-lap in time (Pollock, 1936, pp. 123-124). The Chenesruins are associated exclusively with Florescent-stageceramics (Brainerd, 1949). The Toltec-style buildingsstratigraphically follow those of the Florescent stagein this group. The architectural criteria of these pe-riods, as shown in the buildings of the Monjas group,seem well marked and consistent, and are agreed uponby the specialists in the field, although the final defini-tive architectural description awaits publication ofBolles' monograph.

The ceramic criteria are distinct and well estab-lished for the. Early Mexican substage, one of the twomost fully documented periods on the Peninsula; thePuuc area occupation of the Florescent stage is theother well documented period. The Florescent stageof the site of Chiche'n Itza is much less well known ce-

ramically since few unmixed deposits of that periodwere recovered, doubtless due to the later disturbancesof strata and the deposition caused by the heavy EarlyMexican occupation there. However, there are otherFlorescent collections from sites near Chichen Itza'.

Our sample of the Florescent stage at Yaxuna isquite small but is probably classifiable as early Flo-rescent, as are the collection of whole vessels fromDzebtun to the north (fig. 35), and the Mani Florescentpottery to the west. The Florescent occupation of Coba'is represented so poorly in our collections as to offerlittle clue. However, these collections are sufficient toestablish the fact that the Florescent stage, as definedin western Yucata'n, also existed in the area of Chiche'nItza. The dating of the probable later horizon of theFlorescent stage shown by the major part of the Puucsite collections in relationship to the Early Mexicansubstage at Chich6n Itza is discussed later in this sec-tion. I believe that they constitute a chronological se-

quence with little if any overlap between them.If the assumption of the validity and contemporane-

ity of the architectural and ceramic periods are ac-cepted, the apparent disparity in the finding of onlymixed Early Mexican-Florescent collections in the ex-cavation of a building complex containing Florescentbuildings must be due to the collection of the ceramicdata (point c above).

Several factors combine to make ceramic samplingdifficult at Chichen Itza'. Shallowness of soil and theuse of clean stone fills make rubbish deposits rare,and the reuse of midden earth is inevitable in such a

heavily occupied site as this. The primary purpose ofthe archaeological work, as contracted for with theMexican government, was the restoration of ruins, andthis aim was followed conscientiously to the detrimentof exploratory ceramic trenching within building groups.Thus the ceramics come largely from the occupationof, or after the abandonment of, the latest constructionsin the architectural group. The Early Mexican collec-tions contain strikingly varying frequencies of Flores-cent pottery, but this variation is.usually independentof the chronological position of each collection in theEarly Mexican substage. It seems rather to be a func-tion of the amount of Florescent occupation whichunderlies the spot where the collection was found.There are some suggestions of transitional forms anddecorations between Florescent and Early Mexican ce-

ramics at Chiche'n Itza, but most of the collectionsseem to be mixed rather than transitional. Middle Mex-ican and Late Mexican sherds are often limited to de-posits which overlay fallen roofs, and thus are demon-strably late. The Monjas collections mentioned above(M28 and MHR51), which would be expected to be Flo-rescent, are hard to account for, but they may be con-

taminated either by downward movement of late ma-

terials through the loose stone fill or by faulty handlingof ceramics at or after excavation.

The chronological placing of the Florescent-stagepottery of the Monjas relative to the occupation of thePuuc sites is a matter of considerable interest, andbears directly on the correlation of the Maya andChristian calendars. The similarity of the Monjas Flo-rescent ceramics to early Florescent pottery ratherthan to the later material from the Puuc sites hasbeen mentioned. Since Z Fine Orange pottery occurs

in Florescent and not in Early Mexican deposits, andthe reverse is true of X Fine Orange, the Florescentand Early Mexican deposits must be sequent ratherthan chronologically concurrent (Brainerd, 1941). Lackof evidence of trade from Early Mexican Chichen Itzain the late Florescent Puuc deposits also points strong-ly toward a sequent placing of late Florescent and Early

43


Mexican, as does the occurrence of deposits at Chiche'nItza which are completely free of Florescent potteryin areas where no Florescent architectural evidencehas been found. The hiatus between Thompson's Chi-che6n Itza dates and the traditional date for the arrivalof the Toltecs (Thompson, 1937, p. 190) may be usedquite neatly to cover the hiatus between early Flores-cent and Early Mexican stages in the Monjas materials.This hiatus in dates (10.3.0.0.0 to 10.8.0.0.0) occursat about the right time if we place the span of our ear-ly Florescent ceramics from 9.14.0.0.0 to perhaps10.3.0.0.0. Using Thompson's 10.3.0.0.0 as an inter-mediate point, we may consider 10.8.0.0.0 as the endof the Florescent stage and the beginning of Early Mex-ican, as Thompson has suggested (1941, Scheme B).This allows about 180 years for the early Classic ce-ramics and 100 years for the fully evolved Puuc ceram-ic assemblage, during which span the Monjas and prob-ably the rest of Chiche6n Itza showed only a light occu-pation, demonstrated ceramically by a few scatteredsherds of the more elaborate, later Puuc types. Analternative hypothesis, favored by Proskouriakoff (1950,pp. 169-172) on the basis of Chiche'n Itza' dates and thestyle of stone sculpture, and by Lothrop (1952, pp.111-113) on the basis of metal-work style, is to over-lap Toltec Chiche6n with the Puuc occupation. This re-construction has support from certain documentaryinterpretations. Although the ceramic evidence can beinterpreted to allow some overlapping for these occu-pations, I believe that the evidence seems to favor aminimum of such overlap.

None of our collections, save those from the Puucsites, shows sizeable deposits of the late and moreelaborate Puuc wares. This may indicate, as we havesuggested elsewhere, that there was a major occupa-tional shift to the Puuc region during this period, orour sampling may simply by chance have missed thisperiod in other parts of the country. A less likely pos-sibility, to our mind, but one that should in fairness bementioned, is that the rather elaborate pottery of thelate Florescent stage in the Puuc area was an extreme-ly localized phenomenon, and that the early Florescentceramics from the Monjas and elsewhere, which wehave assumed to date early, are actually chronological-ly late forms showing peripheral lag or regional varia-tion. This hypothesis seems to me somewhat farfetchedin view of the nearness of sites, the extensive popula-tion indicated, and the similarity of pottery types atvarious other time horizons over the whole north endof the Peninsula.

Evidences of Middle and Late Mexican occupationare slight relative to Early Mexican at the Monjas;their strongest representation is in the East Building,where collections from the building itself show 22.4per cent Coarse Slateware and 6.9 per cent CoarseRedware, compared with 60.7 per cent Medium Slate-ware and 10 per cent Medium Redware out of 290sherds of these four slipped wares. The Coarse Slate-ware jar found only slightly broken in this buildingsuggests that the last occupation was during the MiddleMexican period and that the Coarse Redware, whichconsisted mainly of small bowl fragments, should beassigned to the late pilgrimages which have left a sur-face ceramic layer over most of Chichen Itza.

Sacred Cenote.-The dating of the ceramics fromthe Sacred Cenote at Chichen Itza' is of particular in-terest because of the richness and variety of objectsfound in the deposit there, and because of the documen-tary evidence concerning it. The ceramics alone willbe treated here and this study is far from exhaustivefor that material, since I shall not attempt to discussor identify the imported ceramics found. The other

artifacts from this deposit have been studied elsewhere,and are in process of publication by the Peabody Mu-seum, Harvard University (see Lothrop, 1952).

The cenote was dredged by E. H. Thompson between1904 and 1907. A winch and boom from which was sus-pended a steel clam-shell bucket were used in the op-eration; the material from the deposit thus was not re-covered or recorded in any stratigraphic order. Theceramic data available to me come from two sources,tabulations made by G. C. Vaillant during 1926 of sherdcollections which remained at Chiche'n Itza' from Thomp-son's dredging operations, and examination of ceramicscoming from the same operation kindly made availableby the Peabody Museum. These two sets of data willbe described in turn, and their significance discussed.

The collections studied by Vaillant have since beeneither lost or separated, and the labels lost. However,his ware distinctions are quite closely equatable withthose used in this report, save that, although Vaillantis known to have recorded Coarse Slateware for indi-vidual specimens, we cannot be sure he separated itfrom Medium Slateware in his tabulation. This uncer-tainty is minimized by the near certainty that CoarseSlateware did not occur in any quantity in the collec-tions Vaillant worked upon.39 Two collections were re-corded by Vaillant, number 1 only by presence and ab-sence, number 2 quantitatively by number of sherdsper category. The tabulation which follows allows acomparison between Vaillant's quantitative cenote tab-ulation and one of our least mixed Early Mexican de-posits.

Comparison between Chichen Itza Cenote collectionand a collection typical of the Early Mexican period.My equivalents are in parentheses.Porous (unslipped) Cenote Trench South

No. 2 of Southeast

Jars 9.5Plain 4.3Striated 5.2

Cer emonial(blue-painted bowls) 12. 1

Sub-lacquer, transitional(no equivalent) 1.5

Late incensario(figurine incensario) 2.4

Early incensario(pedestal incensario) 0.0

Lacquer (Coarse Redware) 13.6Slatewar e

(Medium Slateware,) 53.8Jar s 40.0Cajetes (basins) 11.1Bowls 2.2Graters (grater bowls) 0.6

Polished red(Medium Redware) 3.7

Polished orange(early variety ofMedium Redware?) 1.2

Fine slatewar e

(Florescent ThinSlateware) 1.5

Fine orange (X Fine Orange) 0.0Total sherds in collection 322.

Colonnade1.3

0.0

8.66.6

13.25.2

0.0

34.20.0

34.l

24.2

0.02.23

1,254.

Several clear-cut facts emerge.1. The cenote sherd deposit consists largely of pot-

tery dating from the Early Mexican and Florescentstages. These two horizons are those of the major oc-

cupation of Chich6n Itza. Both are characterized by

44


Medium Slateware. Vaillant's Fine Slate and PolishedOrange (if our identification of Vaillant's descriptionis correct) occur only in Florescent stage deposits,and thus the presence of these wares in the collectionsuggests a distinct Florescent intermixture.

2. The Early Mexican and Florescent deposit de-scribed above is composed of utilitarian pottery, notceremonial. The preponderance of jars and basins incomparison to bowls and grater bowls suggests thatthe deposit was formed by the use of the cenote as awater source rather than as a ceremonial center.Strongly reinforcing this conclusion is the absence inthe cenote deposit of the so-called pedestal incensar-ios which characterize both Early Mexican and Flores-cent times, and which were used ceremonially.

3. Ceramics of the Late Mexican substage occur inthe cenote deposit mixed with the Early Mexican Flo-rescent deposit. The number of sherds definitely iden-tifiable as of this substage is slightly less than half thenumber identifiable of the early periods (28.1 per centcompared to 60.3 per cent). Of the identifiable LateMexican ceramics, a considerable proportion (all theblue-painted bowls and an indeterminate proportion ofthe lacquer ware) is believed to have been ceremonialin use.

4. Coarse Slateware, the preponderant pottery of theMiddle Mexican substage, of which the most definitivecollections come from Dzibilchaltun, is sparsely re-presented at both Chichen Itza' and Mayapan, and if Icorrectly interpret Vaillant's tabulations, is complete-ly unrepresented in the Cenote sample. The chief con-centrations of ceramics of this period at Chichen Itzaare in above-floor refuse in the Mercado and the adja-cent Southeast Colonnade, where architectural evidenceof decadent reuse of buildings abounds, and above floorsin the Temple of the Wall Panels and the East Buildingof the Monjas. It therefore seems evident that the Sa-cred Cenote at Chiche'n Itza was little used as a watersource by Middle Mex5can times, concurrent with ageneral depopulation of Chiche'n Itza, and that the prac-tice of throwing pottery ceremonial vessels into thecenote was not common until the Late Mexican substage.During this and later times, as detailed elsewhere,there is no evidence for the use of the site of Chiche'nItza save as a goal for pilgrimages.

The Sacred Cenote collection of whole and recon-structible vessels at the Peabody Museum, HarvardUniversity, adds additional evidence and detail to thestory. Figure 96 illustrates 16 coarse paste vesselsof this collection, and an additional vessel (fig. 95, n)is aberrant in slip color but probably of this category.Vessels fig. 96, b and jcontained balls of copal. Sevenof these vessels bear red slip (Coarse Redware) andthe remainder do not, but the paste of all seems thesame. Several are painted in blue, some with the addi-tion of other colors. These vessels correspond in wareand shape to other Late Mexican Coarse Redware fromthe Chichen Itza and Mayapan collections (cf. figs. 94,h; 26; 22, d; 27, A). A plaster-coated incensario (fig.97, a) corresponds to the fragments from Chichen Itzaillustrated with it. This type of incensario can probablybe attributed to the Late Mexican substage on basis ofits occurrence in quantity at Mayapan (fig. 23, d) andits absence from all collections containing Middle, butnot Late Mexican material.

Other Cenote vessels from the Peabody Museumcollection, some of them ceremonial in type, belongto other periods, mainly Early Mexican. Perforatedbody incense braziers without attached figurines (fig.97, h and j) do not appear in the Mayapan collectionsbut do appear in Early Mexican Chich6n (fig. 97, , i).The complete brazier from the Cenote collection (fig.97, h) bears what appears to be a specular hematiteslip on a foreign paste (quite possibly from the Mexi-can mainland), while fig. 97, i, must be a crude, local,unslipped copy. Two small Medium Slateware basins(fig. 73, b and f) fit as to ware with the Early Mexicanbasins illustrated in the same figure; b is probably aminiature and likely Florescent in date (cf. fig. 44, b,41, 43, from Uxmal). An incised Medium Redware jar(fig. 85, a) is certainly Early Mexican (cf. pottery onsame plate); the scratched Medium Redware jar sherd(fig. 88, h, 1) is closely similar to the accompanyingsherds from other Early Mexican Chiche'n Itza collec-tions. The medium paste unslipped fabric stamp fromthe Cenote (fig. 91, i) is placed in time by its closecorrespondence to fig. 91, g, which comes from trench1 south of the Southeast Colonnade, an Early Mexicandeposit.

Three other objects in the Peabody Museum colleo-tion from the Sacred Cenote seem late. The miniaturecup-shaped object (fig. 93, w) shows features like thoseon fig. 93, x, and several specimens in the Regil Collec-tion (fig. 100, L-D1. These show enough similarity tomodern Lacandon god pots (cf. fig. 100, Dj to suggestlateness, and we have encountered no such vessels inexcavated material. Another miniature incensarioprobably from the Cenote (fig. 93, v) likely belongs tothe same group; a vessel found on the ground surfacein a shrine near Coba (fig. 100, d) closely resemblesthis piece. The Cenote bird whistle (fig. 93, aa) resem-bles the others shown with it; bb and cc were made in1940 at Maxcanu.

These objects reinforce the conclusion drawn fromthe sherd collection that the Sacred Cenote at Chiche'nItza' was used as a depository for ceramic offerings,as opposed to its use for dumping and as a water sup-ply, chiefly during the Late Mexican period, which isprovisionally dated as from about 1200 to 1440 A.D.

A considerable body of literature is available atpresent on the Sacred Cenote at Chich6n Itza, and willsoon be augmented by further Peabody Museum reports.The best beginning reference is Tozzer, 1941, whichsummarizes and gives references to much of the infor-mation available. Bishop Landa describes the persis-tence of the ceremonial use of the Cenote in the mid-sixteenth century and the numbers of idols still thereat that time. Some of these idols very likely were figu-rine incensarios of the type illustrated in this report(for complete specimens, see figs. 99-102). The tradi-tion of making pilgrimages to abandoned temple sitesis documented for various Maya groups from Conquesttimes until the present. As with the modern Lacandones,these pilgrimages often included the depositing of in-censarios.40 Chichen Itza must have remained such asite from its decline at about 1200 A.D. until after Con-quest times-abandoned, ruinous, but revered by theMaya as a former glory of their people and as an abodeof their gods.

45

III. CERAMIC DESCRIPTIONSThis section contains the descriptive data upon which

the conclusions of the work are ultimately based. Anoutline of the procedure used in obtaining these data isincluded in the introduction. The descriptive materialis contained in two main sections, a description of thewares and a set of figures with captions illustrating theceramic form and decorative repertories. The figurecaptions are full; into them have been written discus-sions of relationships between shapes and decoration,and references to similarities with other pottery fromthis and other areas. The ware descriptions includeshort, generalized, comparative sections which sum-marize detailed and referenced material from the cap-tions. A third level of generalization will be found inthe sections on historical reconstruction which followthis.

The quantitative tabulations which were made as thepottery was recorded are not included in this monographbecause of their bulk. They are available for study atU.C.L.A. It is expected that their chief usefulness willbe realized when more detailed analyses of substagesare made. They have proved useful already in analysesof the history of shapes and in other detailed work.Their compilation, which required a major segment ofall working time, was made in an effort to preserve de-tailed data which would otherwise have been lost withthe necessary discarding of the collections. Forms andrim shapes have been recorded on the individual lotsto a total of from 30 to 200 categories. This degree ofdetail has not been utilized for the establishment of therelatively loose chronology used here. The full tabula-tions have proved too voluminous for inspectional study.Statistical work now under way on parts of the tabula-tions should allow the refinement of intraphase chro-nology and closer interregional time and distributionalstudies as work proceeds (see Brainerd, 1951). A gen-eral picture of the relative frequencies of the majorwares can be obtained by examination of charts 2 to20 and by reading the chapters on stratigraphy andchronology.

The ware descriptions used here are brief and gen-eralized to facilitate their usefulness as a key to themain sorting groups employed during the study. Minorvariants from the main range of material and rare in-trusive wares are not discussed in detail in this seo-tion, nor are detailed comparisons made to potteryfrom other areas. Data of this sort are included in thecaptions. It was found that paste texture was a usefulsorting criterion; it is defined here simply as the de-gree of irregularity of the fractured surfaces of thepottery as viewed without magnification. Four divisionswere defined and used in the naming of wares; wareswere called "Coarse," 'Medium," 'Thin," and "Fine."There is reason to believe that this variability is dueprincipally to the size of temper particles; fine waresare untempered. In general, wall thickness of vesselsvaried directly with texture; coarse textures werethicker, fine were thinner. See plate I for illustration.While the ranges of color,41 slip, and paste texturesare believed to be distinct and were found to allowready sorting, it should be remembered that the col-lections were sorted separately over a long time pe-riod and not referred to for comparison before finalediting. I believe that a proper study of the regionaland temporal variations certainly present within thecategories here set forth can be carried out intelligent-

[46]

ly only in close conjunction with technical studies. Suchstudies were foregone because of bulk of material andscarcity of time; sample sherd collections have beensaved in hope of later detailed work. This is a survey,and I can hope to do no more than formulate a frame-work for later, more detailed study. Miss Anna Shepardvery kindly examined several thousand fragments witha binocular microscope, and has sectioned samplesherds for more careful study. Her temper determina-tions are attributed when used; other temper determi-nations are mine.

In the sections on comparisons which accompanyeach ware description, an effort is made to describesuch intergradings and distinctions as were noted be-tween wares. The wares chosen in sorting have vary-ing degrees of definition from their accompanying andrelated materials. Each ware was sorted as a groupdefined by the presence of several describable attrib-utes characteristically found in combination on singlespecimens. The characteristics chosen as definitiveof wares were those in which tendencies toward occur-rence together were noted. The classification is thusbelieved not to be an artifact of the sorter; each warerepresents a product determined by cultural condition-ing in the choice of materials, in the techniques usedfor forming and decorating, and in the stylistic conven-tions of form and design. The fact that each of the de-finitive characteristics chosen has an independent dis-tribution of its own makes the drawing of lines betweenwares depend to some degree upon the judgment of thetypologist, even though their tendencies to occur incombination were a prime criterion in the typology.In the judgments made as to ware dividing lines, anattempt was made to use as the definitive characteris-tics those which (a) showed definite binodality ratherthan simple intergrading between the two wares inquestion, and (b) characteristics which might be judgedto have been intentionally varied by the potter ratherthan the result of imperfect control of craft techniques.An effort has been made to describe more closely theoccurrence of these characteristics in the several sec-tions on temper, slip, paint, and forms.

Since ware descriptions are at best simply a devicefor the easier handling of an unwieldy and highly vari-able mass of material, no inherent dignity should beassumed for those defined here, and no value otherthan the fact that they represent, in the author's bestjudgment, the trait clusterings which structure thematerial studied. They have been named for easierreference; use of these names and definitions shouldbe subject to constant scrutiny as to usefulness in anynew work.

The wares used in this report are more broadlydrawn than in most archaeological reports on singlesites or small regions. This is a result of the thinlyspread nature of the sampling and is mirrored in thematerials. For example, there is considerable consis-tent difference between the Regional Medium Redwar eof Cobg, Yaxuna, and Acanceh. These samples are herelumped under a single ware because the excavation oftwenty more sites would doubtless produce severalmore "subwares" of this period, all having roughlyequal variability. This situation must have been causedby many small groups of potters following the samegeneral traditions and styles in their work, but eachat the same time varying in materials available and


in minor cultural idiosyncrasies. Close ware descrip-tions such as those used in reports of single sites be-come cumbersome in a survey with as wide a regionalcoverage as this. Closer definition may advantageous-ly be employed in the more detailed studies whichshould follow this one.

The scheme of classification used in pottery de-scription in this report is hierarchic, with wares asits largest divisions, then vessel shapes, rim shapes,and other details of vessel form. The wares are dis-tinguished by criteria of surface and paste, the vesselsby general shape range, the finer classifications byfurther subdivisions of form. Decoration, although itgenerally conforms to these descending divisions inoccurrence, sometimes crosscuts them, as the othersteps of the classification also do on occasion.

In this generalized report, the wares have beenused as the stage and substage diagnostics. The majorslipped wares have proved particularly valuable aschronological period diagnostics, but in several casesa single ware occurs with no variation, or only slightlynoticeable variation, in surface and paste in two suc-cessive periods. Examples are Flaky Redware betweenthe Late Formative and Early Regional substages,Medium Slateware between the Florescent stage andthe Early Mexican substage, and the Coarse Redwareof the Late Mexican substage and post-Conquest times.It is likely that technical ceramic studies on paste andslip may eventually allow finer subdivisions on a basisof surface and paste but they are unavailable for thisreport, and are not so readily useable in pottery sort-ing as are macroscopic characteristics. In cases wherewares cross major chronological boundaries, the wares

have here been divided on the basis of distinctions invessel shapes, and such less distinctive changes insurface and paste as are observable have been de-scribed for supplementary use in sorting. The descrip-tions of the wares and the section on the sequence ofwares in Yucatan should serve to clarify the criteriaon which these subdivisions were based.

VESSEL SHAPES, DETAILS OF FORM

Vessel shapes in the major ceramic periods of Yu-catan are easily classifiable into certain major cate-gories, which in most cases allow the sorting of bodysherds as well as rims into several major subdivisionswithin wares. Every effort was made to determine thevessel shapes of all periods and to classify all sherdsinto vessel shape groups. This effort was successfulon perhaps 98 per cent of the sherds, due in major de-gree to the information accumulated in notes and pho-tographs taken of nearly 1,000 whole vessels found invarious collections. A description of the nomenclatureof the main shapes follows:

Jars all have a globular body with constricted mouthand a neck showing a marked change in curvature ofprofile from that of the upper part of the body (theshoulder). All jars noted, save in Pattern BurnishedMonochrome and in unslipped ware of all periods, havea bottom bounded by an angular junction with the sidewall. Slipped jars are always unslipped on the interior,thus allowing all concurvate42 sherds with exteriorsurface slipped and interior unslipped to be sorted as

jar sherds.Basins have a globular to hemispherical body, no

neck, and a wide mouth. Slipped basins always havethe interior surface slipped, and usually, although notin all kinds, have the outer surface slipped or partiallyslipped. In general they are large, heavy vessels, al-though this is not true in all periods. Constricted

mouthed vessels which conform to the above specifica-tions are here described as basins, although theseshapes range far from the accepted concept of the wordin English. This was done because such vessels seem

in some groups to form a continuous range with shal-lower wide-mouthed forms. Most basins known fromYucatan have a flat bottom bounded by an angular junc-tion; a few types have a ring base, and one or more mayhave had three hollow legs. Concurvate body sherdswhich have an interior slip can usually be sorted intobasin or various types of bowl sherds by criteria ofradius and kind of curvature, thickness, design, or

finish of exterior surface.Basal break bowls are distinguished by an angular

to subangular junction between bottom and sides. Flooris usually faintly concave. Walls are flaring to outcurv-ing and are well under the bowl radius in height. Slipcovers vessel floor and the interior and usually the ex-

terior sides of walls. Bottom is usually unslipped. Threelegs (tripod) of any of several forms are nearly univer-sal, ring bases and flattened bottoms are rare.

Shouldered bowls have strongly concave floors, an

abrupt change of curvature at about one-half the vesselheight, and rim nearly vertical, often insloping.

Hemispheroid or rounded bowls have no break incurvature of profile. The term hemispheroid is looselyused in this report to include bowls with oval or parab-oloid curves, as well as those whose section includesmore or less than a semicircle. Interior is alwaysslipped, exterior is usually slipped or partially slipped.Annular bases are standard in some periods, flat bot-toms with curving basal junction in others.

Beakers are cylindrical or with slightly flaring sidewalls and with height nearly equal to diameter. Bottommay be flat or convex with a flattened central area,or may have three small legs. They are usually slippedboth interiorly and exteriorly, except for the bottom,and often decorated, usually by incision.

Pvriform vessels are pear-shaped with a bulbousbody gradually constricting to a cylindrical neck. Theyare slipped exteriorly only, and have an expanding(trumpet) mouthed pedestal or tripod rattle legs. Shoul-dered pyriform vessels are a subtype and have an an-

gular break in contour in the lower part of the body.These forms are often decorated.

Cylindrical vessels are similar in form but higherin proportion to diameter than are beakers. Sides maybe straight, outcurving, or incurving. They are slippedexteriorly, the slip extending slightly over the rim.The bottom may be flat; there may be a trumpet base,or tripod rattle legs on a convex bottom. They are of-ten decorated.

Hourglass incensarios have the form of a hemisphe-roid or paraboloid bowl supported on a large trumpetpedestal. They are unslipped, and often decorated withapplique bosses. They usually bear a postfiring coatof lime or white clay.

Grater bowls have an incised pattern enclosed in an

incised circle on their floor. The bowl may be of sev-

eral forms; one distinctive form has a shallow ovalprofile with incurving lip.

Figurine incensarios in Yucatan most commonlyconsist of large unslipped cylindrical vessels on trum-pet pedestals, to one side of which has been applique'da human figurine, hand-modeled, unslipped, but usual-ly painted, with mold-pressed face. They usually beara postfiring white coat, often overpainted in colors.

Of the above forms the beakers, pyriform vessels,cylindrical vessels, and some of the rounded and basalbreak bowls are made with thinner walls (thin ware),are often decorated, and are more often traded thanare the remaining forms. These are called finewares.

47


The naming of various of the less frequent formsmay be obtained from the plate captions.

Several types of encircling moldings are found on

vessels. They usually show such regularity as to sug-

gest the use of a kabal. Descriptive terms are used as

follows.Ridges are small, with cross section approaching

an equilateral triangle with heavily rounded apex (seefigs. 2, h8, i,.

Flanges are considerably higher than are ridges,their sides are less convergent, and the apex more

blunt (see fig. 9, fj.

Skirts have a form similar to flanges or ridges butproject almost directly downward from the vessel (seefigs. 28, ; 58,i)

Basal ridges, flanges, or skirts occur exteriorly atthe juncture between the side-walls and bottom of a

bowl.Cutting and terracing is sometimes performed,

particularly on skirts, ring bases, and slab legs. Cut-ting is done by taking series of notches out of the ridge,flange, or skirt, terracing by cutting out crenelationsor stair-shaped areas.

Bolster rims are thickened on the exterior surfaceof the lip by the addition of a fillet of clay, more or

less semicircular or loaf-shaped in profile (see figs.27, ji 43, 44; 73). They occur on basins.

Bead rims are like bolster rims in form but muchsmaller, never exceeding 5 mm. in depth (see fig. 50,a). They occur on thinware bowls and beakers.

Hoop rims are slightly raised, smooth, wide bandson the vessel exterior adjoining the lip; hoop moldingsare placed lower on the wall, or adjoining the basalangle.

To find the sites at which each ware was found, andthe location of illustrations of each ware, see chart 1.

WARE DESCRIPTIONS

Pattern Burnished Ware

Surface. -Unslipped matte, smooth, lusterless.Pinkish Buff, Light Vinaceous Cinnamon, darkening toGrayish Olive.

Paste.-Medium-fine fracture, Light Vinaceous Cin-namon to deep gray, temper undetermined.

Decoration.-By burnishing in wide and narrow

bands, stripes, and diagonal hachure. Polished areas

faintly lustrous; noticeable only in strong light. In il-lustrations, commas on neck and bottom sherds indi-cate limits of burnished areas.

Shapes.-Bottles, with distinctive narrow, exterior-ly thickened neck; one watch-glass shape.

Comparisons. -Unlike all other wares found in Yu-

catan in form and decoration. Ware color and texturesimilar to badly weathered fragments found at SantaRosa Xtampak and Dzibilnocac in the Chenes area,Campeche. One similar bottle rim fragment was foundat Holactun. This rim (unillustrated) is thinner bothin wall and lip than the main fragments, but otherwiseappears identical. The sherds of the Middle Formativedeposits from the Campeche sites dug in 1949 bear no

distinguishable burnished designs, but include bottle-neck shapes resembling those from Holactun. Thisconstitutes a stylistic link between the Early and Mid-

dle Formative collections.Chronological position. -Sole known Early Forma-

tive ware.

Formative Unslipped Potters

The Pattern Burnished Monochrome of the Early

Formative substage seems to be unslipped. However,this pottery has a distinctively finer texture and smooth-er finish than have the characteristic Yucatan unslippedexteriorly striated jars which began to be made in thisstage and continued almost until the Spanish Conquest.

Striated jars very nearly identical with the Regionalstage unslipped ware jars occurred in several LateFormative stage deposits. In the pure collections ofthis period excavated at Santa Rosa Xtampak and Dzi-bilnocac in the Chenes region, the jars have a bluntoutcurving lip, with neck meeting body in a gradualcurve. No other unslipped vessel forms were noted.

Red Slip over Striated Ware

Surface. -Exterior slipped over a finely striatedsurface; medium lustre. Color Brick Red to MoroccoRed to black.

Paste.-Similar to Formative Monochrome, surfacestriated externally.

Shape.-Jars only.Comparisons.-Completely distinctive. Slip color

deeper than Formative monochromes, surface finishunique. No comparable material from other areas known.

Chronological position. -Late Formative.

Formative Orangeware

Slip.-Lustrous slip; powdery abrasion common,baring paste. Color Vinaceous Tawny to Terra Cotta.

Paste.-Fine texture, no temper visible to eye, frac-ture verges on conchoidal, color Orange Cinnamon.

Shapes. -Bowls.Comparisons.-Much like the Regional Medium Red-

ware from Yaxuna in slip color and paste texture, butshapes are different. Two fragments show trickle paint.R. E. Smith classes nearly all sherds similar to oridentical with Uaxactun Chicanel.

Chronological position. -Late Formative, identifiedonly at Yaxuna.

Formative Monochrome

Surface.-Slipped, smooth, sometimes showing pol-ishing marks; opaque to waxy surface. Color and char-acter of slip variable from site to site; see plate cap-tions. Commonest slip color is a dull orange-red, oftengray or yellow splotched. Slip may appear either opaqueor waxy. Other slip colors found at some sites are adull gray to black, a brilliant polished black, a waxycream color, a clear chalky white.

Decoration.-The waxy orange-red and cream-col-ored slips at some sites show lighter-colored, or oc-casionally darker -colored, trickle paint. This paintshows the characteristics of slateware trickle paint.Combinations of the above colors on exteriors, interi-ors, and lips of single specimens are characteristicat some sites. This ware, with additional study, couldprobably be divided into at least two subwares, the onecharacterized by opaquely slipped surfaces not bearingtrickle paint, the other by waxy slip and trickle paint.The latter variety may be related to Usulutan ware(Lothrop, 1933, pp. 47-51), the paint of which is simi-lar to Yucatan trickle paint. One specimen was foundwhich may well be Usulutan ware (fig. 31, d, 6). A fewfragments among these monochromes bear either redor black paint in designs (see figs. 16, b, 9; 31, e, 25).These may be imports or conceivably may come froma later horizon. Underslip incision and grooving ofvarying character are common, crude painted areasare rarely found, a few sherds show incision or carv-ing after slipping.

48


Paste.-Medium-coarse texture. Color VinaceousCinnamon to Cinnamon Buff to gray. Temper untestedsave at Holactun, where it is all sherd.

Shapes. -Flat-bottomed, basal break bowls with slipcovering bottoms; spheroid shapes; rare jars; cuspi-do r s.

Comparisons.-Ware characteristics do not allowa ready distinction between this pottery and several ofthe Regional coarse and medium paste wares, and itis possible to confuse certain of the waxy slippedpieces with Florescent Slate and Redwares. The ma-jority of the Formative shapes, however, are complete-ly distinctive; Flaky Redware which also occurs inthese shapes is distinguishable by surface characteris-tics. The great majority of the fragments of FormativeMonochrome are classified by R. E. Smith as close oridentical to Uaxactun Chicanel pottery, although certaincharacteristics in the group suggest the earlier Ma-mom phase (cf. Thompson, 1939, pp. 75-85 for similarwares). The fact that monochrome subtypes seem toshow marked segregation among the various sites, andthat Flaky Redware was also made in Formative shapes,suggests that complex relationships may appear in lat-er study of the Yucatan Formative stage.

Chronological position. -Diagnostic for Middle andLate Formative phase.

Regional Unslipped Ware

Surface.-Little irregularity of contour, few project-ing temper particles, texture of surface is rathercoarse. No evidence of polishing.

Color.-Tilleul Buff, Light Ochraceous Buff, PalePinkish Buff, clay color to gray, sometimes smudged,never soot-blackened.

Paste.-Fracture coarse. All examined had calcitetemper.

Decoration.-Striation on the exterior of vesselwalls is very common but not universal on jars of thisware. The striation has been made by raking a toolwith evenly spaced teeth over the surface. Patterningof striation varies by period and area, depending upondirection and arrangement of tool strokes. In some re-

gions, a thin wash of black paint was applied at jarrims. Both of these techniques are also used on slippedpottery of Regional stage. In 1942 it was noted that twotypes of striation could be distinguished on unslippedjars, an early (Late Formative and Regional) type, inwhich the tool marks are shallow and some of the orig-inal surface of the vessel has been left between them,and a later (Florescent) type more deeply striated withlittle if any of the original surface unaltered. The scaleof the later striation is a little larger than the earlier,but does not allow distinction as clearly as does depthof striation.

This typological distinction, although slight and dif-ficult to describe, has proved a ready and reliablesorting criterion, particularly valuable because of thelarge samples (unslipped jar body sherds) availablefor its use. Since unslipped body sherds were not savedfrom Chiche'n Itza', the Puuc sites, Holactun, Coba', andOxkintok, this information is not available for them.It seems likely that bivalve shells were used as thestriating tools. If so, the use of a different species ofshell might produce just such a slight but very constantdifference in striation as that noted. The change in toolmay have resulted from changes either in taste or inthe availability of shell, due to varying marine condi-tions. Or it may reflect a change in Yucatan trade re-

lations. Another hypothesis is that the striations were

made in clays of different properties, or at a differentstage of drying. In a collection from a chultun just

west of the Monjas, Chiche'n Itza (see fig. 65), a uniquestyle of unslipped striated jar occurs with other jarswhich bear striations on the exterior neck wall andFlaky Redware slip on the lip and neck interior. Asso-ciated Flaky Red bowl sherds of mixed Formative andRegional shapes place the striated jars very generallyin time. They may slightly postdate the Yaxuna FlakyRedware horizon, but this is uncertain at best.

Shapes.-The jar with striated outer surface occurs

in all sampled phases of the Regional stage, save theFlaky Redware phase at Yaxuna and the early phase atOxkintok. In the Redware-containing Regional stage de-posits where this kind of jar occurs, it always bearsfine striation. At Oxkintok, the unslipped jar is a ver-

sion of the concurrent slipped jar, and bears the same

black-painted rim decoration; however, striation is oc-

casionally used on jar exteriors at Oxkintok, and ap-pears on interiors of the Flaky Redware jars at Yaxu-na, and on Red over Striated Slip jars of the Mani LateFormative deposits. At Coba', there are unslipped bowlsand basins; at Oxkintok, unslipped basins of uniquewide-flaring form. There are no incensarios of theRegional stage, save those at Acanceh which come

from deposits which bear considerable slateware, andthus are not certainly a part of the Regional repertory,and from Dzibilchultun where the same condition holds.These incensarios bear applique' decoration.

Chronological position. -Regional stage.

Flaky Redware

Surface.-Lustrous opaque slip, with a tendency toflake or spall in areas from a smooth, fine, powdery,cream-colored undersurface. No powdery abrasion ofslip noted. Color a strong red, Ferruginous to Vinace-ous Rufous, with slight if any lighter-colored splotchingbut sometimes grading to dark grays or Sayal Brown.The diagnostic trait of flakiness, which links this ware

with Incised and Punctate Dichrome and with the Trickleon Flaky Redware, may be hypothesized as due to a

preliminary smoothing or burnishing of the vessel sur-

face before the addition of the red slip, a markedly dif-fering coefficient of expansion between paste and slip,a soft or easily disintegrated underslip, or a combina-tion of various of these factors. The powdery, light-colored character of the undersurface in badly spalledspecimens suggests that it may have contained finelydivided calcite, the leaching of which may be responsi-ble for the loosening of the harder red slip. The advan-tage of such a light-colored undersurface might be inthe added brilliancy of red slip color attainable there-by. No technologic test of this hypothesis has beenmade. Shepard (in Thompson, 1939, p. 268) describesa similar effect caused by an underslip on San Jose6 IIpolychromes, but the contemporaneous and earliermonochromes do not show it. Note, however (ibid., p.75), that San Jose I monochromes are described as

having a tendency to flake. Similar slips are describedby Shepard for polychrome of a later phase, BenqueViejo III (Thompson, 1940, p. 11).

Paste.-Coarse texture, color Capucine Buff to Til-leul Buff to gray, usually shows light-grayish temperparti cle s.

Shapes.-Large jars, basins, shouldered bowls.Comparisons. -Distinguishable from other wares of

the same flaky slipped family by its over-all red slipand limitation of decoration to very rare and atypicalsherds such as fig. 6, a, 17 18 21. San Jose Redware,the preponderant Formative ware of that site (Thomp-son, 1939, pp. 75-84), may be related to the Flaky Red-ware described here.

Chronological position. -Late Formative and Early

49


Regional. Specimens distinguished as to horizon bytheir shapes.

Flaky Dichrome

Slip.-Flaky character as that of Flaky Redware,but is Apricot Orange to Vinaceous Cinnamon in color.

Paste.-Similar to that of Flaky Redware. Unslippedinteriors of jars are usually deeply striated. Temperof Balam Canche specimens is calcite; that of othersis unknown.

Decorative techniques. -Painting in large designs,among which rectilinear step-fret predominates; colorranges Vinaceous Rufous, Terra Cotta, Vinaceous Rus-set, Ocher Red. Incising seems to have preceded slip-ping. It is used to outline painted areas, and occasion-ally used alone. A few fragments (fig. 6, a) show in-cised diagonal cross hatching. Punctation is used tofill areas, randomly spaced or in lines. Painting, punc-tation, and incision are usually complementary in de-signs, but occasionally each is used alone.

Shapes. -Large jars, shouldered and basal flangedbowls.

Comparisons. -The stylistic relationships of thisware to basal flange polychrome bowls are discussedelsewhere (see Brainerd, 1948a for a discussion of thedistribution of this decorative technique). Flaky Poly-chrome differs in decorative technique, but seems tobelong to the same ware family.

Chronoloaical position. -Stylistically placed in Ear-ly Regional.

Regional Polychrome

Surface.-Lustrous red-orange slip. Tendency to-ward flaking noted in all save Mani and Chac Cave jars.

Paste. -Texture and color as in Flaky Dichrome;jars internally striated. Chac sample (fig. 3, .i? withgrayish, translucent to opaque calcite temper. MissShepard (1951, p. 243) states that almost all fragmentsare calcite tempered.

Decorative technique.-Painting in black and red on

an orange slip.Shape s.-Large j ar s, basal flange bowls.Comparisons. -Slip and paste color and interior jar

striations suggest Flaky Dichrome; association andintergrading of bowls (fig. 63, a) with Flaky Dichromesuggest nearly the same horizon for all three groups.Jar fragments are associated with basal break poly-chrome bowls at Mani, but in collections of long depo-sitional span. Most of the basal break polychrome bowlsherds are accepted by R. E. Smith as Tzakol poly-chrome. The bowls of fig. 63, a may most logically behypothesized as indigenous. Although there is notenough evidence here to assure the development ofTzakol-style polychrome from the local Yucata'n in-cised dichrome in the face of the very sparse frequen-cies of these wares in our collections, it is difficult toexplain the collection of fig. 63, a as neatly by any oth-er hypothesis. More collections of this horizon are

needed than this thinly spread survey has produced.Chronological placement.-Middle Regional, on sty-

listic criteria.

Trickle on Flaky Redware

Surface.-Slip closely similar to that of Flaky Red-ware, showing same lustre and spalling. Color Vinace-ous Rufous, Fawn Color, Sayal Brown to Deep GullGray.

Paste.-Closely similar to that of Flaky Redware.Texture rough and crumbly at fracture.

Decoration.-Trickle paint of rich, constant, deep-black color (see fig. 7). Although disposition of paintis closely similar to that of the slatewares, the colorin constancy and depth of tone is completely unique.

Shapes -Large jars, basins.Comparisons.-This ware is found associated with

other members of the Flaky Redware family in severalinstances. The decorative techniques of this and thedichrome group are never combined. Two successivegroups are stratigraphically separated at Yaxuna (seefig. 7). This ware documents a continued tradition ofthe trickle method of paint application which lastedfrom Formative through Middle Mexican times.

Chronological placement.-Early and Middle (orLate) Regional.

Thin Red-Brown Ware

Surface. -Slip lustrous and waxy. Color Sayal Brownto Wood Brown to Cinnamon.

Paste.-Smooth, powdery surface where unslipped;medium fine texture. Color Cinnamon Buff and paler.

Decoration.-One sherd bears a red paint splotch.Shapes -Small jars and basins.Comparisons. -Similar to Thin Slateware of Chichen

Itza variety (cf. fig. 4, p). Its presence in the FlakyRedware deposits suggests that it may be an early pro-totype of Thin Slateware. Only 12 sherds found.

Chronological placement.-Early Regional.

Fibre-Tempered Ware

Surface and paste.-Unslipped, very smooth surface,silky to the touch. Color Tilleul Buff; honey-combedappearance due to many cavities left by the vegetaltemper particles. These cavities under magnificationshow transparent, siliceous, unbranched filamentswhich may have come from a grass, agave, or othermonocotyledonous plant. This pottery is immediatelyand easily identifiable in Yucatan collections by weightalone. It is much lighter than any other ware.

Shapes. -Basins.Comparisons. -Fibre-tempered pottery is known

from Florida, although in different shapes, but is un-known from the Antillean area (personal informationfrom Irving Rouse, 1946). Pottery tempered with treebark has been reported from various localities in theAmazon and Orinoco drainages. Fibre-tempered pot-tery belongs in the early pottery horizon of Southeast-ern United States. Association in the Yaxuna Cenotedeposits place this ware with reasonable certainty asearly Regional; its appearance at Coba' in identical ves-sel form suggests a middle or late Regional dating.Inclusive dates may thus be placed at 300-600 A.D. inYucatan. These occurrences can in no sense be consi-dered as valid evidences of cultural diffusion; vegetaltempering as a ceramic trait must have had a longlife, and been independently invented often. It is known,for example, from Neolithic West Asiatic cultures.

Oxkintok Coarse Monochrome

Surface. -Slipped, lumpy with imperfectly smoothedtemper particles on jars. Smooth on bowls. Slip dur-able, lustrous. Three color ranges. The first has Cin-namon Buff predominating, varying through TawnyOlive, Orange Cinnamon, Drab, and often splotchedwith black. This occurs on most jars and a few bowls.Second range is red, varying from Vinaceous Rufousthrough Mars Orange, English Red, Japan Rose, andoccurs on nearly all bowls and some jars. Third is ahigh-gloss black. See plate captions.

50


Paste.-Coarse to medium textured, temper crystal-line or powdery calcite.

Shapes -Jars, basins, basal break bowls; large,very shallow, rounded bowls.

Decoration.-Thin, matte, brownish-black paint on

jars and bowls.Comparisons. -There is considerable similarity in

shape between this ware and the late Regional Yaxunaand Coba slipped wares, and Acanceh wares of thesame period; there are also similarities to Dzibilchal-tun Red on Thin Gray. Basal break bowls and bolsterrim basins are closer to Florescent slateware formsand farther from Peten forms than are the shapes ofthese comparable Regional wares, suggesting that theOxkintok Regional pottery assemblage is more closelyakin to the slateware tradition which seems to haveoverwhelmed all other ceramic traditions before theend of Florescent times. Sorting between Coarse Mono-chrome and Thin Monochrome has been made primari-ly on vessel shapes, along the same general boundariesas in the slatewares of the Puuc sites. As among Puucslatewares, there is considerable intergrading, althoughin general the rounded bowls have finer-textured pasteand thinner walls than do the other forms. See figurecaptions for detailed analysis.

Chronological placement. -Middle Regional.

Oxkintok Thin Monochrome

Surface. -Slipped, lustrous to nearly glassy polish,showing tool marks. Colors Cinnamon Buff to greenish,deep red, high-gloss black, red with fine, evenly spacedgray flecking.

Paste.-Fine textured, hard, thin: 3-5, rarely to 6,

mm. Temper undetermined.Decoration.-By surface alteration, all under slip

and including shallow incision, vertical and diagonalchanneling, low relief appliqued lozenges, appliquedmonkey faces. All decoration done with marked virtu-osity and restraint. This ware, although not showy,ranks high among New World ceramics in the author 'sopinion. More material would justify its subdivisioninto several types. See figure captions for detail.

Shapes.-Shallow oblate spheroid bowls with flat-tened or ring base, low vertically sided bowls, small

cup.Comparisons.-The group shows intergrading in

slip surface, paste texture, and wall thickness withOxkintok Coarse Monochrome. One delicate cup, fig.12, L was classified by R. E. Smith as fitting UaxactunTzakol. One bowl of this ware in the Me'rida Museumis said to come from Campeche (fig. 103, d). No otherresemblances to ceramics from other areas were

noted, although the quality of this ware is equal to thatof most known tradewares. The accompanying Tzakolstyle polychrome consists of only 7 small fragments(fig. 9, c.).

Regional Coarse Redware

Surface.-Medium lustre, adherent slip. Color Ru-

fous to Testaceous to gray brown showing oyster graymottlings and, rarely, dark smudging. Samples from

Acanceh and Mani lack mottling; jars approach a blackcolor.

Paste.-Coarse to medium texture. Color gray tobuff. Temper is calcite and ash, frequencies unknown.

Shapes .-Jars, basins.Comparisons.-Differs from Oxkintok Coarse

Slipped ware in presence of mottling and absence of

Cinnamon Buff color range. Slip somewhat darker and

less orange than Formative monochromes; differen-

tiated from Flaky Redware by adherent slip and fromRegional Medium Redware in coarse paste texture.There is intergrading in all these features at variousof the sites. Shapes are a major diagnostic.

Chronolozical position.-Middle and Late Regional.

Regional Medium Redware

Surface.-Very similar to that of Regional CoarseRedware, but smoother, more lustrous finish; mottlingcommoner, smudging absent. Acanceh and Mani sam-ples rarely have mottling. One fragment has a gessocoating.

Paste.-Medium to fine textured, ash temnpered atYaxuna and Coba. Shepard (1951, p. 243) suggests thatthis ware may be foreign. Color Ochraceous Salmonto Ochraceous Orange. At Acanceh temper unknown,color darker to gray.

Shapes.-Medium size jars, bowls with basal ridge,hemispheroid bowls.

Comparisons.-Western variety of this ware seemsthe prototype of Red on Thin Gray, and probably gradesinto regionally, or immediately follows temporally theOxkintok monochromes. There also seems to be anintergrading into a dark-colored slateware, and maybe intergrading into the Florescent redware of the Puucsite s.

Chronological position. -Middle and Late Regional.

Coba' Dichromes and Polychromes

Surface.-Polychrome has a glossy, soapy, slippedsurface, with underlying Ochraceous Salmon fine-tex-tured paste with a tendency to a powdery, abraded sur-face when slip wears off. Pastes are tempered withcalcite or tuff. Dichromes have a matte surface; pastesimilar to polychrome, presence of slip dubious.

Shapes -Deep bowls with increased curvature atwall-bottom juncture; bottoms faintly flattened. Di-chrome also made in a shallow bowl with sublabialflange and ring base.

Comparisons.-Some of these may be variants ofthe Regional Medium Redware of Coba; their frequencysuggests local manufacture. The polychromes were ac-ceptable by R. E. Smith as Tepeu 1 wares, but Shepard(1951, p. 243) suggests a local origin on the basis oftechnological analysis. See captions to figs. 2 and 3for details.

Chronological position.-Middle or Late Regional;Tepeu 1 cross-tie.

Fine Gravwar e

Surface. -Seemingly unslipped; powdery to smooth,with faint lustre. Color same as paste when powdery;one shade darker, Dark Dull Gray, when smooth.

Paste. -Very fine textured; probably untempered,but no microscopic test made. Color Pale Neutral Gray.

Decoration.-Incised, punched, grooved; see figurecaptions.

Shapes.-Lower beaker with flat or double bottom,flat-bottomed outcurved rim bowl.

Comparisons. -Very distinctive because of extremefineness of paste and clear neutral color. Forms inter-grade with wares of other areas (see captions) and withcertain decorated slateware vessels.

Chronological -position.-Regional-Flor e scent tran-sition.

Thin Blackware

Surface.-Matte to slight gloss; slipped. Color dead,

51


brilliant black to Neutral Gray. One fragment with ges-so coating.

Paste.-Fine textured; most fragments from Holac-tun are sherd tempered; some sherds bear fine, trans-lucent, colorless, rounded grains. Some fragmentsshow sulphur colored areas in the paste, as do certainlustrous blackware fragments from Xpuhil, Rio Becar ea.

Shapes. -Similar to those better known in Fine Gray-ware.

Comparisons.-A scattering of blackware fragmentsfrom the Puuc sites may belong to this group, but di-rect comparisons were not made. Xpuhil fragmentslikely the same ware.

Chronological position.-Regional-Flo re scent tran-sition. Perhaps continuing into early Florescent times.

Red on Thin Grayware

Surface.-Thin, sometimes abraded slip over asomewhat irregular surface. Slip matte or polished.Color red to Cinnamon Buff. Buff slip commonest onbasal angle bowls.

Paste.-Coarse to medium texture. Deep Gull Grayto Light Gull Gray in color. Paste somewhat friable.

Shapes.-Medium size jars, basal angle bowls.Comparisons.-Wall thinness is to some extent diag-

nostic, but approximates that of Coarse and MediumRegional Redware. Slip color range is similar to thatof Oxkintok Monochromes. Gray, friable paste is bestdiagnostic from both these wares. Red on Thin Gray-ware is shown to be definitely contemporaneous withMedium Slateware at Dzibilchaltun, and to be gradual-ly replaced by slateware during the time of deposition.This situation has been used to place the site sequen-tially.

Chronological placeme nt.-Regional-Florescenttransition.

Dzibilchaltun Fine Orange

Surface.-Unslipped, or may bear areas of ApricotOrange, clear white, or Quaker Drab slip. Unslippedsurfaces either powdery or faintly glossy. White slipopaque with medium lustre; black slip thin, a stain.

Paste.-Very fine-textured; probably untempered,although untested microscopically. Clay color and textture distinctively like that of fine orange of other pe-

riods.Shapes.-Outcurve-sided low beakers, barrel-shaped

cylinders, flat-bottomed sharp basal angle bowls;hemi spheroid bowls rare.

Decoration. -Incised through slip; removal of back-ground areas rare. Black painted lines and designspresent.

Comparisons.-Differs from its most similar ware,Puuc Fine Orange, in that incised lines are thicker,less variable in width, and clumsier in execution; rimshapes are mostly outcurved; designs, motifs differ.

Chronological position. -Regional-Florescent tran-sition.

Florescent Unslipped Ware

Surface.-Like that of Regional Unslipped ware;smooth to porous, no polish. Color commonly TilleulBuff to gray, sometimes smudged. Texture of surfaceporous.

Paste.-Fracture coarse. Temper predominantlyor wholly calcite in jars. Some incensarios show sherd,others volcanic ash temper, suggesting that they maybe made of Medium Slateware paste left unslipped.

Decoration.-All jars have externally striated bod-ies; this striation is deeper and coarser than that onRegional-stage unslipped jars. Incensarios bear seve-ral varieties of hand-modeled applique' decoration ontheir exterior walls, and often show traces of a thick,white, calcareous coat, applied after firing. The appli-que decoration is varied: applied crude tetrahedronsin patterns, cut and sometimes reed-stamped encir-cling fillets, variously shaped lugs.

Shapes.-Large to medium size jars, incensarios,perhaps some bowls, although such fragments may al-so belong to incensarios; rare incense ladles with tu-bular handles.

Comparisons.-The unslipped, striated-surfacedjar is found in the Central Maya area in late Forma-tive times43 and continues through all periods. Its ear-liest certainly documented appearances in Yucata'n areat Yaxuna in a phase strongly influenced from the Peten(Tzakol and Tepeu 1), later at Dzibilchaltun and Acan-ceh in a phase showing a marked mixture of slatewareswith monochromes which seem to derive from the ear-ly Oxkintok phase (where fully developed unslippedstriated jars are absent). In the Chenes sites (1949season), striated jars occur in slateware depositswhich probably are equivalent to the Yucata'n Regionalstage. Thus the unslipped, striated jar would seem tohave appeared from the south, accompanying bothslatewares and Peten-like monochromes, but withforerunners in the Flaky Redware and Early Oxkintokhorizons.

The unslipped incensario is earliest recognizableat Dzibilchaltun and Acanceh with the Regional-Flores-cent horizon, and, like the unslipped striated jar, seemsto run through the Florescent and into the succeedingMexican stage as a continuous tradition which showsno striking changes in form or decoration. Unslippedincensarios begin in the Formative phase in the Guate-mala highlands (De Borhegyi, 1950, 1951), but neitherform nor decorative technique closely approaches thosetreated here. At Uaxactun in the Tzakol period (Smith,1936a, fig. 4, 8-10) and in the corresponding San JoseII period (Thompson, 1939, p. 87), unslipped incensar-ios with applique spikes not unlike those on the Yucatanspecimens appear. This suggests that the spiked incen-sario may have diffused north from the Central Mayaarea into the Yucatan peninsula, a hypothesis mademore likely because of the known high-ceremonial de-velopment of the Peten Maya during this time period.

Tubular-handled incense ladles are rare. They prob-ably represent either stray sherds of the Early Mexi-can substage, or an evidence of Mexican influence ante-dating the Mexican substage. Some of the Puuc archi-tectural detail also suggests such influence. Incenseladles, but with trough-shaped or solid cylindricalrather than tubular handles, are contemporaneous withthe Florescent stage in Guatemala (see Kidder, Jen-nings, and Shook, 1946, fig. 201, a, b for examples).

Flor e scent Medium Slatewar e

Surface.-Slipped, slight to medium lustre; waxy,translucent appearance, varying occasionally to opaque;often shows faint encircling marks such as might re-sult from applications of cloth to a turning vessel.Sometimes bears an inconspicuous, regular crackle;dendritic white opaque, or purple markings common,which seem to have been caused by contact of rootswith the buried pottery. Depressed slipped areas of-ten chalky and thus suggestive of a polishing technique,but lack of tool marks suggests a soft implement. Co-lor corresponds quite closely to that of the body claysurface, and ranges through oyster gray to brownish

52


grays, brownish reds, purplish grays; Avellaneous,Light Pinkish Cinnamon, Cartridge Buff are common.

Maertz and Paul colors seem a closer match: 14A5,12B3, 13C4, splotches of 56L7, 11E4, 37A1, 12B2,94H6. Color often shades to a dark brownish gray inlarge areas. Light-grayish slip tints usually are foundon calcite tempered pastes.

Paste.-Texture medium, sometimes grading tofine. Colors Vinaceous Cinnamon in reddish areas toAvellaneous in yellowish areas; light gray also seen.

Tempering materials include calcite of various types,powdery to crystalline and in various stages of subdi-vision, tuff in considerable frequency, reddish claylumps which may have been a constituent of the clay,and crushed sherds; several of these types sometimesoccur mixed in single sherds. Frequency of tempertypes varies markedly at the various sites tested;light paste color is definitely correlated with calcitetemper. Temper classifications, if done on all collec-tions, give promise of much aid in determining region-al and chronologic differences in this ware, the pres-

ent classification of which is far from satisfactory.The finer-textured pastes of this group are identicalwith that of Thin Slateware; sorting distinctions havebeen drawn in this classification on vessel shape andthickness of wall rather than upon coarseness of pastetexture, which is only in general positively correlatedwith wall thickness.

Decoration. -In trickle paint; decoration ranges

from a color lighter than slip to a deep black. Com-monest range is a deep brown; cinnamon tints common.

Outlines normally blurred and indistinct. Color oftenhas a clotted appearance on lighter background, some-

times no color save for faint dark outline. Lines usual-ly 1 cm. or wider. Jars and basins invariably bear ir-regular vertical trickles, exteriorly on jars, both ex-

teriorly and interiorly on basins. Bowls bear rim blobsand floor decorations, some elaborate. Although unan-

alyzed, this paint behaves like an organic pigment; thevague outlines and occasional reversed or disappearingcolor are both characteristic of this type of paint. Inaddition to the trickle paint, this ware is occasionallydecorated with preslip or postslip incising, specularred slip bands, or incising through red slip bands.Thumbed or design-stamped fillets and mold-presseddesigns on hollow legs were also used as decoration.

Shapes.-Large, medium, small, and miniature jars.Jar lids. Basins, small basins, basal break bowls withvariety of tripod leg shapes, some with basal ridgesand skirts. Rounded bowls with exterior beveled or ta-pered lip, and ring or flattened base. Drums with heavy,exteriorly striated, unslipped rims.

Comparisons.-This ware is the distinctive and pre-dominant ceramic product of the northern YucatanPeninsula. Its distinguishing characteristics are thewaxy slip and the trickle paint. Both go back to theFormative stage, where certain of the slipped mono-chromes bear waxy slip, though of an orange cast nottypical of developed slateware, and also bear the char-acteristic vaguely bordered trickle paint. The earliestknown ware approaching the form repertory of Flores-cent Slateware is Black Trickle on Flaky Redwarewhich, however, lacks the slip and paint characteris-tics of slateware. With the coming of the Mexicanstage, the waxy slip grades into opaque white. CoarseSlateware follows with white opaque slip and the lastappearance of trickle paint decoration.

Chronoloaical placement. -Flore scent into EarlyMexican. Earliest appearance is hard to date, and was

likely localized. In Yucatan, on evidence of the earlyOxkintok collections, it seems to follow 9.2.0.0.0. AtSanta Rosa Xtampak it was almost the only slipped

ware before 9. 16.0.0.0. At Yaxuna it seems definitelyto have been present, but not dominant, at about9.12.0.0.0. There seems good reason at present tohypothesize that it developed with mosaic masonrydecoration of the Puuc-Chenes-Rfo Bec style, andoriginated in those areas contemporaneously with theRegional wares of the Yucatan area; gradually, withits accompanying architectural style, it supplanted theearlier wares of that area. It must be emphasized, how-ever, that sampling has been scanty and sporadic; fu-ture work may alter the picture.

Thin Slateware

Surface.-Slip with medium lustre, less waxy andtranslucent in appearance than Medium Slateware;characterized by a fine crackle in which predominantdirection of line is vertical. Dendritic purple markingscommon, vessels all over-slipped. Colors, Pallid Pur-plish Gray to Cartridge Buff to Pale Olive Buff, arelighter than those of Medium Slateware.

Paste.-Is fine-textured at fracture, but is tempered;light gray to buff in color; walls are thin, averaging 4to 5 mm., rarely less than 3 mm. or more than 6 mm.thick. Tempered with fine calcite in lightest-coloredspecimens, others with tuff, clay lump.

Decoration.-Vertical preslip channeling and preslipincising are rare. Trickle paint, nearly universal onMedium Slateware, never occurs. Gesso coatings,white, rose, and green, occurred on a few sherds.

Shapes.-Tall and low beakers, hemispheroid bowls.Comparisons. -Considerable intergrading with Me-

dium Slateware in surface and paste characteristics.Vessel shape and wall thinness were valuable as sort-ing criteria, and were followed rather than paste in afew cases where fine-textured paste was used in shapesnormal to Medium Slateware. Certain variants of thisware, found in early Florescent collections and at Chi-chen Itza' in collections of uncertain date, suggest thatthis group may have derived forms from such earlierfinewares as Fine Grayware and Thin Blackware; thecommon hemispheroid bowl shape in this ware seemsa newcomer. The Thin Red-Brown Ware found at Ya-xuna (see fig. 6, 1) may represent an early stage in thedevelopment of Thin Slateware.

Chronological position. -Flor e scent.

Florescent Medium and Thin Redware

Surface.-Slipped, medium lustre. Slip adheres close-ly to paste; no tendency toward crackling or spalling.Rarely with dark blotches, sometilnes bears purpledendritic markings. Slip appears translucent and waxy,varying with paste color. Color ranges: Ferruginous,Vinaceous Rufous, Japan Rose, Congo Pink.

Paste.-Medium to fine texture, but never untem-pered. Color always red, slightly lighter than that ofslip. Temper about equally divided between tuff andcalcite; sherd and clay lump temper rare.

Shapes.-Beakers, thin- and thick-walled hemisphe-roid bowls, small basins, basal break bowls.

Comparisons.-This ware may be considered as acolor variant of Medium and Thin Slatewares since itoccurs with the slatewares, and in the same shape rep-ertories. Redware in the Puuc sites is largely re-stricted to vessel shapes of the 'Thin" category. Dif-ference between slatewares and redwares on basis ofcolor is complete, with no intergrading noted. The col-or difference is probably caused by choice of bodyclay. Color and slip qualities as well as form makethis ware distinguishable from the Regional and laterredwares.

53


Chronological -position. -Florescent stage.

Holactun Slateware

Surface.-Slip lacks the waxiness of Medium Slate-ware, appearing opaque; ranges lighter in color (Til-leul Buff to Light Gray to Neutral Gray to white); anddoes not match underlying paste color. In unpaintedareas there is a tendency toward spalling.

Paste. -Medium-textured fracture. Color CreamBuff, Apricot Orange, or Deep Neutral Gray. Temperall or nearly all sherd.

Decoration. -Paint only. Lustrous black to NeutralGray. More resistant to weathering than slip. Appliedin regularly and closely spaced trickles, with heavy,bulbous thickenings merging into horizontal zones nearthe rim.

Shapes.-Small jars with either nearly vertical rimwith T lip, or flare rim with outbent lip. Two opposedvertical loop handles at shoulder rim angle have astraplike cross section. Jar lids found. Basins have Tlip, strap handles are horizontally placed. No bowls ofthis ware were identified at Holactun.

Comparisons. -Although this ware is closely simi-lar in general form and decoration to the MediumSlateware of the Puuc sites, it can readily be sortedby slip and paint color range, reinforced by rim vari-ations and the prevalent sherd temper. The prevalenceof sherd temper in the Formative Monochrome fromHolactun suggests a possible continuity of ceramictradition there and, as a corollary, the placement ofthis ware earlier than Florescent. The presence ofthis ware in smaller quantity in Puuc sites argues forcontemporaneity and trade with Holactun. Holactunslate is reported at Etzna by Ruz (1945, pp. 50, 61),who implies that it is abundant there in associationwith slateware of Puuc type and that it postdates ste-lae 18 and 19, which date 9.12.0.0.0 and 9.13.0.0.0 andare in turn associated with late Tzakol or early Tepeuceramics. The almost complete absence of HolactunSlateware at Santa Rosa Xtampak and Dzibilnocac inthe Chenes region is rather surprising, in view of theabundance of the ware at Holactun and Etzna, and ofits presence in Puuc sites. Since these Chenes sitesmust be at least partially contemporaneous with thePuuc sites, a cultural barrier of some sort betweenthem and Holactun and Etzna is suggested.

Chronological position. -Florescent stage.

Puuc Fine Orangeware(Z-type Fine Orangeware, see Brainerd, 1941, pp.173-178, plate III, g, i-cc, for detailed descriptions.)

Surface. -Smooth, often bearing red or white glossyslip. Both paste and slips usually show chalky abra-sion. Unslipped surface ranges from between LightOchraceous Salmon and Light Ochraceous Buff throughLight Ochraceous Salmon to Drab Gray.

Paste.-Untempered, with a fine, chinalike frac-ture. Color through paste matches surface color; seeabove.

Decoration.-Modeled gadrooning, incising throughslip into leather-hard or dry clay, painting of areaswith red or black slip. A gesso slip is found rarely;two colors, rose and green, were noted. The whiteslip is of glossy finish, clear color, fine texture, thin-ly applied, and quite opaque. Red slip bears mediumgloss, ranges nearly to a black color, and when thinlyapplied often shows a streakiness which is used in de-signing (fig. 59, , 3). Black slip appears like a stain,not altering the surface gloss of the paste and of noappreciable thickness. In some specimens, the streak-

iness caused by light application is used in designing(fig. 103, m). Incising always penetrates the slip andis drawn rapidly with a smooth calligraphic techniqueon a leather-hard to dry surface. Most designs includeincised background areas. Drawing is more delicatelydone than on either Chich6n Itzg (X type) Fine Orangeor Dzibilchaltun Fine Orange, utilizing changes in pres-sure for variations from a thick to a thin line. Designarrangement seems to be limited almost exclusivelyto paneled bands, most panels showing two alternatingdesigns. Common elements are rounded or squarishcircles and scrolls, and delicately drawn curlicueswhich are used as space fillers.

Shapes.-Bowls of parabolic form with tapering lips,with or without ring or low pedestal bases. Barrel-shaped to wide-mouthed pyriform vessels with flatbottom, perhaps at times with pedestal base. Jars arerare, and of variable form; all vessels are small andthin walled.

Comparisons. -The pastes of the four kinds of fineorangeware are easily distinguishable from all localpotteries, and are closely similar to the paste of figu-rines from the Island of Jaina. Since the Jaina figurinesare presumably of local manufacture, they evidence alocal use of an untempered, orange-colored paste. Ruz'sreport of abundant Z Fine Orange from excavations onthe Island of Carmen (Ruz, 1945, p. 69), as well as theexample from a Carmen collection illustrated as fig.36, i, suggest that this area may have been at or nearthe locality of manufacture. Two vessels reported fromJaina, our fig. 103, m and possibly that illustrated byPifta Chan (1948, figs. 22, 23), although the paste ofthis latter piece may not be fine orange, lend supportto the theory that the Florescent styles of fine orange.ware may have Campeche coast origins. The chrono-logical precedence of Dzibilchaltun Fine Orange overPuuc Fine Orange, as well as their placement in asingle tradition, seems unquestionable; the form dif-ferences between the two wares are in accord withform sequences known in Thin Slateware, Fine Gray-ware, and the blackwares; their decoration intergrades,and their pastes seem nearly identical. Fragments ofPuuc Fine Orange illustrated from other areas (seeBrainerd, 1941, pp. 178-179) when identical are likelyindicators for trade with the Campeche coast. PuucFine Orange is quite distinctive from the later Chichenand Mayapan Fine Orange wares both in form repertoryand in design.

Chronological position.-Flor e scent stage.

Mexican Unslipped Ware

Surface.-The same as Florescent Unslipped Ware.Color also the same. Texture of surface porous.

Paste.-As in Florescent Unslipped Ware. Temperbelieved to be exclusively calcite.

Decoration. -Striation on jar exteriors continuesthrough this stage. Crudely modeled animal-head lugsare added to jars in the Late substage. Incensario de-coration is augmented by the use of applied discs andperforations, and later by vertical flanges and appli-qued, elaborately modeled human figures with moldpressed faces. Hand-modeled decoration occurs onthe forms and hand-modeled animal figurines arefound. A thin, irregular red wash is found in EarlyMexican times, on "Mixtec braziers" and on smalltripod, basal break bowls. White calcareous paint anda pale blue paint, as well as occasional use of othercolors, are characteristic of incensarios and also oc-cur on unslipped vessels other than jars.

Shapes.-Striated jars continue through the stage,with change in rim shape during the Early substage.

54


In the Early substage, the following shapes also occur:comales are very rare; "Mixtec braziers" with perfo-rated sided pot-shaped body, tubular handle, and twotall legs; small basal break bowls with tripod legs.Middle substage collections are small, and unslippedware forms not clearly known. In the Late substage,the "Mixtec braziers" and comales have disappeared.Large flattish bowls, perhaps with pedestals, and ba-sins with constricted rims appear. Also probably ofthis substage are hand-modeled hollow animal figu-rines, masks, and animal effigy bowls.

The incensario form shows changes during Mexicantimes. A new, taller, more constricted form on a ped-estal appears and, in either the Late Mexican substageor during the succeeding interregnum, the figurine in-censario, with cylindrical body on a pedestal, appearswith vertical flanges and appliqu6d human figurineswith mold-pressed faces and hand-modeled body anddetail s.

Comparisons.-Origin of the new jar rim shape isobscure. Comales are common in the Guatemala high-lands and in the Motagua Valley. The new incensarioshape has also been reported from the Motagua Valley.The animal figurines have parallels at Santa Rosa,British Honduras, and at Cerro de Las Mesas, Vera-cruz, as do the figurine incensarios, whose verticalflanges suggest Guatemala highland forms. Attachedhuman figures on incensarios occur earlier in Oaxaca,and have a tremendous range along the east coast ofthe Mesoamerican area in immediate pre-Conquesttime s.

Chronological position. -Mexican stage.

Mexidcan Medium Slateware

Surface. -Slipped, smooth with faint lustre. Slip isadherent, with little tendency toward weathering orspalling. It is often grayish-white to gray in color,appearing opaque, and lacking the soapy, translucentappearance of Florescent Medium Slateware. Slip col-or is often independent of paste color. Considerableintergrading with the soapy, translucent slips of Flo-rescent Medium Slateware was noted.

Paste.-Medium texture of fracture, color rangingfrom reddish-buff, the commonest color, to gray. Tem-per appears to be exclusively volcanic ash.

Decoration. -In trickle paint, with character seem-ingly identical to that on Florescent Medium Slateware.There is some difference in the designs of the two. Sand U elements are used on the shoulders of mediumsized jars (figs. 71, ; 72, e, k, ); series of spiralsare pendant from the rim on bowl interiors (fig. 74, c).

Underslip incision is rarely used as on FlorescentMedium Slateware, usually in designs of X Fine Orangestyle (fig. 72, a, b, c, h).

Note in fig. 72, a that black paint has been used asa slip band, also as in X Fine Orange, cf. fig. 80. Thefine orange designs on this ware are always slippedafter incising. Incising on drum rims (fig. 72, k) andon grater bowl interiors (fig. 74, i) is usually done onunslipped areas. Graffiti drawing (fig. 72, d) was prob-ably an idle time pursuit on fired vessels or on sherds.

Shapes.-Large and medium-sized jars with eithertall, concave, cylindrical necks or short to vestigalthickened necks bearing external banding grooves. Ba-sins in general Puuc shape and size range, but withdirect rims, or bearing bolsters differing from Puucforms in that they lack an internal angular terminationof the lip; low bowls which in profile range from havingan unbroken curve from bottom to lip, through suban-gular bottom rim juncture. Most of these bowls haveflattened bottoms and no base or legs; a few of the

subangular specimens bear thin, hollow, rattle legs.Oblate hemispheroid grater bowls, floor with circularincised area, bear hollow rattle legs. Rattle legs arenormally tenoned into the bottom and then luted. Ped-estal stands with flat top on an annular base. Tall cy-lindrical vessels with trumpet base. "Pestles" withpunched bottoms, curved taper shank topped by crude-ly modeled animal head. Drums are probably indicatedby large diameter, heavy outcurving rims with un-slipped, vertically striated exteriors. A few effigyvessels were found.

Comparisons.-This ware shows close similarityto Florescent Medium Slateware. In the tabulation ofthe materials done in 1940, the defining limits betweenthe two wares were drawn on the basis of shape anddecoration; the somewhat variable difference in slipappearance was used only secondarily. Most of the col-lections at Chiche'n Itza' show some mixing of EarlyMexican and Florescent forms. All sherds not fittinginto the Puuc repertory and having a light slip colorwere assumed to be of Early Mexican substage, sincethe succeeding Coarse Slateware of the Middle Mexicansubstage is distinctively different. Thus, the primarysorting criteria between Florescent and Early MexicanMedium Slatewares were not those of ware (paste andslip), but of form and decoration.A weak point in the sorting technique used above is

the possibility that some pottery types not found inPuuc-region Florescent collections may have occurredin the Florescent stage of the Chiche'n Itza area. Wehave no large, pure collections of this stage from thearea of Chiche'n Itza'. That this error did in fact occuris evidenced by the finding in 1949 of certain formsformerly believed to be Early Mexican diagnostics atChiche'n Itza in the substructure of structure 1 at Xpu-hil in the Rio Bec area. This typical Rlo Bec structure(Ruppert and Dennison, 1943, frontispiece) is almostcertainly of pre-Mexican date. There are two of theseforms, the pedestal stand shown in fig. 71, e-i, and the"grater bowl" or bowl with scored floor (fig. 74, i, j).The Xpuhil pedestal stands are almost identical withthose from Chiche'n Itza; the bowls are of differentform, but still bear the scored pattern on the floor.The amazingly close correspondence between pedestalstands and 'stools" from Marajoa on the Amazon hasbeen remarked (see fig. 71, caption). The grater bowlhas a long history (see fig. 74, caption), but the mostlikely antecedents to the Chichen Itza forms wouldseem to be those from Xpuhil, southeastern Campeche.The designs show similarity to one of the two stylesof Mazapan pottery (Linne', 1934, figs. 70, 71, 73, 94,a) which may be significant. If so, the Mazapan bowlsmust owe their designs to contact with Yucatan, wherethese designs have a long history.

The tradition of Early Mexican Medium Slatewaremust stem from that of the Florescent Medium Slate-ware which seems to have been used throughout theNorthern Yucatan Peninsula, although our only goodrepertory comes from the Puuc sites. The local (Chi-che'n Itza) variant of this ware may reasonably be hy-pothesized as influenced from the Rlo Bec area to thesouth on evidence of the pedestal stands and graterbowls. A more pervading stylistic influence in bothform and decoration from imported X Fine Orangepottery from Veracruz is abundantly evident (see cap-tions for figs. 71-74, 82, 83).

Chronological position.-Early Mexican substage.

Mexican Medium Redware

Surface -Smooth, medium polish, varying fromwaxy, translucent surface with grayish "bloom" and

55


deep-red undertone often broken by purplish dendriticstaining, to a strong, more opaque deep red, some-times flecked with holes and cracks with black weath-ered edges. Surface often variable on same vessel.Color Etruscan Red to Testaceous. Mottled areasWood Brown to a pewter or greenish-gray suggestiveof plumbate, but with less reflection. Occasionally iri-descent. Occasional pieces are blackened.

Paste. -Similar in texture to that of Medium Slate-ware, but always red in color (Vinaceous Cinnamon toLight Pinkish Cinnamon). Temper of this ware as sam-pled by Miss Shepard is almost exclusively volcanicash.

Decoration.-Underslip incising, save on three bowlsherds which are incised through slip. Incised areasometimes covered by a white or black slip band. Nev-er bears painted designs. Decoration follows the styleof X Fine Orange incising (see figs. 82, 83 for compar-isons), but may also show evidence of local tradition.As in the Medium Slateware of this period, the tradi-tion of underslip incising has carried over from Flo-rescent times.

Shapes.-Medium-sized jars, pyriform and cylindri-cal vessels with cascabel tripods and trumpet bases,bowls, drum-shaped vessels, basal break bowls, shal-low spheroid bowls.

Comarisons.-This ware is certainly grounded inFlorescent tradition, the Thin Redware of the Puucsites. Slip and paste of the two wares show markedsimilarity. However, the style of shape and decorationof Mexican Medium Redware shows more influencefrom X Fine Orange ware than does the style of Mexi-can Medium Slate (see figs. 85-88 and captions). Formsseem to be X Fine Orange copies, save perhaps forfig. 86, a-c, as does the decoration, save perhaps forfig. 85, a, although there are various suggestionsthroughout of influence of the local tradition in shapeand design. This strong influence of the Mexican main-land imported fine wares is not surprising; this ware,like its Florescent predecessor, was fabricated intomedium-sized and small vessels, many of which aredecorated and which must have had a restricted "upperclass" use. The "upper class" people of this periodwere culturally oriented toward the Mexican mainland.

Chronological position. -Early Mexican sub stage.

Red on Medium Slateware

Surface and paste.-Variable; some pieces have a

waxy Medium Slateware slip and characteristic paste,others have finer paste texture and slip of an atypical-ly neutral gray color with darker streaks of PallidNeutral Gray and irregular areas of Pinkish Buff.Others have a slip ranging toward buff in color, whichtends to abrade from a medium-textured paste. It islikely that several wares are represented under thisloose rubric.

Decoration.-All this pottery is decorated in rough-ly applied red paint, minimum width of line 4 mm.,average 10 mm.; irregular rounded circles or roundedblobs of paint are common. This paint is a pigmentrather than a stain, and often shows relief against theglossy slip. Some painting on typical slateware slipshows bad abrasion, in other cases it seems to havebeen burnished into the slip. The paint adheres wellto the other slips. Pigment is of two types, a purplishred showing specular reflections, and an orange rednot showing sparkle. The paint on the gray slippedspecimens tends toward a purplish color with "bleached"gray or whitish areas. It shows brush rmarks as highridges, must have been very viscous when applied,and is quite hard after firing.

Shapes.-Large to small jars, small basins, lids,incense ladles, hemispheroid bowls.

Comparisons.-This ware, like Mexican MediumSlate and Redwares, illustrates the strong effects offoreign stylistic influence on the local ceramic indus-try. The technological task of separating the localslate and redwares from fine orange imports was easy,the untempered orange paste allowing macroscopicsorting. Red on Slateware has been inspired by main-land products, specimens of which probably lie in thecollections, as yet unseparated from the local ware.The origins of Mexican inspiration-if not of Mexicantrade-can be at least partly fixed. The Tula collectionsshow decoration closely similar to certain of the Redon Slateware specimens; the Xochicalco collectionscontain specimens markedly similar to others (seefig. 75, caption). The Tula specimens are of Toltecto Mazapan date and, since Tula almost completelylacks Aztec I (Culhuacan) style pottery, were probablymade before that period. At Chichen Itza, it has beenimpossible to demonstrate any chronological separa-tion between Red on Slateware and the X Fine Orange-ware which shows such close similarities to the AztecI decorative style. Either the Early Mexican subphaseat Chiche'n Itza may cover successively the Tula-Toltecand Aztec I periods, or X Fine Orange painted designstyle may have been in existence in Veracruz beforethe Culhuacan Aztec I pottery was made. But both ofthese hypotheses depend upon the assumption of rela-tively rapid and uninhibited spread of design stylesthrough the Valley of Mexico during this period. Thepresence of Red on Slateware style decoration at Xo-chicalco (where it seems to have been more used thanat either Chichen Itza or Tula) may aid in the datingof this puzzling and important site.

Chronological position. -Early Mexican substage.

Chichen (X type) Fine Orangeware(For additional description, see Brainerd, 1941, pp.163-172.)

Surface. -Very smooth to lustrous but never waxyin feel. Surfaces, particularly when unslipped, oftenshow a powdery abrasion. Most sherds bear an opaqueslip somewhat darker than the underlying paste (Vina-ceous Tawny to Terra Cotta to Smoke Gray). Slip hasat times collected in thickened, lustrous-surfacedareas on unexposed surfaces. Rattle legs are luted inplace and show no tenoning. Both legs and luting areso circular as to suggest kabal manufacture. Bandsor panels of either white, black, or purplish-red slipshowing sparkle are used in conjunction with inciseddesigns. The black appears like a stain, uneven indensity of color and not altering surface lustre. Thewhite is glossy, clear in color, opaque, laid on heavi-ly, subject to abrasion and occasional spalling. It mayhave been burnished. Encircling scratches of veryeven contour commonly occur on unslipped surfaces.

Paste.-Texture of fracture exceedingly fine,chinalike; no inclusions are visible to the naked eye,but fracture shows a faint, even graininess. Color isVinaceous Cinnamon or Orange Cinnamon, occasional-ly ranging to Neutral Gray.

Decoration.-Of two types, incised or plano-reliefand painted. Incised lines range .5-1 mm. wide andof about the same depth. They seem to have been en-graved into nearly dry clay; outline is crisp, withlittle or no burr. Drawing is precise, and curves areevenly rounded. Cut-away areas show the action of achisel-shaped tool with faintly and unevenly serratededge; a wooden tool would give this appearance. Whereslip was present, the incising was always done after

56


slipping and always penetrates the slip so that theorange paste shows good contrast with the slip color.Incised design occurs alone or in combination withpainted design, which may also occur alone. Painteddesign is done with the same black, stainlike paint thatis used for secondary slipped areas under incised de-sign. Drawing is done freely, often carelessly, in lines2-4 mm. wide. Occasionally, black-filled areas areused. For discussion on the styles of incised andpainted design, see Brainerd, 1941, Brainerd, 1953,and captions for figs. 76-84.

Shapes.-Medium jars, pyriform and cylindrical ves-sels with cascabel tripods and trumpet bases, flat andround-bottomed basal break bowls with cascabel tri-pods, hemispheroid bowls, effigies, vertical doughnutvessel, etc.

Comparisons. -The Chichen Itza collection of thisware is the largest in existence (1,053 sherds and 11whole vessels), unless the scattered collections fromIsla de Sacrificios, Veracruz, be considered as one.The predominant types from Isla de Sacrificios arenot found at Chichen Itza', and the types of the Chiche'nItza collections are not found predominant among thedecorated wares of any other site. On the other hand,there is no doubt that the Fine Orangeware of the Ear-ly Mexican subphase at Chichen Itza' was not made inYucata'n, and little doubt that it was made somewherein coastal Veracruz or in Tabasco. The absence fromChiche'n Itza of a wide variety of fine orangeware ves-sels known from Isla de Sacrificios, Cerro Montoso,and elsewhere, which have boldly drawn white designswith black and at times deep red outlining and second-ary design, suggests that this type of design may notbe contemporaneous with the Early Mexican substagedeposits at Chichen Itza. The proveniences of the fewsherds decorated in this color combination found atChichen Itza suggest that these vessels may date laterthan X type Fine Orange.

The marked similarity between X Fine Orangepainted design and Aztec I painted design has beendiscussed elsewhere, as have the distinctions betweenthe painted and incised design styles and the possibleorigins of both styles. (See captions of figs. 82-84.)The cross-dating of the Yucata'n and Mexican mainlandchronologies is well served by this ware. The prob-lems of the origins and interaction between the twostyles must be sought on the Mexican mainland. TheChich6n Fine Orange sample from Mayapan (see fig.28 and caption) may represent a late variant of thisware, and may be dateable to the Middle Mexican sub.stage.

Chronolo gical pOsition. - Ear ly Mexican sub stag e.

Plumbate(See Shepard, 1948, pp. 130-13 1, for Yucatan Plumbate.)

Miss A. 0. Shepard has described this ware so ex-haustively that no further description need be given.No Yucata'n Plumbate found is identified by Miss She-pard as of other than Tohil type, which is the typefound elsewhere associated with X type Fine Orange.One hundred and seventy-eight fragments come fromthe Chichen Itza' deposits, most of which are clearlyfrom Early Mexican substage deposits. Two fragmentscome from top levels at Uxmal, and one each comesfrom surface finds at Oxkintok and Mani.

Chronological position.-Early Mexican substage.

Coarse Slateware

Surface.-Slip with faint lustre, but with lumpy sur-face-a fine-grained slip over a paste bearing protrud-

ing temper particles. Color is grayish-white occasion-ally smudged to dark gray; a variant has a dull orangecolored slip (Orange Cinnamon or Pale Cinnamon Pink).The white slip may well be identical with the opaquewhite slip which is found on much Mexican MediumSlateware.

Paste. -Coarse texture on fracture. Usually distin-guishable from medium-textured paste of Florescentand Early Mexican wares by fracture as well as byrough surface. Temper determined by Miss Shepardon 36 specimens, all calcite. Paste color ranges grayto red-orange flecked with gray temper.

Decoration. -By black trickle paint. Paint occasion-ally disappears in spots, but never has the brownishrange of Florescent slateware paint. Orange variantof this ware usually does not bear trickle paint.

Shapes .-Large to medium jars, basins, basal breakbowls, hemispheroid bowls, grater bowls, cylindricaland pyriform vessels on pedestals, effigies.

Comparisons.-The distinguishing feature of thisware is the use of a coarse paste with a new undersur-face finishing technique. Wall thickness is greater thanthat of its closest relative and predecessor, MexicanMedium Slateware, but form and design show onlyslight changes. The effigy legs from bowls of this ware(fig. 92, e, I, 2) are hand-modeled and show no closesimilarity to any forms I know. Although our samplesof this ware are small for careful study, the lack ofobservable similarities to other areas, as well as theclose following of Early Mexidcan Slateware traditionin form and decoration, suggest a cultural continuumwith a minimum of outside influence during the sub-phase. Similarities to Huastec Black on White are notjudged close enough to imply contact when the distanceis considered.

Chronological position. -Middle Mexican subphasediagnostic. This ware occurs only with heavy mixtureof Early Mexican potteries in the Chich6n Itza collec-tions, and seems limited to surface deposits, save pos-sibly for the Mercado and Southeast Colonnade whereit sometimes occurs in about equal quantity in singlecollections with Medium Slateware. Another concentra-tion of Coarse Slateware is in the Temple of the WallPanels and adjacent buildings, including the extremeeast end of the Monjas complex. In all separate collec-tions from this area, however, Coarse Slateware neverequals Medium Slateware in frequency. Thus at Chi-che6n Itza we have as yet no certain evidence of build-ing activity as late as the Middle Mexican subphase,save perhaps in the dais of the Mercado. At Dzibilchal-tun, Coarse Slateware occurs in the higher levels oftrenches mixed with Regional Florescent wares, butwith very few wares of the Early and none of the LateMexican substage. At this site there must thereforehave been an occupation during which Coarse Slatewarewas the only slipped pottery-a "pure" Middle Mexicansubstage occupation.

At Mayapan, Coarse Slateware is low in all depos-its and in no collection does it equal Coarse Redwarein frequency. Thus we have evidence of either a Mid-dle Mexican or more likely a transitional Middle-LateMexican substage occupation at Mayapan; more exca-vation is needed to determine which.

Coarse Redware

Surface.-Ranges matte with irregularities throughsmooth to faintly lustrous. Slip fits well, little spallingor abrasion noted. Color a solid red with occasionaldarker smudged areas. (Ferruginaceous, VinaceousRufous to dark gray).

Paste.-Coarse fracture; gray to buff, with grayinclusions. Temper calcite.

57


Decoration. -Rare. Occasional under slip inciseddesigns, sometimes also covered by a white slip band(see fig. 85, a). These designs derive from X FineOrange. Other incised designs at Mayapan, some un-der slip and some over (figs. 26, c, d 27, _, i: 96, c),are less certain i/n origin. The incised designs on bowlskirts may derive from, or have influenced, similarbut less elaborate designs on Mayapan Fine Orange(cf. fig. 28, a). Effigy vessels show modeled decora-tion (figs. 25, .gZ 27, e., f, h). Effigy vessel legs aremold-pressed (figs. 27, g; 95, b2 96, .g) and also showMayapan Fine Orange similarities (cf. fig. 28, a, 37).Vestiges of blue paint are quite commonly found onbowls of this ware. This paint is not fired in, and isfound on exteriors, interiors, slipped and unslippedsurfaces.

Shapes.-Large to medium jars, basins, basal breakbowls, hemispheroid bowls, grater bowls.

Comparisons.-Coarse Redware is the potterywhich Vaillant called Red Lacquer ware. He arguesthat this ware shows evidence of a resurgence of Ma-ya influence at Chichen Itza (Vaillant, 1927, pp. 364-367). This hypothesis seems farfetched and not usefulat this point; although slipped wares with the generalcharacteristics of this ware are known in earlier Ma-ya pottery, we know of no Maya group of this periodor near it from which this ware is likely to have de-rived. The main direction of influence during the pe-riod seems to have been from the Campeche-Tabascocoastal region, which in turn was probably subject tostrong Nahua-Aztec influence.44

Evidence of foreign influence in Coarse Redware isnot extensive. The Early Mexican form repertory withits Mexican innovations is continued with few changes.Vertical sublabial incision, interior encircling grooves,and recessed jar bottoms are new, grater bowls losetheir rim incurve; skirts, some notched and engraved,are attached to basal break bowls. Zo6morphic andanthropomorphic legs are used on bowls. Small, trun-cated, conical bowl legs are in use; some are hollowwith two exteriorly placed holes, one above the other.Skirts and legs probably come from Tabasco; the oth-ers cannot be traced. The terraced slab leg, rare inthis ware (fig. 94, h, 32) and present at Cintla (fig.103, L 7, .2), shows Valley of Mexico connections,where it is reported for Aztec III (Brenner, 1931, pl.A).

Chronological position. -Late Mexican diagnostic.

Mayapan Fine Orange Ware

Surface. -Smooth; appears to bear a very thin slip,of color and texture very close to that of the paste.Color Vinaceous Cinnamon or Orange Cinnamon.

One fragment is covered by a purplish-red slipwith sparkle like that of specular hematite. Some bearwhite slip all over, or encircling bands, some whiteand some black. These white and black slips are verysimilar to those on Chichen Itza Fine Orange. Evi-dence of forming shows in marked irregularity of con-

tour (see fig. 28, a), and these irregular undulationsof the surface go very evenly around the vessel insuch a way as to resemble closely those made by use

of the modern kabal. Skirts are also extremely circu-lar, and appear to have been formed from the claymass of the vessel rather than by applique. Rattlelegs are luted without tenon.

Paste.-Fine texture; looks untempered through a

10-power glass, but has not been checked microscopi-cally.

Decoration.-Incising in lines averaging .5 mm. in

width and of about half that depth, with semicircular

cross section. Most designs show no cutout areas andare extremely simple and limited to vessel exteriors-save for one bowl with an incised grater pattern onthe floor. Incising through black and white slip bandswas noted. Painted decoration is in black with the samestainlike quality of that on Chiche6n Itza Fine Orange.Designs again are very simple, with banding lines thecommonest feature. The hemispheroid bowl rims withwhite slip, or slip band, and black, encircling linespainted over it may deserve classification as a sepa-rate type, but the sample is small. Effigy legs are theonly type of basal support found. They bear mold-pressed fronts, with possibly hand-modeled back por-tions. Skirts are notched, as are those of the accom-panying Coarse Redware.

Shapes -Bowls with sharp basal break, skirtedbowls with very rounded basal break, either or bothwith effigy legs; two shapes of spheroid bowls, graterbowl, pyriform vessels, constricted cylindri cal-neckedjars.

There is a strong suggestion that several types,each characterized by the almost exclusive presenceof several traits of form and decorative style, are in-cluded in this group. Since there is no evidence atpresent that all these several types were made by thesame potters, or even potters having close culturalinterrelations, Mayapan Fine Orange cannot at presenttechnically satisfy the implications of the classificationof a ware, although it has, for convenience, been giventhat classification here. The above remarks applyequally to Puuc and Chichen Fine Orange, and to Dzi-bilchaltun Fine Grayware, none of which have been in-vestigated in their original cultural context at sitesnear their center of manufacture.

The types which seem apparent in our sample are:(type A) Bowls with rounded basal break and undulatingwall profiles, with bent and thickened lips. Many bearskirts and probably most bear effigy legs. Decorationon exterior wall of lightly incised vertical lines witha horizontal encircling incision at wall-skirt junction.Examples are figs. 28, a, 1, _-_l 23-26 30-37 39-4l.(type B) Hemispheroid bowls with out-bent lips; evenwall profile and black-painted band designs on interiorand exterior. Examples are fig. 28, ,c,, 9-10. Otherpainted sherds with form of type A, figs. 28, a, 22,

, 28, suggest culture connection between the twogroups. (type C) Deep, hemispheroid bowls withblack-painted, horizontal lines on a white slip band(fig. 28, b, 1-3). In the sample available, this typeshows no definite traits in common with the other two.The remaining sherds are indeterminate; 28, c, 18may be related to type B by reason of the painted de-sign; several may belong to Chichen Fine Orange onbasis of their plano-relief decoration.

Comparisons. -Mayapan Fine Orange stands inmuch the same relationship to the local wares as Chi-chen Fine Orange did during the Toltec period. It wasimported, and was copied to some extent in local wares;the notched skirt and effigy legs, but not the designsor rim shapes, are followed on Coarse Redware. Ma-yapan Fine Orange, like Chiche'n Fine Orange, containstypes which were not necessarily made by the samecultural group. The relationship of Mayapan FineOrange to Chichen Fine Orange may be to a degreereconstructed. First, although the paste, and probablythe paints, of the two groups of wares are macroscopi-cally identical, forms and styles seem to be distin-guishable enough so that complete vessels of the twogroups would show no intergrading of either form ordesign. Thus a common tradition of techniques andmaterials seems to link the two groups (though thisstatement is subject to the results of technical analy-

58


sis), while the styles (form and design) of the twogroups are quite distinct. Distribution of the twogroups is, at least at our present level of knowledge,equally distinct. The center of occurrence of Chiche'nFine Orange seems to be in central Veracruz, whileMayapan Fine Orange seems to center on the Cam-peche coast with related wares to the west on the Ta-basco coast.45

Chichen Fine Orange occurs in the minority, butmixed with Mayapan Fine Orangeware, in the Mayapandeposits. Although the present samples do not allow a

stratigraphic separation of the two wares, it is prob-able that the Chichen-style ware ranges earlier indate and that it belongs with the Middle Mexican ware

assemblage. A suggestion of time overlap between thewares is found in two of Seler's Uaxac Canal speci-mens (fig. 89, s, .), one each of Chichen and MayapanFine Orange, which, if they were found in a singlegrave (and this is likely though not certain), would

show contemporaneity between Chiche'n and MayapanFine Orangewares. There is also some evidence thatthese later samples, now included under Chich6n FineOrange, can be stylistically separated from those ofthe Early Mexican subphase (see fig. 28, caption).

A comparative estimate of direction and amount ofceramic trade at Chiche6n Itza and Mayapan may be ofinterest. Fine orange is by all odds the major foreigntradeware at both sites. At Chichen Itza the fine orange-ware from Veracruz runs about .5 per cent of totalsherd count (body plus rims); at Mayapan the fineorangeware from the Campeche coast runs .43 per

cent of total sherd count. These figures are of thesame magnitude; the amount of ceramic trade was

roughly the same. Chich6n Fine Orange seems to doc-ument a long-distance coastal trade and Mayapan FineOrange a shorter haul, a good part of it overland.46

Chronological position. -Late Mexican substage.

59

IV. CHRONOLOGICAL ANALYSIS

Evidences of chronologic sequence have come tolight at odd times and in irregular order during thecourse of this survey, and nearly every evidence ofsequence after its first recognition was confirmed byadditional evidence from other deposits. Aid in theplacement of the earlier ceramic assemblages wasgiven by interlinking through tradewares and similari-ties to the well-dated and documented Uaxactun se-quence to the south. Although all sequence evidencewill of necessity not be given here, an effort will bemade to demonstrate convincingly the placement ofthe major pottery wares and the basis on which thegeneralized Yucatan stages have been set up.

THEORY

Time periods in ceramic studies are most common-ly defined from characteristic assemblages of potterydescribed by ware and form repertory; the best stud-ies give detail on variation and on relative frequencyof wares, shapes, and decoration. These assemblagesnormally consist of pottery found in direct association,or of a series of collections which possess so manyattributes in common that they are assumed to havebeen made at a single time and place. Such assem-blages have generally been called phases in the Mayaarea, although there has been a tendency to extend theuse of the original phase name to all assemblageswithin a considerable area if they are believed to beroughly contemporaneous with the type collectionsfrom which the phase was originally defined.

When time changes have been recognized at a siteor in an area, the phases have been sequenced andfixed in absolute dating as far as possible from avail-able evidence.

Time is a continuous variable; the phases describedthus represent selected points or spans in time. Thearchaeologist most commonly chooses his points orspans on the basis of large collections which, for var-ious reasons, he believes to be "pure"-that is, tohave been deposited within a relatively short timespan and to be unmixed with material of other age.Such collections are often supplemented by use of ad-ditional specimens placed with the dated lots on thebasis of similarity. Phase descriptions thus may usu-ally be most accurately described as cultural crosssections taken at points along the time axis, at a sin-gle geographic location.

The fact that period or phase descriptions are merecross sections of a continuous body of material chang-ing through time has often been ignored in analysisand even more commonly ignored in later exposition.The phases or periods described list blocks of waresand forms, and the assumption is sometimes at leasttacitly made that changes take place simultaneously inall wares and forms at the same time, at a sort of in-terphase boundary line. A corollary of such an assump-tion is the sudden replacement of one culture by anoth-er. The acceptance of such culture replacement withoutsufficient docurnentation has allowed such errors asthe incorrect reconstruction of a "Pueblo Invasion" inthe Southwest United States47 and, it may be suspected,many of the hypotheses of migration and displacementof human groups which have been made in American,European, and Near Eastern prehistory.48

[60]

Ceramic horizon descriptions are certainly the bestmanner of obtaining an idea of the cultural repertoryof a group, and can be used as basic materials forphase ethnography. In Mesoamerican archaeology theirmisinterpretation has been guarded against by someworkers through emphasis on their discontinuous na-ture through time. Ceramic assemblages are common-ly designated as phases and given names which do notcommit them to a fixed, closed, chronologic succes-sion, as do numbers as used to define horizons. If usedintelligently, this system can allow the introduction ofnew phases intermediate in time, and even permits thetransposition of the named phases in a chronology with-out dislocation of the nomenclature. Freedom is alsogiven for the definition of regional variations by thenaming of phases from new excavations, these phasesbeing placed by chronologic equivalence to those al-ready known.

In addition to the description of phases, there hasbeen in the last few years a tendency toward the useof general inclusive names for the major time divi-sions in Mesoamerica. Kidder (Kidder, Jennings, andShook, 1946) and Armillas (1948) have named such di-visions for Mesoamerica. Such terms are of courseof great value in generalized comparative work.

NOMENCLATURE IN THIS REPORT

Because of the relatively large area and long timeperiod covered by this survey, a generalized nomen-clature is essential. The following system has beenchosen as conflicting least with those currently in usein neighboring areas.

Yucatan Yucata'nStages in this report Thompson.

1945Early FormativeMiddle Formative )FormativeLate FormativeEarly Regional )IniesMiddle Regional )Inritial

Late Regional )earlyFlorescent

)lateEarly MexicanMiddle MexicanLate MexicanInter regnum

)ColonialPost-Conquest

)Modern

Mexidcan)Mexican)Absorption

Central Maya Area(Uaxactun)equivalents

Mamom

Chicanel, San Jo s6 I

)Tzakol, Tepeu 1, 2,)and early 3; San)Jose II-V; Holmul)I-VTepeu late 3; SanJose' late V

The nomenclature given the Yucatan chronology issuperficially quite different from the commonly usedand at present quite adequate generalized Mesoameri-can stage sequences of pre-Classic or FormativeClassic or Initial Series, post-Classic or Militaristicor Expansionist. The stage nomenclature presentedhere is meant for Yucatan alone. It is based dispro-portionately upon ceramic history, but there is someevidence that architectural history in Yucatan alsofits this outline at most points. The question of the fi-nal usefulness of this terminology in general descrip-tions of Yucatecan culture history must, however,


await future testing. The time span has been segmentedto conform as far as possible with major ceramic andarchitectural changes; the stages have been renamedseveral times during the study in an effort to avoidconfusion with the ever-increasing welter of terminol-ogy in Mesoamerican archaeology.

The Formative stage in Yucatan should cause noconfusion; it roughly equates with the Formative orpre-Classic stage elsewhere. The Central Maya Clas-sic or Initial Series stage has been subdivided in theYucatan sequence because of the striking differencesbetween the monochrome- and slateware-bearing sitesof this stage. What descriptions are available suggestthat the architecture of this stage in Yucatan presentsa dichotomy which parallels the ceramic evidence.Monochrome-bearing deposits are always earlier thanslateware in all parts of Yucata'n, but there is consid-erable evidence that slateware began to be made ear-lier in the Puuc region than on the northern Yucata'nplain. Thus the Yucatecan Regional (monochrome-bear-ing) and Florescent (slateware-bearing) deposits can-not be sequenced along chronological lines alone; thereis evidence of chronological overlap and regional dif-ference between them. The beginnings of the Regionalstage in Yucatan can be set as considerably antedatingthe beginning of the Maya Initial Series period on evi-dence presented elsewhere in this report. RegionalIncised and Punctate Dichromes, with their associatedFlaky Redware, antedate the Tzakol-phase pottery ofUaxactun, the earliest use of which in turn is believedto antedate the first Initial Series dates.

The Mexican stage may have begun later in Yuca-tan than the end of the Initial Series dates, as has beensuggested by Thompson (1945). The 'Early" subdivi-sion is based on the ceramics of the Toltec buildingsat Chiche'n Itza. The Late Mexican substage is basedon the ceramics which preponderate in the Mayapancollections. The Middle Mexican substage is definedby a distinctive ceramic assemblage (Coarse Slate-ware is the diagnostic), which is early at Mayapan,late at Toltec Chiche'n Itza', and found in nearly puredeposits at Dzibilchaltun. The architecture of this sub-stage is unknown.

The time period between the historically recorded"fall" of Mayapan in 1441 A.D. and the Spanish Con-quest a century later was one of internal disunion inYucatan. Of the three major rulers of the Confederacyof Mayapan, one, the Xiu ruler, is reported to havesettled in Mani. This town became a major Spanishecclesiastical center in early Colonial times, doubt-less favored by the Spanish because of the aid giventhem by the Xiu family. Much of the pottery found atMani was dateable as Colonial by tradeware; some ofit may come from the century of 'interregnum" beforethe Conque st.

The problem of nomenclature of the various depos-its from the individual sites is difficult to handle bythe system of named phases described above. Themethod here chosen is not meant as a tacit criticismnor indeed as an improvement of that method. The useof the present method has been made advisable by thenature of the material handled. Our ceramics comefrom some 20 sites, several of which show evidenceof occupation during a major part of the chronologicspan. Some of these sites are at a considerable dis-tance from others, and certainly show regional dif-ferences. In other instances, the collections of certainhorizons in various sites are so small as to make ituncertain whether further excavation would indicatethe naming of a new phase. In fact, the interpretationof phases in the description of regional culture is farfrom uniform in Mesoamerica. If the collections were

conscientiously named by phase, there would be sucha number that the author, let alone the readers, couldnot carry them easily in mind.

The system used has been to name the assemblagesby site name modified by either a stage name or, whenthat might be equivocal (as in cases where the positionof the collections in the Yucatan chronology is notknown exactly), by a number giving the early-lateposition of that phase in the chronological sequencedefined at the site. Since, except in the case of Chiche'nItza', the excavations were merely exploratory trenches,there seems no reason why these phase names andnumbers should be perpetuated if extensive excavationsare later made at any of these sites.

The deccription and sequencing of pottery assem-blages is of primary value to the service function ofthe survey; such definition allows the relatively easydating of other archaeological materials by means ofaccompanying pottery. To this purpose the unusuallylarge and voluminously annotated corpus of pottery il-lustrations is provided, in addition to the ware descrip-tions and notes as to their chronologic placing. Thedetailed tabulations, averaging sixty or more sherdcategories per site, are not here published, but havebeen preserved and are available for later detailedanalysis. Ware percentages for most assemblages canbe read from the stratigraphic diagrams which accom-pany the next sections of the report.

The stratigraphic evidence for the Yucatan chronol-ogy has been supplemented by additional kinds of evi-dence. The cross-finds with Uaxactun have in manycases substantiated stratigraphic placing and in othercases have actually given clues to the sequence. Theplacing of two or three of the phases depends uponseriational evidence, as does the unraveling of muchof the detail in the copiously represented Florescentand Early Mexican periods.49

Seriations in the sense used here include all sequen-tial placements made without stratigraphic evidence.All of these depend for their mechanism on the theorythat the form of the varied products of cultural activitygradually changes as does the culture itself.

If we can correctly place a group of products ofculture in order of change, and determine the direc-tions of time in the sequence, we may attain the sameresult as if our series were found in stratigraphicdeposition. The difference should be stressed betweensequencing or seriation by the above procedure, andthat produced by chronological placement on an apriori assumption that material culture increases incomplexity, conventionalization, or virtuosity throughtime. We know that material culture does not always'improve" through time, but that it does change. Ourseriations are based on change, not improvement, asa criterion.

In seriational analysis, the best criterion for theisolation of the results of factors other than time,lies in the fact that time is a unidimensional factoracting upon all the facets of culture. If two or moreobjectively defined criteria in pottery (A, B, and C)gradually change during a series of time increments(1 through 10), we will never find a pottery vesselcharacterized by the formula Al, B5, C8, but the var-iants of A B C will always be those of correspondingor nearly corresponding time periods (as A3, B3-4,C2-5). If the variations in the three criteria, A, B,and C, are linked to regional rather than to time dif-ferences, we may expect to find the coefficients of A,B, and C to vary more widely in their combinations.50The placement of the specimens on fig. 63 as inter-mediate between Incised Dichrome and Tzakol-stylepottery was made by this method.

61


The statistical work which has been done on theYucatan collections (Brainerd, 1951) has been directedtoward the determination of chronology within the timedivisions used in this report. This attempt at refine-ment has been dropped for a time in favor of the prep-aration of this outline of the total ceramic sequence.

STRATIGRAPHY

The chronological placing of the ceramic assem-blages of the Yucata'n survey is supported at mostpoints by multiple lines of evidence. The first evidencegained for most stages was that of stratigraphy. Thisevidence will be outlined in the order in wbhch excava-tions were made, and the general outlines of the chro-nology determined. Fuller discussion of much of thismaterial is given in the sections devoted to the sepa-rate sites.

Before the beginning of the 1942 season's excava-tions, the validity of the previously determined se-quence from Puuc sites to Toltec Chiche'n Itza' hadbeen checked from Roberts' material. Strong similar-ity among the Labna, Sayil, Kabah, Sabacche, and Ux-mal collections was apparent from their typologicalclassifications, which I made independently from therim collections gathered by Roberts. The Holactuncollections seemed a regional variant of the same pe-riod, also containing Formative deposits dateable fromthose of Uaxactun. The collections from Chiche'n Itzashowed a marked preponderance of wares and shapesnot present in the Puuc collections, and some ratheruncertain suggestions of stratigraphy. The CoarseRedware and figurine incensarios in these collectionswere limited definitely to surface and postoccupationalcollections (as noted by Vaillant, 1927, pp. 363-364).I identified Coarse Slateware, unrecognized by Robertsin the Chiche'n collections, and tentatively placed it be-tween the Toltec and Coarse Redware horizons mainlyon stylistic grounds, but also on its presence in build-ings judged to be late, and in superficial collections.The non-Puuc pottery in the majority of the Chiche'nItza collections was tentatively placed as of the Toltecperiod, anchored there by the associated distinctivetradewares. The Coba' collections and parts of the Ya-xuna collection made by Roberts were recognized byRobert Smith in 1940 as of Tepeu 1 date, thus fittingthe Coba' stela dates. Architectural evidence of veneer

masonry, rare at Yaxuna, better preserved than therest of the site, and stratigraphically on top, suggestedthat the slateware found there dates later than the pre-dominant Tepeu 1-like wares. The early Oxkintok col-lections which I dug in 1940 were recognized by RobertSmith as having Tzakol similarities, thus matching indate the 9.2.0.0.0 lintel which had recently been foundat that site; the stratigraphically later collectionsfrom Oxkintok were recognized as markedly similarto those from the Puuc sites, particularly to thosefrom nearby Uxmal.

Thus by the end of 1940, seven horizons could beisolated and sequenced with some degree of certainty;three of these, the Puuc, Toltec Chichen, and CoarseRedware, corresponded to previous determinations byVaillant, supported by Roberts. Various sequencesand conclusions based on architecture also addedweight to some of these placements. I had tentativelyplaced the Coarse Slateware in the sequence; the threeearlier phases were anchored into the Peten sequence,but of these only Oxkintok was stratigraphically placedas yet in Yucatan. The Incised Dichrome collectionfrom Balam Canche, the Guest House Cenote collectionfrom Chiche'n Itza, and certain other collections, re-

mained distinctive and consequently intriguing but withno real clues as to placement.

The choice of Yaxuna in 1942 as the initial site forexcavation was made on the basis of Roberts' collec-tions from there. Four sequent periods were recog-nized from the excavations, placed in reasonably cer-tain stratigraphic order which was afforded furthersupport by Peten crossties determined by Robert Smith,who examined the pottery that season. Charts 2, 3, and4 demonstrate these sequences in three deposits. Form-ative Monochrome predominates in the bottoms of thethree deposits, followed by Flaky Redware in charts2 and 3, and by Regional Medium Redware in chart 4.Medium Slateware preponderates in charts 2 and 3surface cuts, and is prominent in chart 4. This allowsa sequencing from bottom up of Formative Monochrome,Regional Medium Redware, and slatewares, with FlakyRedware probably below Regional Medium Redware(note chart 4), possibly above it, but still below theslatewares. The crossties with sherds identified byR. E. Smith as Peten-like are shown to the right ofcharts 3 and 4. Chicanel seems roughly contempora-neous with Formative Monochrome, Tzakol with FlakyRedware, Tepeu 1 with Regional Medium Redware, andthe slateware horizon shows a decline in Peten trade-wares. These Peten equivalents of course increasedthe likelihood of our poorly documented stratigraphicplacement of Flaky Redware, as did the Tzakol Petentradewares found in a nearly pure Flaky Redwarephase collection in the Yaxuna Cenote. The presenceof Incised Dichrome in the Yaxuna Cenote Flaky Red-ware collections also fixed the Balam Canche IncisedDichrome collection in the sequence; marked similari-ties between the Regional Medium Redware assemblageand the Coba' collections, coupled with the Coba'-Petenplacement and stelae dates, fitted the main Coba' col-lections to this sequence and gave check dates in theMaya calendar.

Dzibilchaltun was next dug, and instead of the latedeposits expected because of the Colonial church, aphase transitional into the wares of the Puuc sites wasdiscovered, with obviously disjunctive, and thus easilyisolated, Coarse Slateware admixture. Charts 5, 6,and 7 summarize the stratigraphy. The Formativeand Regional wares are very lightly represented here;charts 5 and 7 carry little authority due to small sam-ple size, but are in agreement with the sequence. Chart6 shows the trends of the main deposits, with the slate-wares increasing toward the top. This evidence, coupledwith the transition noted at Chich6n Itza' between Flo-rescent and Early Mexican, suggested that these Dzi-bilchaltun deposits probably somewhat antedated thePuuc assemblages, a hypothesis confirmed by workat Acanceh and elsewhere.

Acanceh gave hopes of more pre-Florescent mate-rial through the presence of the probably early stuccofagade and the large stucco masks on the northernpyramid. The stratified Regional and Regional-Flores-cent deposits (charts 8, 9, 10, 11, 12, 13) gave moredetailed information than any others encountered. Thedeposits, segr egated by superimposed floor s, supportthe Yaxuna Cenote evidence that Flaky Redware wasnot made concurrently with slatewares (see chart 8,cuts f and g), and soundly establish it as earlier thanCoarse and Medium Regional Redwares. RegionalMedium and Coarse Redware are invariably associatedwith Medium Slateware in all collections of significantsize. That this mixing is due to later deposition seems

unlikely in view of the large numbers of trenches, bothhere and at Yaxuna, where this condition holds con-

stantly true. That slateware was made early in theChenes area is suggested by the 1949 collections, as

2


yet incompletely analyzed. The Acanceh deposits there,.fore document an increasing use of slateware, displac-ing the Regional wares of this area, a situation alsoprobably true at Yaxuna. The relative extent to whichregional and chronological factors are at work throughthe various sites of this Regional-Florescent sequenceis of major importance in placing the northwesternand eastern Yucata'n sites chronologically in relation-ship to the Puuc sites, and needs further excavationto clarify. Coba, where pure Regional Redware depos-its were found, may lie beyond the geographic rangeof the earlier slatewares.

The chronologically sequent position of RegionalRedware and Red on Thin Gray at Acanceh, and theclose stylistic resemblance of their jar forms, placesthe main Dzibilchaltun occupation as quite clearly laterthan that of Yaxuna III, Coba, and other occupationsstrong in Regional Coarse and Medium Redwares. Theearly Oxkintok collections are with considerable cer-tainty placed earlier than the Regional Coarse andMedium Redware occupations by the absence of slate-ware in Regional Oxkintok deposits, by various stylis-tic criteria, by the Peten ceramic correspondences,and by the Oxkintok 9.2.0. 0. 0 Initial Serie s date.A good representation of Coarse Slateware at Acan-

ceh confirms again its stratigraphic position aboveRegional-Flor e scent deposits. Early Mexican potteryis too scarce, unfortunately, to allow a check by itsplacing here.

The presence of Flaky Redware in late Formativestage vessel shapes in the lowest strata at Acancehwith Regional stage forms of the same ware, coupledwith the appearance of a Formative Monochrome hori-zon at nearby Mani, suggests that the two wares maycharacterize successive horizons of the late Formativesubphase. This possibility is supported by certain dif-ferences in rim profiles between Formative Mono-chrome and Formative Flaky Redware. The FormativeFlaky Redware vessel shapes grade into those of theRegional stage Flaky Redware.

Mayapan showed well-documented stratigraphythrough two sequences: (1) Coarse Redware followsCoarse Slateware; and (2) Figurine Incensarios growprogressively more frequent toward the close of theCoarse Redware occupation. Charts 14, 15, 16, and 17show both these trends. Maximum frequency of CoarseSlateware found at Mayapan is about 35 per cent inlevel g of trench 7 (chart 14), lying under a plaza floor.It is questionable whether Figurine Incensarios were

in use during the earlier part of the Coarse Redwareoccupation; several large collections of that phasewere entirely or nearly free of them.

The excavations at Mani yield evidence of long con-

tinued occupation. The trenches dug in the cenote near

the center of town produced the earliest deposits yetto come from Yucatan. Charts 18 and 19 show its posi-tion better than the other sequences. As may be seen,the strata are mixed, probably due to construction ac-

tivities during the building of the adjacent Colonialstage ramp. Burnished Design Ware preceded all otherceramic assemblages here. Following this Early Form-ative deposit come good Late Formative samples, thenRegional, Florescent with a few Early Mexican forms,then Middle Mexican, very lightly represented LateMexican, followed by an enormous concentration ofpost-Mayapan material. Since the Xius, one of the rul-ing groups of Mayapan, are said to have settled atMani after 1440 A.D., the period following this date isof particular interest. The post-Mayapan pottery fromthe Mani Cenote is closely comparable with that of theMani Monastery, the end date of which is early 19thcentury; none of the Mayapan specialized or trade

wares was found in the deposits. The placement of thelate Mani pottery thus may well be post-Conquest. Itis possible, however, that some of this pottery was

made during the pre-Conquest interregnum.The chronologic placements at Chiche'n Itza are not

shown in the graphs, and thus need explanation. CoarseRedware and Figurine Incensarios are in general foundsuperficially in debris of fallen Florescent and EarlyMexican buildings. Coarse Slateware is found in theSoutheast Colonnade and Monjas East Court, which byarchitectural stratigraphy date later than adjacent Ear-ly Mexican buildings on evidence detailed in J. S. Bolle'sunpublished report of the Monjas excavations. Severaltrenches dug by Roberts at Chiche'n Itza show Flores-cent styles of Medium Slateware giving place to EarlyMexican styles. Formative Monochromes give way toFlorescent Slateware in a trench at the southwest cor-

ner of the Caracol. The Puuc collections all have Me-dium Slateware predominant. At Kabah, however, a

small showing of Formative Monochromes tends tohold to low trench levels.

Figure 21 shows the more diagnostic wares ar-

ranged in a sequence, top to bottom. Wares which ap..peared through more than one subphase are bracketed.This early-late sequence is, through most of its length,supported by stratigraphy. Arrows indicate stratigraph-ic evidences of superpositions of wares. For accurateplacement of a ware in sequence, the graph shouldshow solid arrows connecting it with its two immedi-ately preceding and sequent wares. The sequence pairswhich do not show such stratigraphic documentationare: (a) Coarse Slateware above Medium Slateware;this placement is documented by architectural stratig-raphy described above. (b) Coarse and Medium Red-wares above Oxkintok Monochrome; this sequence hasbeen assumed from Initial Series dates at Coba andOxkintok where these wares are found, as well as byPeten ceramic crossties with both wares, and by sty-listic sequences in forms. (c) Oxkintok Monochromeabove Flaky Redware; form repertory of OxkintokMonochrome shows similarities to that of Coarse andMedium Redwares and to Medium Slateware, whereasFlaky Redware does not and therefore must be furtherremoved in time-a stylistic placement. (d) RegionalFlaky Redwares above Formative Flaky Redwares-on stylistic evidence again; the closeness of RegionalFlaky Redware bowls to early Tzakol bowls, and evi-dences of transition in Incised Dichrome, are discussedelsewhere. Also, forms of Regional Flaky Redware are

similar to those of other Regional wares, whereasFormative Flaky Redware forms are nearly identicalwith the Formative Monochrome repertory. (e) Forma-tive Flaky Redwares above Formative Monochrome-a stylistic placement based on the fact that Flaky Red-ware forms bridge the Formative-Regional stagetransition, whereas monochrome forms do not. Thisplacement was further strengthened in 1949 (see Brai-nerd, 1949) by the discovery of a new substage lyingbetween Burnished Design Ware and the FormativeMonochromes shown here. The collections of this Mid-dle Formative substage bear a preponderant amountof monochrome very similar in ware characteristicsto the Formative Monochromes described in this re-

port.In the evaluation of the evidence for chronology

here presented, the variable of regional differenceshould be held constantly in mind. Although the strati-graphically and stylistically determined trends pre-sented seem valid, the 1949 excavations have borneout previous suspicions that slateware was made dur-ing most of the Regional phase in parts of the area.

This finding suggests that other regional differences

63


may have complicated the chronological picture; ourceramic samples are thinly spread over both time andspace. There is considerable evidence that form anddecoration are more sensitive through time than areware changes, and forms have given us our majorchronological clues for the above placements. Tradepottery has also provided checks and new clues in thedevelopment of our chronology.

CERAMIC EVIDENCE BEARING ON THE

MAYA- CHRISTIAN CALENDRIC

CORRELATION

The following comments on the correlation prob-lem can best be understood by reference to chart 22,of which this section is an explanation.

Columns I-V demonstrate the manner in whichchronicles events in Yucatan and Maya Initial Seriesdates enmesh, using the three most widely acceptedMaya-Christian correlations. Column I represents the12.9.0.0.0 correlation, column I the 11.3.0.0.0 corre-lation, column III the 11.16.0.0.0 correlation. It willbe noted that 1539 A.D. in each of these correlationsfalls on a Maya katun ending on the day 13 Ahau, thespecification given by early post-Conquest sources,and that katuns 13 Ahau, as well as all other katunending coefficients given on the righthand side of col-umn V, fall 256 1/4 years apart. The assumption thatthe Maya and Christian calendars correlate in thiscyclic manner depends upon coincidences singularenough to be convincing, and has remained almost un-challenged. The three positions shown, as will be seen,are those which most nearly fit several groups of sep-arately derived evidence.

Points A, B, and C shown in each of the first threecolumns match katuns 8, 8, 4 respectively on the Ma-ya calendar, column VI. Point A, the fall of Mayapan,which occurred within 100 years of the Conquest, isunquestioned in its placement. Points C and B indicaterespectively the dates at which the Itza' arrived andabandoned Chiche'n Itza, according to the Roys-Thomp-son scheme derived from the interpretation of theChilam Balams and other chronicles and from certaincarved dates at Chichen Itza.51

A reexamination of the evidence for the length ofthe time spans A-B, B-C in Yucata'n is in order. B-C,the span of the Chiche'n Itza occupation, is quite uni-formly fixed by the chronicles as about 200 years, andthis span also checks well with the katun ending dates.Landa and Herrera give the Mayapan span (A-B) as

500 years, which is made improbable by several fac-tors, perhaps the most obvious of which is the small-ness and decadent construction of the site. Variousother references suggest that the 500 years may applyto the duration of the League of Mayapan, out of whichtime span Chiche'n ruled alone for the first 200 years.The Chilam Balam of Tizimin gives 280 years for thejoint rule of the lords of Uxmal, Mayapan, and ChichenItza.52 The ceramic evidence of a period (Middle Mex-ican substage) postdating the architecture of ChichenItza and found sparingly and only in the lower levels ofMayapan, suggests a possibility of a longer span forA-C. In the author's opinion, the consensus of evidencefavors the time for A-B as shown, about 260 years.

Assuming that points 1539 A.D., A, B, and C, arein their correct order and span, the choice between thethree Maya-Christian correlations depends upon therelationship between points C and D; C is computedfrom the Christian calendar and D is placed in theMaya Initial Series.

That D preceded C we know ceramically. Thereforethe 11.3.0.0.0 correlation, column II, is ruled out ifthe range of our period A-C is allowed as correct.Remaining for consideration are the 11.6.0.0.0 and12.9. 0.0.0 correlations.

Column VII shows in summary form the Yucatanceramic stages and some principal deposits, and theirplacement on the Maya calendar. These placementsare accurate within quite close limits between 9.0.0.0.0and 9.16.0.0.0 on the evidence both of Maya inscriptionson Yucatan sites and of correspondences through tradewith the well-anchored Uaxactun ceramic sequence(lower part of column Vll).53 The placement of the Mex-ican ceramic stage has here been put in harmony withcolumn III, the 11.16.0.0.0 correlation, but the essen-tial criteria for its length have been drawn from thechronicled sources shown as points A, B, and C. Thethree Mexican ceramic substages shown are differen-tiated by major and easily recognizable differences inthe preponderant slipped wares, coupled with manyother less striking ceramic differences. Their timesequence is well documented as shown in the chart,and evidences of the time duration of the substageswhich show archaeologically do not dispute the chartedplacements. Thus the Mexican ceramic stage as herediagrammed is dated from the Christian calendar, andthe solution of Maya-Christian calendric correlation,if all above assumptions are acceptable, may be re-duced to a judgment of the length of the latter part ofthe Florescent ceramic stage (span C-D on the chart).

The qualification should be clearly stated that ce-

ramics cannot yield an accurate estimate of absolutetime duration. The ceramic criteria of stylistic changeand quantity of material are subject to several unknownvariables. The stylistic change during the period inquestion is very small relative to that of periods inthe sequence of known duration, and the relativelyslight stylistic changes in architecture noted by Pol-lock in the Puuc area bolster this determination. Thequantity of ceramic and architectural material belong-ing in this period is very large, particularly in thePuuc region. The predominant slipped ware of the Flo-rescent stage is nearly identical with that of the ToltecChichen (Early) segment of the Mexican stage, the ce-ramic differences being mainly in shape and decorationof vessels, and in the trade wares. The architecturalchanges between the two corresponding periods at Chi-ch6n Itza' are to some degree analogous in that construc-tion methods remain quite similar, but form and deco-ration change.

The addition of 260 years to the later end of theFlorescent stage, or such an addition to the Mexicanstage, would be necessary if the 12.9.0.0.0 correlationis to be accepted. Such an addition would make the Flo-rescent stage double to quadruple the length of thedated ceramic stages of the sequence, and these suc-

cessive stages, the author believes, are defined byvery roughly equal degrees of ceramic changes. Thiscriterion, however, is admittedly a dubious one, as

witness the extreme conservatism of Yucatan ceram-ics over the last 500 years. The documentation of thetime span of the Mexican stage, both in Yucata'n andon the Mexidcan mainland, does not easily allow a 260-year addition. Therefore the author favors the11.16.0.0.0 correlation over the 12.9.0.0.0 correlationon ceramic grounds, but favors the 12.9.0.0.0 correla-tion over the 11.3.0.0.0.

Other evidences also favor the 11.16.0.0.0 correla-tion. Thompson's (1937) readings of the Chiche'n Itzashort dates and of several dates in the Puuc area

match this sequence nicely, except for two late datereadings. Beyer's readings for several of these dis-

64


puted dates, however, favor the 12.9.0.0.0 correlation.A further argument for the 11.16.0.0.0 correlation isDr. Kidder's placement (1946) of Teotihuacan III atabout 9.5.0.0.0 (Tzakol phase), through trade potteryat Kaminaljuyu. Only one well-marked, distinctiveceramic horizon intervenes between Teotihuacan IIIand Mayapan-Tula in the Valley of Mexico (Coyotlatel-co-Azcapotzalco-Teotihuacan IV-V). Since Mayapan-Tula crossties with Toltec Chichen Itza' so closely inboth architecture and ceramics, the 400 years (rough-ly 500-900 A.D.) indicated by the 11.16.0.0.0 correla-tion seems a reasonable estimate of time for this pe-riod, while the 12.9.0.0.0 correlation would leaveabout 650 years to be covered by this relatively insig-nificant period. Although the 11.3.0.0.0 correlation,leaving about 150 years for Coyotlatelco, would seem

to fit the present Valley of Mexico sequence best, itseems impossible, as can be seen from the diagram,to compress the Tzakol-Toltec Chiche'n Itza' intervalin Yucatan to a mere 150 years, making Toltec Chi-ch'n begin at about 9.16.0.0.0 Maya.

Wauchope (1947, 1948), working with highland Mayachronicles, computes successions of rulers back to

the arrival of the Quiches and Cakchiquels in the Guate-mala highlands. By assuming that the end of his TohilPhase, linked to Tula-Toltec Chich6n Itza by presence

of effigy plumbate pottery, coincided with the conquer-

ing of Zacualpa by the Quiche King Quicab, he placesthe end of the plumbate period in the early fifteenthcentury. Wauchope advances this as evidence for the11.3.0.0.0 correlation. Objections to this scheme are

clear in both the Yucatan and Valley of Mexico chro-nologies, which document successions of several plum-bate-free ceramic periods following the plumbate hori-zon. An interpretation of Wauchope's Zacualpa materialwhich fits our thesis may be made by accepting his fif-teenth-century date for the end of the Tohil phase, butextending this phase to a time length of 450 years on

the evidence of both mold-pressed effigy-head supportsand effigy plumbate in the same phase. In Yucatan,mold-pressed effigy-head bowl supports closely resem-

bling certain of his Tohil specimens54 are entirely re-

stricted to the Mayapan (latest) substage of the Mexicanstage, and plumbate is limited just as definitely to theToltec-Chich6n (earliest) substage.

65

V. THE DEVELOPMENT OF YUCATANCERAMICS: A TRIAL FORMULATION

The initial purpose of a ceramic survey such asthis is the formulation of a chronology, a frameworkto which later archaeological reconstructions can befitted. This is essentially a service function; otherhuman products found associated with pottery arethereby made dateable, and can be fitted to an histori-cal account. The classification and analysis used inproducing a chronology can and should proceed with-out dependence on any attempt at the reconstructionof culture, which is properly an interpretive ratherthan an analytic technique. Since the reconstruction ofculture history is dependent upon the chronologicalframework set up by typological analysis, it must log-ically follow that stage of the work.

The preceding sections of this report have beenbasically descriptive and analytic. In this section aneffort will be made to reconstruct the historical de-velopment of the ceramic craft in Yucatan. In the fi-nal section a more general cultural reconstruction isattempted.

To aid in the understanding of ceramic techniquesas practiced in Yucatan, a general description of mod-ern pottery-making there will be given first. Thereis every reason to believe that this craft has not greatsly changed during the past few centuries. Followingthis section the archaeological ceramics will be dis-cussed.

POTTERY-MAKING IN MODERN YUCATAN

The descriptive summaries given here are meantmerely to outline the problems of this branch of thestudy and to allow archaeological comparisons. Theinformation and opinions given are mine (often cor-roborated by others), save where other sources arespecifically cited; I should be held responsible for anyinaccuracies and errors. The sources for informationand theories follow: first, I owe much to discussionwith, and information from, Miss Anna 0. Shepard,who initiated me into the technology of Yucatan potterybefore I began work in Yucatan, and who has generous-ly given technical advice and theoretical clues duringthe course of this study. My work in Yucatan was pre-ceded by work on Near Eastern and Southwestern pot-tery under Professors Arthur Watts and Arthur Baggsin ceramics and under Professor W. J. McCaughey inoptical mineralogy at Ohio State University. The in-terest and kindness of these men in giving me manyhours of their time I can never repay. Descriptionsof modern Yucata'n pottery-making may be found inMercer, 1895, 1896, 1897; E. H. Thompson, n.d.; Ren-don, 1947, 1948; Henry B. Roberts' diaries, 1932, II,pp. 72-75, and III, pp. 1-2, Ticul pottery-making;1935, II, pp. 20-23, Tepekan pottery-making. I haveobserved pottery-making at Maxcanu, Becal, Ticul,and Uayma, and have taken a moving picture of theprocess at Becal. (See Brainerd, 1946, for a descrip-tion of kabal forming.) Except in the particulars notedin the text, no changes in technique over the 50-yearperiod covered by these reports has been noted. Re-gional variations are not so marked as to offer diffi-culties in making a single presentation of the ceramiccraft in Yucatan. The disadvantages of a composite

[66]

account are believed to be outweighed by the greatersimplicity and clarity possible in the single accountattempted here. It should be warned, however, that notall stages of the process have been checked for eachpottery center, and that more regional and chronologi-cal differences exist than appear in this account.

Materials

No striking differences among types of body claysare recognized by the potters in modern Yucata'n, norwould such be expected from the geological uniformityof the area. The clays55 come from sedimentary depos-its, many or all marine, and produce buff to reddishcolors in the firing. These deposits are localized, andform lenticular areas in the native limestone. Modernclay beds are famous among potters over the wholePeninsula and so must be of limited occurrence. Sincethe Peninsula is largely composed of limestone, theclay may be suspected of having a variable but oftenconsiderable calcium carbonate content. The modernYucata'n ceramics are unglazed. Modern potters claimthat formation of gas (carbon dioxide) during firingprecludes the use of glaze over native clay, but suchgas formation may be due to the universal modern useof calcite as an aplastic. a

Aplastic materials at present are of two sorts, sas-kab and hi'. Saskab is a soft gray earth, composed ofimpure crypto-crystalline (or very minute crystal)limestone. Various qualities of saskab are recognizedby the Yucatecans; the more impure may well be marlswith a considerable clay content. The harder and per-haps more lime-rich varieties are used for pottery.Saskab is mined as an earth; it is employed for vesselsnot to be used for cooking. Hi' is a form of limestone,characterized by a sugary crystalline structure. It isof localized occurrence, and nowadays is often hauledconsiderable distances; for example, horse-drawncarts now haul clay (k'at) from Yokat, near Ticul, toMaxcanu, returning with hi' from near Maxcanu. Clayis also reported to be brought from Becal to Maxcanu.At Ticul, charges for hi' and clay were said to be forlabor and cartage; no charge for mine ownership wasinvolved. Rendon (1947, p. 116) was told that hi' is insome regions calcined before use. I was told at Ticulthat a fire is built in front of the face of a hi' bed tosoften it for quarrying. These reports are to be sus-pected of inaccuracy and should be checked, since cal-cining would release quicklime which would adverselyeffect the working of the clay. Hi' from the beds nearMaxcanu is characterized by minute clayish red lumpsamong the crystals in some parts of the deposit (E. H.Thompson gives a Maya term for the red hi': Kanhi'or chac hi'). Hi' is used for refractory wares and isadded to the wet clay in quantity equal to or greaterthan the wet clay up to a 2 to 1 proportion by volume.

The slip used by modern Yucatan potters is of thered, leached, surface clay (k'ankab) which fills thecrevasses in the prevailing limestone. Imported redochre is often used to mix with it or as a substitute,and is more attractive because of its brighter color.Uayma vessels show a lighter-colored mottled orangeslip, probably caused by a regional variation in thek'ankab. As far as is known, k'ankab is never used as


a constituent of the body of the pottery. It is applied tothe surface with a cloth or brush.

In the preparation of materials for use, the grind-ing of hi' is of importance. It is mined in rocklikefragments which are crushed by placing them on acleared area of bedrock and rocking over them a large,crudely rounded limestone chunk (kamush, accordingto Thompson) which must weigh 300 or 400 pounds.Screens are used for hL'. Thompson reports a sieveof palm ribs (chachab) set at a slope, but sloping wirescreens have now replaced this. No reports describewinnowing to sort for particle size. Thompson de-scribes washing. The clay is sometimes picked over,and lumps containing orange streaks and areas areremoved. It is then dried, crushed crudely, and left tosoften in a vessel filled with water. The water is de-canted, and tempering material added and worked into produce a mix of good consistency for forming. Thepeople of Ticul say that if the proper quantity of tem-per is not added, the vessels will crack. Thompson de-scribes a marked swelling of the clay when water isadded.56 The clay is readied for forming into vesselsby molding into plump rolls perhaps 1 1/2 inches indiameter and 10 inches long.

Forming

Forming tools are arranged before beginning. Alow stool (kanche) of special form, with its seat aflat-bottomed low trough, is placed facing a smoothedboard of zapote or other hard wood (limia). This boardmay be rubbed with soap or grease. On the board isplaced the k'abal, a cylinder of wood commonly about5 inches in diameter by 5 inches high, on which thevessel is to be made. Kabals are made in variantforms and sizes, according to the requirements of theworker and the size of vessel base; a square top issometimes added to the kabal. The rolls of clay areplaced on the board to the potter's right hand, and abowl of water with tools such as a wooden or steelblade, an oval piece of gourd (box), and a cloth orleather strip by the potter's right side.57

The potter sits with knees widespread and the toesof the right foot and ball of the left pressed against theopposite sides of the kabal. A flattened pat of clay isfirst pressed to the kabal and the fillets are pressedinto place around it while the kabal is inched clock-wise by the feet. The work is done on the side of thepot at the potter's left. Clay is squeezed from theright hand onto the interior surface of the pot. Themotion of the kabal at this time seems to be directedchiefly toward keeping the vessel in a convenient posi-tion for work. The pressure for forming is given main-ly by the hands. The technique at this stage of formingis similar to that used in Pueblo pottery-forming inthe United States (see Guthe, 1925, pp. 37-42). A flow-er-pot shaped vessel is thus formed, with inch-thicksides and uneven surface.

The next stage consists of spreading and thinningthe pot. The rim of the half-formed vessel is trimmedto an even edge by a successive series of fingerpinches. The gourd tool is then placed interiorlyagainst the vessel wall, the left hand exteriorly, andthe walls spread and thinned into their final form.During this stage, tools and hands are often dipped inwater and the clay kept softer than is characteristicfor hand-modeling. A variable amount of the forceused in forming come s from the motion of the kabal,which forces the vessel wall between the stationaryhand and tool. Clay is scraped off and added as neededto make the walls even in thickness, while an extreme-ly even circular form is attained by the kabal motion.

The kabal is never spun through more than a fractionof a total revolution during this stage of the procedure,and it was noted that, whereas at Becal the force forforming came from this motion of the kabal, at Uaymanearly all the force applied came from the hands. Thisresulted, at Uayma, in uneven contour and surfacesand lack of circumferential forming marks on the fin-ished pieces. It was also noted that during this stagethe kabal was sometimes rotated counter-clockwiseby one potter, seemingly to rest his legs.

Simple vessel forms such as open bowls and basinsare formed at a single sitting. The potter proceeds di-rectly to finishing them from the stage detailed above,first by very careful evening of the rim, done by pinch-ing off clay and adding small dabs of clay with the fin-gers as the kabal is turned; then by a throwing or turn-ing procedure performed as follows: the potter wetsthe rim of the vessel, using the strip of leather whichhas been hanging on the rim of the water bowl, and wetshis hands. He then rearranges his feet, right heel andleft toes now coming in contact with the kabal. Theleather strip is held clamped over the rim to the pot-ter's left side, the left hand steadies the outside of thepot, and the kabal is rapidly spun clockwise through itsmaximum possible rotation, which is about a revolutionand a quarter. During this revolution the center of thekabal does not move more than one-half inch on theboard as measured by moving pictures taken, and thismovement is compensated by a shift of the hands. Theresulting rim is extremely even and true, approximat-ing that produced on a wheel (Plate I). The techniqueis in effect very close to that of wheel throwing. Onlythe axle is lacking, and this lack only slightly affectsthe hand technique and the qualities of the pottery shapeattainable. By varying tools and hand position, raisedmoldings and grooved lines may be formed (Plate I).It should be noted, however, that the kabal techniqueis capable, through variations of usage in individualpotters, of producing a complete range from handmadeto apparently wheelmade forms. Forming noted atUayma, for example, was much less exact than thatseen at Becal.

In the making of more complex forms than that de-scribed above, additional steps are taken. To form apitcher, the body up to the shoulder-neck junction isformed and allowed to become nearly leather-hard;then the upper part of the vessel is added by throwing,and finally the handle is luted to the partially driedsurface. In forming vessels with large, flat bottoms,the bottom is first formed and allowed to dry partially;then the edge is scored and dampened, and sidewallsbuilt upon it. To economize on time, several vessels,each on its kabal, are worked in rotation.

For the making of annular and trumpet bases, thevessel is first formed on the kabal, then allowed todry leather-hard and cut loose with a knife. The ves-sel is inverted and placed on a special kabal which hasa circularly cut board nailed to its top, with four equi-distant holes bored around its projecting border. Intothese holes have been inserted corncobs which projectvertically. The vessel body (such as the one shown infig. 34, g) is inverted and set among the corncobs sothat their soft surfaces hold it in place. The flat bot-tom of the body is then scraped off until round; it ismoistened, and a trumpet-shaped base built by coilsonto it. The same technique is used for pitchers withannular bases.

In Becal and Lerma all potters are men; in Tepekanit is reported that only women work on the kabal (seeRendon, 1947, p. 115). Mercer had the impression thatonly women were potters in Yucatan; his observationswere made near Me'rida. Tozzer (1907) illustrates a

67


woman working at a kabal in Ticul. At Uayna bothmen and women work. At Ticul both sexes work, butthe occupation is carried down patrilineally (Rendon,1947, p. 115). At both Maxcanu and Becal there is atleast a tendency for the women to make the figurinesand whistles, but the men use the kabal. For over 50years, at least, potters' wheels have been in use inYucatan, as has been the making of glazed potteryfrom imported materials. These trends, as might beexpected, have centered about Me'rida, but seeminglyhave never assumed larger proportions than individualor very small group enterprises. There is no evidencethus far of any dropping off of the local techniques.The best potter I observed, an elderly man in Becal,told me he had learned wheel throwing in Ticul in hisyouth, and that he had found it a much more rapidtechnique but preferred the kabal. Some kabal work-ers use a pivot in the form of a nail projecting fromthe board into the kabal; other workers disdain thisas slovenly practice.

Slipping

A slip is applied to most vessels, with the excep-tion of cooking pots. The slip is a creamy mixture ofk'ankab in water, in which the larger clay particlesare allowed to settle. Commercial ochre is sometimessubstituted or added to produce a brighter hue. It isapplied with a cloth pad or brush over all or part ofthe surface of the dry vessel. The cloth pad is pre-ferred by careful potters, perhaps because it permitsa more even, impacted surface than does the use ofthe brush. The vessel is then dried again before firing.

Fiishini

Thompson gives the name for a polishing stone

(vultunich), indicating that this tool must have been inuse within a half century. I have seen no modern pot.tery which shows evidence of stone or stick polishing,but the post-Conquest pottery of Mani contains sherdsso polished. The wiping of the surface with wet fin-

gers, cloth, or leather is done as a smoothing proce-dure, and particular care is taken to smooth out areasshowing protruding temper particles by application ofbits of clay and by rubbing with a damp finger. Oil andsoap are said to be rubbed into the slip before firingto produce a greater gloss, and I have seen them usedafter firing for this purpose.

Decoration

The vessels are decorated in various manners. Atechnique used in Becal and Lerma is the mold press-ing of low relief designs and inscriptions by meansof pottery stamps (see fig. 34, g, for an example, pl.

III for stamps). Occasionally, stamping is accompa-nied by simple incised work. These impressions are

normally placed in a band bordered by raised mold-

ings. The band is left unslipped, while the remainderof the vessel is usually slipped. The stamps are nega-tive molds of floral patterns and letters in bas-relief.These are used by holding the stamp to the exteriorof the freshly formed vessel and pressing the clayfrom the interior of the vessel wall into it. The tech-

nique can be called embossing or stamp (mold) press-

ing. The application of mold pressed low relief de-

signs by luting them to the vessel wall was also de-scribed by informants. In the making of animal figurinebanks at Ticul, a more elaborate use of the same bas-ic technique was noted. A layer of clay was pressedinto a series of fired clay molds (Rendon describes

plaster molds) of various parts of the animal's body,and the molded segments were then luted togetherwhile still damp. This technique is less efficient thanthe employment of slip casting in multipiece molds,which is widely used over the world for this purposebut has never been reported for aboriginal America.Hand-molding is used for some of the finer parts ofanimals made at Ticul, and for animal-shaped ocari-nas made at Maxcanu in forms similar to those madebefore the Conquest (see fig. 93, r u vz -_cc

At Tepekan, Campeche, painted water jars aremade (plateI) . The colors used are red and brownishblack. The red is k'ankab, the black isdescsribed byRoberts as made from "a small stone which occursin a type of earth called hacalche." Rendon describesthis material as calledta'kook meaning excrementof the turtle, and I found this name used at Becal. These

descriptions, coupled with the color produced, makeit nearly certain that a manganese oxide is used-prob-ably psilomelane, with its botryoidal or reniform as-

pect. These paints are applied to the dry surface be-fore firing and become permanent in the fire. The

paints are applied in an aqueous suspension. I do notknow if an organic vehicle is used in the water. Post-

firing colors are also used, sometimes on vessels and

commonly on figurines. The Maxcanu whistles andocarinas are covered with a white coating. Accordingto Rendon, this coating is of finely ground sascab,sorted by settling and decanting in water. The coatingis much like that on Lacadone and pre-Conquest Mayaincensarios. Additional decoration is painted on with

store-bought, water-mixed pigments. The modern Ti-cul animals are decorated in oil paint.

One further method of decoration should be recorded.The unslipped,hi' tempered cooking vessels, upon

being removed hot from the kiln after firing, are

splashed on their exterior s with a vegetal paint. This

paint is made by macerating and steeping for two days,or boiling, the bark of thechucum, tzalam, or pixoVtree. The dark, watery liquid so made is kept in abowl and applied by slapping it on with a corn husk or

palm frond. This splashing, as I observed it at Max-canu and Becal (Rendon also describes it), consistedof two or three splotches with additional spattereddrops, and produced no more than indistinct, faintlydarker areas. The potters insisted that no other pur-

pose than that of decoration underlies the use of this

paint, and seemed embarrassed when I suggested itssmall value for this purpose. This technique may very

likely be a cultural survival.

Firing

The use of a kiln for firing seems to be universalfor vessels in Yucatan. Rendon describes the openfiring of toys. Construction of the kiln is usually of

rough limestone with mud (k'ankob) chinking in a sim-

ple beehive shape.58 A small flue, which is oftendamped, is cut at top rear of the kiln in some cases;other kilns seen had no flue. The main kiln opening isat ground level, is usually irregular, and is closed by

dry-laid rocks during firing. A covering of palmfronds is usually placed on top of the oven to stoprain or to retain heat, or both. No kiln furniture seemsto have been noted, save for a ring of rocks used atTicul to support the pottery against the kiln sides.The kiln is sometimes preheated, and thin green twigsare used for this "to make greater heat." Thin woodsticks are used for firing, and several loads of woodare inserted at intervals. Inspection and rearrangingof the pottery during firing is performed with long

poles of green wood. Some potters close the kiln open-

68


ing with unmortared stones; others leave it open. Fir-ing lasts, by report, from 1 1/2 to 7 hours. In somecases, and certainly with the vegetal-painted cookingpots described above, the pottery is drawn from thekiln hot and allowed to cool on a bed of ashes. I haveother reports of pottery allowed to cool in the kiln.

Rendon gives a detailed account of the organizationand economics of the pottery industry and of market-ing and distribution. In the villages I have observed,the industry is undertaken by the family and carriedon in typical family quarters. The necessary equip-ment is simple and no specialized structures are usedexcept the kiln. Pottery is usually marketed by mer-chants, some of whom are itinerants who carry ves-sels and figurines on their backs to the village mar-kets. A lashed pole frame, which may be pre-Conquestin origin (fig. D) is used for carrying large jars andcooking pots. It is carried by a tumpline. Yucatanand Campeche, at least those parts within reach of therailroads, constitute a single trade province at thepresent time; in Merida, regular stocks of glazedearthenware from the Mexican mainland reach themarket but usually in badly chipped condition and athigh prices.

The most serious omission in this account of mod-ern pottery is the absence of data on its use. The Yu-catan industry is still overwhelmingly Maya and cen-tered in outlying areas of the state. Although metalvessels and china have partially displaced earthen-ware, I suspect that whenever the earthenware is usedits place in the household has changed but little sincethe Conquest, and thus may present valuable clues tothe archaeologist.

There is no doubt in my mind that the pottery-mak-ing practices now used in Yucatan are in overwhelm-ing proportion pre-Hispanic. Vessel forms and mate-rials are very close to those of the Late Mexicansubstage which shortly preceded the Conquest. Theyare so close that it may safely be said that the last500 years' history has witnessed probably less, andcertainly little more, change in the ceramic craft thanthe most conservative periods of similar length whichpreceded it. These facts should be clear from the sec-tions which follow, in which the many similarities tomodern ceramics will be drawn upon for inferencesof ancient practices.

ARCHAEOLOGICAL MATERIALS

Clay

The clay used in Yucatan archaeological potteryhas not been subjected to technological study. It givesthe impression of relative uniformity, and probablycame from sources similar to those used there atpresent.59 There is good reason to believe that a greatpart of the color differences noticeable in pastes ofthe archaeological pottery result from two causes:(a) differences in firing atmosphere and possibly intemperature; oxidizing atmosphere causes buff color,reducing causes white, gray, to black; (b) type andamount of added aplastic (tempering) materials. Col-or of the paste is also suspected to vary not only dueto the actual color of the aplastic particles, but totheir chemical effect on the fired color of the clay.Sherd and volcanic ash probably add to the porosityduring firing, thus facilitating complete oxidation ofthe paste and resulting in buff to reddish color. Cal-cite in finely divided form is known to bleach the redcolors produced by iron oxides when fired in oxidizingconditions. In the Medium and Thin Slatewares of the

Florescent stage, calcite tempered pastes average alighter, grayer color than the sherd or tuff temperedpastes, probably due to this bleaching effect in theirfiring.

Temper

"Temper" is here used in the incorrect but nowtime-honored archaeological sense of aplastic addedto the pottery clay. Microscopic sections and chemi-cal and optical tests were made on Yucatan pastes byMiss Shepard, and followed by microscopic examinationof several thousand sherds. These determinations weremade in 1940, and thus include only samples excavatedbefore that date. The temper identifications used hereare Miss Shepard's save for those of Yaxuna and Co-ba, which I had made previous to her work. Miss She-pard, whose authority on these matters is unequaled,has made her notes available to me, as well as freelygiven information and discussion during the course ofthis study, but should not be considered responsiblefor any misuse I have made of her data. Many of myinferences are based on scant information, so thisdiscussion can in no sense be considered as a defini-tive report on the technology of Yucatan ceramics; itcomes closer to being a prospectus organized fromclues and leads which have accumulated during thestudy. It is presented here, though unfinished, becauseI believe that only by a consideration of the underlyingtechnology can cultural reconstruction (as opposed toarchaeological analysis) be made. The value of futuretechnological study would be enormous in establishingthe craft history as well as interregional influencesin Yucatan ceramics. (Shepard, 1951, includes addi-tional information on this subject.)

Tempering materials, although poorly sampled asyet, show suggestive clues to craft traditions and posecertain problems as to origin of materials. Miss She-pard in October, 1940, summarized her determinations.She named five principal types of paste: calcite tem-pered, volcanic ash (tuff) tempered, sherd tempered,clay lump, and "untempered." The great majority ofthe sherds can be assigned to one of these groups. Theuntempered type is qualified, since the naming is in-ferential; the minute inclusions contained in the pasteare such as would be expected to occur naturally inthe clay.

Clay lump tempered pastes are also a qualifiedgroup; we do not know if the "lumps" are native tothe clay or are added, and if so what is their source.The lumps are in the general size range of the tempermaterials, subangular to rounded in shape, brick redto brown in color, in shades darker than that of thesurrounding paste. They occur in pastes which usual-ly, but not always, contain a variable amount of calcite;they also occur with variable amounts of volcanic ash.Miss Shepard noted specimens showing volcanic ashand calcite in such proportions as would indicate in-tentional mixture. A logical assumption from the abovedata is that the clay lumps are a natural constituentof certain of the Yucatan pottery clays. An alternativeassumption, depending upon the observation of smallreddish inclusions in the beds of hi' near Maxcanu, isthat the clay lumps were added in the temper, a vari-ant of hi'. This question might be resolved by study ofcollections of modern Yucatan pottery materials.

Miss Shepard describes a considerable variety ofvolcanic ash, variable in texture and structure, aswell as in the materials adherent to and intermixedwith it in lumps. These intermixed materials are insome cases flaky and calcareous, in others are a com-

pact buff-colored clay; they are probably included in

69


the volcanic ash as a result of secondary sedimentarydeposition. No coarse pumiceous ash or ash with min-eral inclusions was noted.

The sources for volcanic ash temper in Yucatanhave not been located. In my opinion, the quantitiesof this material which must have been used in thelarge vessels of the Florescent and Early Mexicanstages preclude the possibility of importation of vol-canic ash from the nearest areas where vulcanismnow occurs, a distance of some 500 kilometers. Theonly locally occurring material of volcanic natureknown at present is pumice, which is sold in the M65.rida market for scouring and is said to be found inlumps on the beaches. This pumice may have beencarried by water from the volcanic areas of Guatema-la and Mexico. But Miss Shepard's determinationsrule out this origin for volcanic ash temper on thebasis of its structure, which is never pumiceous. Thenature of the material and the physiography of Yuca-tan suggest sedimentary deposits, where ash fallshave been entombed in geologic time, as the probablesource of this temper material. Collections of sam-ples from the exposed strata of cenote banks mightpossibly produce volcanic ash of the kinds used fortemper, but a search has not yet been made.

Sherd temper presents no problem of geographicsource. Its use may well be more reliable as a cul-ture trait than are other temper types, since it is a

material universally available to potters and wouldnot be confused with various types of minerals. Sherdtemper particles often show inclusions of other tem-pers, thus establishing a sequence of temper use. Thehistory of this type in Yucatan is of interest and willbe discussed below.

Calcite temper shows some variety which can per-haps be explained by reference to modern Yucatecanpractice. Crystalline and crypto-crystalline calciteare distinguished by Miss Shepard. Crypto-crystallinecalcite ranges from opaque white through grays toblack, presumably depending on impurities in the min-eral and upon degree of smudging in the fire. Crystal-line calcite is transparent and shows cleavage faces;occasional crystals of dog-tooth spar occur. Dolomiteis believed to occur occasionally. Coarse and mediumtextured pastes are often tempered with coarse crys-talline calcite; much smaller crystals are also used,notably in the Thin Slateware of Uxmal. These oftenoccur in sugary-textured aggregates. In several knowninstances, crystalline calcite is found in distinctivelydifferent pottery types than is the crypto-crystalline.

It seems likely that the crystalline calcite of thearchaeological specimens is the substance now calledhi', and that the crypto-crystalline calcite is the mod-ern saskab, but it is certain that the modern restricteduse of crystalline calcite for cooking vessels was notuniversal in ancient times.

Vegetal temper, consisting of short bits of fibrousmaterial which may come from a monocotyledonousplant, was found in sherds of a single ware and vesselform at Coba and Yaxuna (fig. 1, e). This pottery israre and aberrant in other regards than temper; no

attempt will be made to fit it to the local craft history.

Slip s

The description of the use of slips and paints inYucatan is even more hampered by insufficient tech-nological data than is the study of temper. No analyt-ic work has heretofore been attempted on Yucatanpottery slips, although such studies would undoubted-ly shed much light. This section, like that on temper,must present theories and a statement of some prob-

lems. The qualities of slips and paints were noted atthe time of macroscopic sorting, and were describedfrom groups of 'typical pieces." Thus the general de-scriptions given here will not always be definitive insorting of wares sherd by sherd. Nevertheless, I thinkthere is little doubt that this descriptive material de-lineates the major technical peculiarities of the slips,and as such it is presented. The ware descriptions in-clude further qualifications. The techniques used insorting slips and paints are of particular importance,since the definitions of wares, the major potterygroups, are primarily dependent upon them.

In the sorting of slips on Yucatan pottery, severalcriteria were used: (1) color, although valuable, was

not as effective a sorting criterion in Yucatan potteryas it usually is among archaeological wares; too littlevariation in color is found; (2) degree of waxiness(translucency) or its opposite, opacity. Translucencyis initially dependent upon a compact, glossy surface,but many glossy surfaced slips are not waxy or trans-lucent. Translucent slips are distinguishable by simi-larity of slip color to underlying color, and by the par-tial visibility of landmarks such as cracks and protru-sions on the paste surface at the base of the slip. Theseslips are often sectile to the fingernail.60 Translucentslips frequently consist of pale paint which reverses

color on smudging (see below for description of thispaint; (3) tendency to flake off. This tendency is pre-sumably due to the presence of one or both of the fol-lowing conditions: (a) poor fit of the slip-that is, a

difference in shrinkage during drying, in coefficientof expansion during firing, and in final fired dimension,between the slip and the underlying body; and, more

important, (b) lack of bond between slip and body. Thismay be due to a burnished or smoothed body surfaceunder the slip, or to an underslip (underlying primaryslip) which for some reason has partially disinte-grated.61 These flaky slips are not to be confused withslips which, despite crazing (tendency 1 above), stilladhere tightly to the body. The loose bond of the slipto the underlying surface would seem to be the mostcritical causative factor of flaky slips.

Waxy slips were always found to be closely adher-ent; although they often show crazing, there is no ten-dency toward flakiness. Their frequent decoration bymeans of pallid (probably organic) paint, which oftenreverses color with smudging, has been remarkedabove. Their colors are always to some degree gray-ish, ranging from rare instances of deep red withgrayish to purplish surface bloom through orange red,dull buff, oyster gray and deeper grays to dull browns.They include much of the Formative Monochrome, allthe Florescent slatewares and redwares, and are oc-

casionally found among the Early Mexican slatewareand redware.

Flaky slips are all opaque and strongly colored;colors are red and orange. They include Flaky Red-ware, its decorated variants which are Incised Di-chrome and Trickle on Flaky Redware, and RegionalPolychrome (see fig. 63 for example). The two latterwares are Early Regional in date; the unpainted FlakyRedware is both Late Formative and Early Regional.

All other wares noted in the Yucatan sequence bearopaque, adherent slips. The Late Formative Mono-chrome from several of the sites belongs to this group,as do most of the Early Oxkintok Monochromes, theRegional Redwares, and the Middle and Late Mexicanslipped wares. Paints on these wares are strong andsolid in color.

Paints

In Yucatan, from Late Formative through Middle

70


Mexican times, a tradition of pottery decoration bytrickle paint seems to have persisted. And althoughLate Mexican pottery does not show evidence of thistradition, the process described for the decoration ofmodern unslipped pottery may still be an historicalremnant of this same long-lasting technique. The paintmay with reasonable certainty be called organic, al-though exhaustive tests have not yet been made on allwares here classed as having organic paint. But evenwithout a complete testing, its peculiarities are sostriking as to make its recognition through time rea-sonably certain. The manner of application of thispaint, as well as its appearance, suggests the samelengthy tradition and adds weight to its identification.

The characteristics of the organic paint, if we cor-rectly interpret it as such, are as follows: the paintreverses color when the area is smudged by reducinggases or smoke in firing. This reversal results in alight color in the painted area, while the slipped sur-face has been grayed. When found on a waxy slip, thepaint is grayish, pale, variable in color and oftenvague in outline, and does not alter the lustre of theunderlying surface. When found on an opaque slip, ason Trickle on Flaky Redware or on some specimenson Holactun Slateware and on Coarse Middle MexicanSlateware, it is strong black in color and sharp in out-line, save in areas which show reversal of colorthrough smudging. No reversal by smudging has beennoted on Trickle on Flaky Redware. The manner ofthe application of this paint is even more characteris-tic than its surface appearance. It is always appliedin broad, freely drawn lines, and shows a constanttendency to run over the surface. It is because of thisquality that we call it "trickle paint. " The paint ap-pears in long vertical blobs which seem to have beencaused by applying it at the rim of the upright vesseland allowing it to run down the side, sometimes withguidance. Figures are drawn on the floors of bowls,and these, as would be expected, do not show tricklingbut do show the characteristic vague outline. Someslateware designs show oval dark areas within thepainted area, which suggest a variability in its thick-ness caused by considerable viscosity in the paint attime of application, or by its very rapid drying. Iknow of no evidence as to whether the paint was ap-plied before firing or on pulling the vessels hot fromthe fire, as is modern practice.

It has been demonstrated by Shepard (Kidder andShepard, 1936, pp. 411-422), and found independentlyby the author, that the type of organic substance usedmakes little difference, but that the type of slip clayused is of prime importance to this process. The sug-gestion given above that bentonitic clays are availablein Yucatan is of interest, since Miss Shepard foundthat clays of this type are very susceptible to colora-tion by fired organic pigments.

The testing of samples of the various Yucatecanwares for organic paint is badly needed. Miss Shepardreports that the one sherd of Yucatecan slatewarefound at San Jose' (Thompson, 1939, pp. 269-270) wasdecorated with organic paint, and that on slatewaresherds from Chiche'n Itza' "the optical properties ofthe clay suggest that it is of the type which is highlyabsorptive and thus protects organic matter." In arecent brief report (Shepard, 1951, pp. 243-244), She-pard says of Yucatan slateware: "Tests of the painthave shown that it is organic, which makes the wareunique among analyzed Mesoamerican wares.... Theonly other example of carbon paint that has been sentto the laboratory is that on classic resist decoratedsherds."

The writer has noticed on Pueblo III Black-on-White

organic-painted pottery in northern Arizona,62 as wellas on experimental firing tests which he has made oforganic paints on pottery, that tones of paint and back-ground may reverse themselves, depending on firingatmosphere. In a fire which is kept smoky during itsfinal stages, the organic paint may act as a resist toprevent the absorption of smoke, while the backgroundbecomes dark gray in color due to smudging. Yucatanslateware vessels occasionally show this color rever-sal, and the trickle-painted Formative monochrome,which on other grounds I suspect to be decorated withorganic paint, shows it much more commonly. It islikely that the rare negative-designed fragments foundin Yucatan collections (see for example figs. 12, f;24, b) were so made. It thus seems likely that negativesmudging, as it may be called, is a widespread charac-teristic of organic paint, and that most New World re-sist or negative pottery decoration is of this type.63

Black mineral paints seem to be limited to poly-chrome pottery in Yucatan. The modern use of manga-nese ore for paints is nearly certain. They are presum-ably composed of manganese or of a manganese-ironmixture, but have not been tested. Red paints are oftwo types, easily distinguishable on the basis of hue.Hematite and ferruginous clay paints range towardorange in color, paints containing specular hematiteare more rare; they ran,ge toward purple and show acharacteristic sparkle.6, It may be added that specularhematite was found used over a ferruginous paint onstucco building facings at Dzibilchaltun and Yaxuna.

HISTORICAL ACCOUNT

Chart 23 shows the characteristics of temper, slip,and paint of the major slipped wares recognized inYucatan, ranged in rough chronological order frombottom to top. From the chart it may be possible totrace craft traditions through the sequent combinationsof these traits in wares. It must be emphasized, how-ever, that only three alternatives exist in this chartfor each of the three mutually exclusive criteria oftemper, slip, and paint. The presence of a single typeof temper, slip, or paint over a long continuing timeperiod does not of itself prove cultural continuity,since the number of possible choices is not greatenough to preclude chance. Moreover, the interruptionof any type, followed by a resumption in use, must becarefully examined to determine if chance rediscovery,a blank in the archaeological record, or reintroductionfrom an outside area is the most likely hypothesis.

Tempers

It is worth first commenting on the fact that allpottery believed to have been made in Yucatan is tem-pered.65 This may possibly be due to a lack of natural-ly occurring aplastic minerals in Yucatan clays. With-out such materials, shrinkage and consequent wastagein firing might be expected to be heavy. The coastalplain to the west, beginning at Champoton, seems tohave produced untempered wares. This area is char-acterized by river drainages carrying water fromvolcanic mountains, and naturally formed, aplasticmineral-bearing transported clay would be expectedfrom such an area. Although geological determinationsare lacking, we suspect the Yucatan pottery clays tobe relatively free of materials which act as aplasticsunder local pottery firing conditions.

Without question, calcite is the most generallyavailable aplastic. Its use is shown as absent onlyfrom the Early Mexican substage in chart 23, a mis-

71


leading absence since it is characteristic in the un-slipped ware of this as well as of all other horizons.A detailed study of the use of various types of calciteand of dolomite, the presence of which I suspect insome wares, would unquestionably yield more sharplydrawn distinctions. A crystallized calcite of fine, uni-form grain size and structure is characteristic of theThin Slateware of some Puuc sites. In some Yaxunadeposits the unslipped ware divides neatly into twosubgroups, each showing a combination of certain ves-sel forms with a definite type of calcite, one crystal-line and the other crypto-crystalline. Modern Yucatanpotters distinguish between crystalline and crypto-crystalline calcite.

Sherd temper is rare in the collections, was earlyin use, and has a distribution broken in time. Thepresence of Holactun Slateware irregularly distributedin Florescent deposits suggests the possibility of acraft tradition holding over from Formative times,although no documentation is available for sherd tem-per in the Regional stage. The likelihood that some ofthe slatewares of Yucatan are contemporaneous withthe Regional monochrome wares and represent a con-tinued craft tradition stretching on from Formativetimes is discussed below and at other points in thetext. There is no documentation of Holactun Slatewareor of any sherd tempered wares in Regional deposits,but much of our Regional pottery has not been testedfor temper and some may contain sherd temper. Ruz'work in Campeche demonstrates that the geographicrange of Holactun Slateware extends well down intoCampeche (Ruz, 1945, p. 61), and his Etzna pottery,as yet unpublished in detail, might yield valuable evi-dence for the persistence of sherd temper in Regionaltimes. In the central Maya area at Uaxactun and atSan Jose, sherd temper declines sharply in use afterthe Formative stage; a reintroduction of the techniqueinto Yucatan from the south in Florescent times istherefore unlikely. I can think of no reason why thistechnique should have been developed early and thenapparently discarded, unless desired qualities of hard-ness, durability, color, or behavior in heating mayhave dictated changes in temper. There are sugges-tions in our material that all these qualities are tosome extent influenced by tempering materials. Theorigins of sherd temper might theoretically be allo-cated to an area such as a broad flood plain, whereno aplastics occur naturally, but I know of no evidenceyet from the Maya area to support this hypothesis.

Volcanic ash temper is earliest known as a domi-nant type in our samples from the Regional stage atCoba' and Yaxuna.66 It is very common in the MediumSlatewares of the Florescent stage in the Puuc sites,and was used almost exclusively in the Early MexicanMedium Slatewares and Redwares at Chiche'n Itza'. Noash was used in later wares; calcite was used exclu-sively. In the central Maya area volcanic ash (tuff)temper is listed as present in Mamom pottery butabsent from Chicanel, reaching a peak of over 50 percent in Tzakol timnes (Smith, 1940). San Jose' showsthe same trend, but material from the Mamom hori-zon is lacking at this site.

The historical relationships of temper types can

be only generally and provisionally discussed becauseof lack of adequate archaeological sampling and of a

sufficiently refined typology through most time pe-riods and regions.

For the Late Formative stage we have sampled theHolactun collection and six sherds from Kabah. Thesesherds are all of Formative monochrome, nearly allwithout trickle paint. Sherd temper is predominant inthe Holactun sample, but some crystalline calcite,

both coarse and fine, is present, as well as some vol-canic ash.67

The Flaky Redwares sampled, all of Regional stage,show crypto-crystalline calcite temper. The sampleis small and restricted in provenience. The early Ox-kintok wares are virtually all calcite tempered, almostequally divided between crystalline and crypto-crystal-line with some fine crystalline calcite; volcanic ashtempers are present but very rare. According to mydeterminations, the Regional Medium Redwares ofCoba68 and Yaxuna are tempered principally with vol-canic ash, the Coarse Redwares from Coba' with cal-cite. Unfortunately, we have no temper determinationsas yet on Red on Thin Grayware, nor on its associatedslatewar e s.

The sites of the Florescent stage show marked var-

iability in the tempering materials used in the majorslipped wares. Among these, Holactun Slateware, wher-ever found, is almost entirely sherd tempered. Varioussherds sorted as Medium Slateware also show sherdtemper, but they never constitute a large percentageof that ware. The Thin Blackware of Holactun (a verysmall sample) is also sherd tempered, for the mostpart. Without much question, Holactun Slateware datesas late as 9.16.0.0.0; it is found in abundance in super-ficial deposits around the temple bearing that date.At both Uaxactun and San Jos6, sherd temper dropsoff markedly in popularity after the Formative stage.The use of sherd temper this late at Holactun suggestsa continuing tradition of its use, the intermediate timehorizons of which are thus far missing from our tem-per samples.

Calcite temper is concentrated at Oxkintok, wherethe Florescent slipped ware collection sampled (270fragments) contains only 3 per cent of other tempers.Ash temper is concentrated in general in the slippedwares of Sayil and Kabah, but this is not of itself a

particularly illuminating observation because thereare unquestionable tendencies toward relationship be-tween tempers and the various sorts of slipped wares,forms, and rim shapes, as well as between temper andpaste and slip colors in the Puuc area; these can even-

tually be more revealing than uncorrelated observa-tions. For example, Thin Slateware (sorted from Me-dium Slateware mainly on criteria of form) is almostexclusively tempered with fine crystalline calcite atUxmal, Holactun, and Oxkintok, but is largely ash tem-pered at Sayil and Labna and about evenly divided be-tween calcite and ash at Kabah. The differences intemper are paralleled by differences in wall thicknessand slip color; thin walls and light slip color are char-acteristic of fine crystalline calcite tempered sherds.Medium Slateware hemispheroid bowls are largelyash tempered at all sites, in distinction to the rest ofthe ware which is highly variable in temper from siteto site. Medium Slateware basal break bowls are main-ly calcite tempered at all sites sampled, and suchbowls are characteristically light in color. Clay lumptemper occurs most often in Medium Slateware jars,frequently in basins, but almost never in MediumSlateware bowls of any sort.

The variations above have been described by ware

and shape. Although sampling is meager, there is goodreason to believe that type of temper may be quiteclosely correlated with minor variations of rim andvessel shape. These, in turn, are unevenly distributedamong the sites, thus corresponding with the markedlydifferent temper proportions in the sites as describedabove. For example, several of the better defined ofthe dozen or more Medium Slateware jar rim typestabulated show consistent predominance of a singletemper type wherever found, but are markedly uneven

72


in frequency in the different sites. The inference mayplausibly be made that a series of groups of pottersin a certain area made vessels of consistent formstyle, using a characteristic temper, and that each ofthese types was widely traded over the area. This sit-uation is what one might expect from the modern in-dustry in Yucatan and throughout Mesoamerica, fromcertain results of Miss Shepard's Pecos study (Kidderand Shepard, 1936, pp. 581-583), and from the unevendistribution of the wares themselves among thesesite s.

The relative effects of regional and time differencesin producing this maze of interrelated variations amongthe Florescent-stage sites of the Puuc area are as yetincompletely known and are still under investigation.The separation of time in the series can profitably betackled first, and probably has the lesser effect on thetemper types; at least this is true if it can be assumedthat techniques and sources of material have changedthrough time more slowly than have shapes. The prob-lem of the cultural interrelations among these tremen-dous and closely spaced sites should be an unusuallyinteresting one, and the further study of their ceram-ics seems the easiest way to approach the problem.

The temper sampling of slateware from Yaxuna forthe Florescent stage is small and shows marked vari-ety in comparison with Chiche'n Itza', a few miles tothe north. At Yaxuna the determinations made on theslateware of the Roberts collection shows all calcitetemper in jars, more ash than calcite in the basins,and almost exclusively ash in the basal break bowls.A collection sent by Roberts to Peabody Museum showsmainly ash. These samples contain 183 sherds in all;the temper determinations are mine. My findings ona very small slateware sample from Coba' are partash, part calcite.

The slatewares from the mixed collections foundin the Hacienda Cenote at Chichen Itza' are unique inware characteristics, and I have provisionally placedthem as Regional on the basis of style and association(see caption, fig. 66). From a good sample, Miss She-pard determined them to be largely ash tempered.Vessel fragments of Florescent Medium Slatewarefrom Chiche'n Itza' were also determined to be prepon-derantly ash tempered. Thus a long, continuous tradi-tion in temper is indicated at that site.

The tempering materials of the Mexican stage havebeen sampled only at Chich6n Itza'. The Middle andLate substages show use only of calcite, and becauseof the widespread homogeneity of the wares of thesehorizons it is probable, though unproved as yet, thatno other temper was used on the Peninsula and thatthis tempering material has continued in exclusiveuse until the present. The tempering material ofslipped wares of the Early Mexican substage at Chi-che'n Itza is nearly all volcanic ash. At this site, Me-dium Slateware, both of Florescent and Early Mexicanforms, and the accompanying Medium Redware, showthe same overwhelming frequencies of ash temper.Thus there seems to have been no change in temperingmaterial during the transition between the Florescentand Mexican stages, and the later shift from ash tocalcite temper may be an underlying cause of the dif-ference in appearance which allows a cleancut sort tobe made between Medium Slateware (Early Mexicansubstage diagnostic) and Coarse Slateware (MiddleMexican substage diagnostic). The constant small per-centages of temper other than ash at Chiche6n Itzaduring both Florescent and Early Mexican times sug-gests trade. There is good reason on architecturaland other grounds to suspect such trade, at least inFlorescent times, from both the Puuc area to the west

and the Rlo Bec area to the south. There are no EarlyMexican substage remains yet identified from eitherof these areas in quantity sufficient for comparison.The presence of rare silica-sand tempered sherds atChiche'n Itza', as well as of the fine orange paste pot-teries of the Florescent and all subsequent substages,points to trade from the west and southwest, and tradefrom these directions seems to have continued throughthe Late Mexican substage.

The vegetal temper found very sparingly at Cobaand Yaxuna lies completely outside all known Meso-american traditions, and no plausible historical con-nections have been found in neighboring areas.

Slip s

The criteria which have been used for macroscopicsorting of slips are of two kinds: first, translucencyor "waxiness" and its opposite, opacity; second, flaki-ness and its opposite, adherence. Waxy slips are al-ways adherent; this leaves three kinds of slips plottedin chart 23: waxy adherent, opaque adherent, andopaque flaky.

Waxy and opaque adherent slips were both used inLate Formative times. I suspect that the marked dif-ference between various sites in this regard is due toregional rather than chronologic difference, althoughthis is far from certain as yet. Waxy slips usuallybear trickle paint. Flaky slips seem to be confined tothe later part of the Late Formative substage, continu-ing into the early part of the Regional stage; the evi-dence for this placement is that of the pottery forms.These occurrences probably reflect the introductionof a new technique for the preparation of surfaces forslipping. Flaky slips do not persist beyond the earlyRegional stage in Yucata'n. It is quite likely, as MissShepard implies (Thompson, 1939, p. 268), that theirpurpose may have been the heightening of slip color.If so, the Maya preoccupation in Late Regional andfollowing times with monochromes of muted color mayhave accompanied the abandonment of this somewhatlaborious double-finishing technique.

Waxy slips are absent from the Regional ceramicassemblages known so far, but characteristic of theFlorescent-stage ceramics. If a cultural continuity behypothesized for this sort of slip through the Regionalstage, it must be assumed either that it was reintro-duced from another region in Florescent times, orthat the slatewares, which in this area are character-istic of the Florescent stage, had started to be madein certain regions by early Regional times.69 Thereare reasons for suspecting that the latter hypothesisis correct. The Florescent stage is characterized bywaxy slips throughout; I have seen but few slippedsherds of this stage which do not have a waxy, trans-lucent surface. The superceding of waxy slips by opaqueslips in the Early Mexican subphase is striking in ourmaterial from Chichen Itza, although there is evidencethat it was gradual. The Chiche'n Itza Early Mexicanslipped wares are tempered exclusively with ash,which probably causes or at least contributes to thestrongly red paste color of these wares and may haveprompted the use of an opaque slip if a white surfacewas desired. The trend to opaque slip is unmistakable,although technical studies would add important details.The white, opaque slip of Middle Mexican Coarse Slate-ware is indistinguishable from the Early Mexicanwhite, opaque slip, but is often applied to a rough sur-face. The change from opaque white slip to the redk'ankab slip which is still in use today unquestionablyrepresents a change in slip clay source. I have no in-formation as to the source of the translucent and

73


opaque white clays; red clay is almost universal inoccurrence.

Paints

The distribution of paint types in time is for themost part closely correlated with that of the slips.The somewhat equivocal names used in chart 23 arenecessary at this stage because of the absence of tech-nological determination on Yucatan paints. A minimumof technical work should allow a clearer and morecertain set of determinations. As discussed above, thepaint called "trickle" is strongly suspected of beingorganic in nature, as is the modern type describedearlier. Opaque paints are mostly of mineral origin.The ones considered here are only those of waresfound in major proportion in the collections, all ofwhich are presumably of local manufacture. The paintof Flaky Dichrome is a mineral red. More reds and ablack (probably also mineral) ornament Regional Poly-chrome, much of which is probably of local manufac-ture. Oxkintok Monochromes are often decorated in ablack which is probably also mineral. Occasional useof red mineral paint occurs in the Florescent stageand in the Early and Late Mexican substages. The viv-id black paints found on Trickle on Flaky Red, and on

Middle Mexican Coarse Slateware almost certainly be-long to the organic group. They do not alter the sur-face reflection of the slip; they are applied in the same

trickle technique as the other paints of the series pre-sumed organic. The Middle Mexican Coarse Slatewarepaint at times shows color reversal (resist smudging),as do organic paints. The strong black color of thesepaints can with some reason be laid to a quality of theslip. Both slips are opaque; the former red, the latterwhite. These are the only trickle paints found on

opaque slips in Yucatan.The preceding text and chart 23 have described the

distributions of tempers, slips, and paints throughtime, and indicate in general a long persistence ofeach of the traits considered. Although the sporadicsampling of sites in Yucata'n now available is not suf-ficient for a sound regional study, it may be said thatthe evidence to date suggests regional variation inthese traits during several periods.70

Wares

The marked differences among Late FormativeMonochromes from various sites have been describedearlier. Some of these differences are probably chron-ological, since certain vessel and rim forms also ap-pear in Formative Flaky Redware which continued tobe made during the Regional stage, while a differentgroup of forms shows similarities to the Middle Form-ative collections dug in 1949. Regional form differ-ences may also be present; no thorough study has yetbeen made of these stylistic variations, and samplesare small and from widespread locations. The use oftrickle paint on Late Formative Monochrome alsoshows a marked difference in frequency among collec-tions from various sites, as do several other shapesand decorative techniques. At present we cannot sep-arate temporal and regional variations from this massof data with any degree of certainty. It seems likely,however, that this period was one of rapid technologicdevelopment, as the more or less contemporaneousMiraflores phase in highland Guatemala may also havebeen.71

The Regional stage has been so named because ofthe marked evidence of regional variation in its ce-

ramics. Regional Flaky Redware, because of its Form-

ative stage vessel form similarities and its stylisticlinkages to early Tzakol in design style, must be placedvery early in the Regional stage; its period of use verylikely preceded that of the earliest Maya calendric in-scriptions. It is distinctive in form and finishing tech-niques as well as in ware characteristics. Its distribu-tion is wider than that of any other Regional stage ware:

Yaxuna, Chiche6n Itza, Dzibilchaltun, Merida, Acanceh,and (if the seemingly related Regional Polychrome beincluded) at Chac Cave in the Puuc region, at Mani,and at Mayapan. I am inclined to believe that this ware,at present known only from small collections save atYaxuna, represents a widespread, well-developed hori-zon as yet only slightly known due to vagaries of sam-

pling. The distribution of Trickle on Flaky Redware atYaxuna (see fig. 7) makes it reasonably certain thatthe periods of manufacture of Flaky Redwares and Re-gional Medium and Fine Redwares overlapped at thatsite. Flaky Dichrome, a decorated variety of FlakyRedware, is of particular interest as a horizon markerin Mesoamerica (see Brainerd, 1948, and Wauchope,1950, for distribution and discussion), as well as in-dicating possible parallel development in the sequenceof ceramic decoration of the Mesoamerican and Andeanregions. There is little question that this ware is a

close relative of Yucatan Regional stage Polychrome,which in turn is a regional variant of the Tzakol Poly-chrome of the Peten region.72 There is, additionally,good evidence at Yaxuna that Flaky Dichrome precedesTzakol-like polychrome pottery. This evidence isstrengthened by the E. H. Thompson collection shownin fig. 63, a, which contains combinations of forms anddecoration which consistently grade from Flaky Di-chrome into Regional Polychrome with Tzakol featuresin both decoration and form (see fig. 33, caption).73Further confirmation of the close relationship betweenFlaky Dichrome and Regional Polychrome is the pres-ence of interior striation of jars as a common charac-teristic found only in these two wares. The markedsimilarity of the slips and paints of the two wares,particularly the unique brilliant orange slip, also sup-ports their interrelationship.

It is strange that our best documentation of the ante-cedents of Tzakol Polychrome, a pottery ware knownas a Peten Maya early Classic diagnostic over Meso-america, should appear from Yucatan, an area notedfor its absence of a polychrome style of decoration.No pottery with the decorative or shape repertory ofFlaky Dichrome has been found at Uaxactun, San Jose',or Holmul. Thompson (1939, tables 17, 18) suggestsa disjunction between Chicanel and Tzakol, and betweenSan Jose I and II, in which he places Holmul I andMountain Cow II. This period must be contemporaneouswith Flaky Redware in Yucatan since we can crosstieboth the preceding and following periods. Although thecollections from Holmul and Mountain Cow do not con-

tain material directly comparable to Flaky Dichrome,the collections are small and I find it easier to believethat the absence results from chance sampling ratherthan being an indication that Yucatan is the restrictedregion where Maya polychromes originated. The find-ing of incised dichromes in the Motagua Valley andthe Guatemalan and Mexican highlands adds evidencefor the view of a wide distribution of the repertory ofdecorative techniques which characterizes Flaky Di-chrome.

Regional Polychrome has not thus far appeared as

a major constituent in any excavated Yucatan collec-tion. At Yaxuna it forms 1 to 2 per cent of the cenote

collections, whereas Flaky Dichrome ranges about 10per cent. In the collections of trenches 24b c w andc, and 21-26 _, b (see tables 3 and 4), which are be

74


lieved to date later, the polychrome (here labeled Tza-kol) far outnumbers the Flaky Dichrome. In Mayapantrench 2, d, e (see table 17), the polychrome outnum-bers both Flaky Redware and the Regional Redwares,and no Flaky Dichrome was found. At Mani trench 12(see table 20), almost no Flaky Redware or Flaky Di-chrome was found, but polychrome bowls as well asjars occur in some quantity. The jars are orange ad-herent slipped and interior striated, much resemblingthe jar sherds from Chac Cave. All these collections,however, show mixtures of pottery from various otherhorizons. The Balam Canche collection (fig. 64 andcaption) is limited to Flaky Dichrome large-jar necksand a few slateware jar sherds. It is a surface collec-tion from a cenote cave and therefore limited in ves-sel repertory. Probably it has been additionally se-lected in its gathering. The E. H. Thompson collectionhas doubtless also been selected for decoration but, aspointed out above, seems to document a transition fromincised dichrome to three-color Tzakol-style poly.-chrome bowls.

The most likely reconstruction from these data isas follows: the early Regional substage is in generalcharacterized by Flaky Redware, a ware which beganto be produced during the Late Formative substage.Early Regional Flaky Redware evolved from thisthrough evolution of a distinctive shape repertory, andbegan to be accompanied by Trickle on Flaky Redwareand Flaky Dichrome at Yaxuna, and at least by FlakyDichrome over much of Yucatan. The Flaky Dichromeseems to have evolved locally into Regional Poly-chrome, the bowls of which are nearly identical withTzakol Polychrome in the Peten to the south, thus sug-gesting a cultural similarity over the whole area atthis time. Accompanying these bowls in western Yuca-tan, best sampled at Mani, are orange-red jars whichoften bore painted decoration. In the east, at Yaxuna,Flaky Redware is displaced by Coarse and MediumRegional Redwares. Trickle on Flaky Redware lasts,with vessel shape changes, into this later horizon atYaxuna.

The Oxkintok Regional ware assemblage is remark-ably distinctive and shows almost no trade with othersites sampled. Oxkintok Coarse Monochrome probablyinfluenced the later Red on Thin Graywares from Dzi-bilchaltun and Acanceh, but Oxkintok wares are knownelsewhere only from scattered sherds at Acanceh andby one bowl reputed to come from Campeche (fig. 103,d). The high technical quality of the Oxkintok ware rep-ertory argues for well-developed antecedents and cor-

relatives over a considerable area. The Yucata'n sitesdo not provide these but do show Tzakol-like poly-chromes which must be at least nearly contemporane-ous. Since Oxkintok is peripherally located on thewestern outlyers of the Puuc range, coastal Campechemay well have been the home ground for the develop-ment of these wares, a group which may be importantin influence on Puuc pottery. Unfortunately for theabove hypothesis, Ruz' carefully done ceramic recon-

naissance of the Campeche coast produced little ifany Oxkintok-like pottery. But Ruz' material alsoseems lacking in Early and Middle Regional ceramicsof any sort, which suggests this absence may repre-sent a sampling gap rather than a true absence of Ox-kintok-like pottery along the Campeche coast. Thusthe distribution of Oxkintok-like ceramics still cannotbe given.

A problem is raised by the presence of a RegionalFlaky Redware horizon at Mayapan, Mani, and ChacCave. The Oxkintok and Regional Flaky Redware ce-

ramic assemblages show practically no evidence oftrade, and this suggests that they are not contempora-

neous. Both show Tzakol-style polychrome bowl frag-ments, frequent in Flaky Redware deposits and veryrare in the Oxkintok Regional-stage deposits. Thereare reasons, given above, to believe that RegionalFlaky Redware immediately follows the Late Forma-tive substage. Therefore, it is likely to precede theOxkintok samples, whose later position is additionallysuggested by stylistic similarities to Dzibilchaltun andAcanceh wares. The date of the Oxkintok lintel, whichis 9.2.0.0.0, points to the occupation of the site at thatearly date; the pottery is not stratigraphically asso-ciated with the lintel, although it comes from a majoroccupational deposit found near the lintel building. Ifthe hypothesis be used that 9.2.0.0.0 is the date of thedeposit, the Regional Flaky Redware deposits mustantedate this. Since 9.2.0.0.0 is only 160 years laterthan the date of the earliest inscription, the RegionalFlaky Redware deposits likely antedate the beginningof inscriptions, thus falling within the Formative or

pre-Classic stage, if this stage be thus defined.74 Thehorizon has been called Early Regional in this reportbecause in our present sequence in Yucatan, this hori-zon begins with major breaks in the style of vesselforms and of decorative techniques from the precedingLate Formative horizon. It is possible that the markedstylistic changes between Late Formative and EarlyRegional ceramics in our sequence may reflect a dis-junction in deposits rather than a rapid cultural change.If this proves true, this departure from the centralMaya area stages may be unjustified.75

The Regional Redwares of Yaxuna and Coba showmany features in common and are accompanied by di-chromes and polychromes, particularly at Coba, whichalso show close correspondence with the central Mayaceramics as exemplified by the Uaxactun and San Josddeposits. The documentation of Coba as an outpost ofthe Classic Maya hierarchic cult is strong from evi-dence of architecture and sculptured monuments, andit is not surprising that Yaxuna, at the other end of thegreat causeway, should show similar affinities. Thecentral Maya crossties of the regional wares are tothe Tzakol and Tepeu 1 phases of Uaxactun; San Jose'crossties substantiate this assignment. The AcancehRegional Redwares show only general similarities inpaste and slip to those of Coba and Yaxuna. Vesselform correspondences, however, are much more spe-cific than those between the wares. There is a goodsample of Peten-style polychromes from Acanceh(fig. 9, c-d). Those accompanying the Regional Red-wares are Tzakol in date; no Tepeu polychromes were

identified by R. E. Smith from the collection. Thiscrosstie places the Acanceh Regional Redwares some-what earlier than those from Yaxuna and CobS and ap-proximately contemporaneous to the early Oxkintokassemblages, which Smith believes synchronous withhis Tzakol 3. This synchronism is reinforced by thefinding of occasional Oxkintok-like slipped fragmentsat Acanceh (see fig. 19, ) and the likelihood of a re-

gional variation in ceramics between the two sites isreinforced thereby.

Red on Thin Grayware comes from two sites, Dzi-bilchaltun and Acanceh. It is accompanied by slate-wares at both. There is good stylistic evidence fromform that the Acanceh sample partially precedes theDzibilchaltun sample in date of deposit, and trade evi-dence that the two Acanceh varieties synchronize withTepeu 1 and Tepeu 2 respectively'at Uaxactun. Theforms of this ware give clues as to its stylistic ante-cedents from various Regional wares. There is more

doubt as to its ware characteristics. The two alterna-tive colors found on slip, Cinnamon Buff and red, sug-.gest a relationship with the Oxkintok Monochromes

75


where the same two interchangeable colors are found.This suggestion is strengthened by the lack of smooth-ing of the body and slip surfaces of Red on Thin Gray-ware which is paralleled in Oxkintok Coarse Mono-chrome, but in the absence of technologic study itcannot yet be proved to be due to similar techniqueand materials.

The slatewares have presented the most complexproblem of analysis during this study. They have beenfound stratigraphically above Formative and Regionalmonochrome wares at almost all sites tested in Yuca-tan save in the Puuc area, and have been found as anunquestionable part of assemblages which also containRegional monochromes at Dzibilchaltun and at Acan-ceh. At these sites and also at Mani, the slatewareshave shown certain common characteristics of wareand form which distinguish them from the bulk of thePuuc slatewares, and which in some degree ally themwith slatewares from Chichen Itza, Yaxuna, Oxkintok,and Holactun.

The relative effects of regional variety and chron-ologic change have been hard to assess. If the differ-ences can be assumed to be in the main part chrono-logically caused, the Puuc sites must follow a "Region-al-Florescent transition" seen at Dzibilchaltun, Acan-ceh, and these other sites. But the absence of Red on

Thin Grayware at Mani despite the presence of slate-wares quite similar to those of Dzibilchaltun andAcanceh, and the near absence at Mani of RegionalRedwares which at Acanceh and Yaxuna occur quiteconsistently in association with the slatewares, sug-gest that regional variation may also be playing a part.

The alternative to the assumption that chronologicchange is responsible for the situation outlined aboveis to assume contemporaneity for all these slatewarecollections. This assumption, if pursued to its conclu-sion, requires that the Puuc sites be dated as contem-poraneous with Dzibilchaltun and with the other sitesenumerated above as producing similar slatewares.This solution is unlikely because the thin and finewares are so markedly different between Dzibilchal-tun and the Puuc, and these wares, because of theirwide trade connections, are significant in their ab-sence as well as presence, and thus allow crosstyingwith foreign chronologies.

The 1949 Chenes-Rio Bec ceramic survey, not yetcompletely analyzed, gives much additional informa-tion, but thus far not a complete solution to the prob-lem. The two Chenes sites dug have produced slate-ware assemblages which resemble the Dzibilchaltun-Acanceh-Mani group more closely than the Puuc, withgood evidence that typical Puuc slateware overliesthem in the trenches. This adds strong evidence for a

chronological subdivision for the slatewares, sincethe Puuc area has now been "surrounded" by evidenceof earlier ceramic types, and stratigraphic evidenceof the superposition of Puuc-style slatewares on theearlier slateware types has been produced. It shouldbe held in mind, however, that definition betweenPuuc-style and earlier slatewares is not possiblefrom small lots, and that the evidence for superposi-tion rests only on changing percentages of potteryshapes, not on ware changes or total replacement ofshapes. Also, there is not even near identity betweenthe Chenes slatewares and those of northern Yucatan,although they resemble each other more than they dothe Puuc ceramics. Some of this evidence has beenconsiderably reinforced by statistical seriation stud-ies of the Yucatan, Chenes, and Ri<o Bec sites, the re-

sults of which will be published later. The two Chenessites sampled, Santa Rosa Xtampak and Dzibilnocac,although producing slatewares known to be earlier

than those from the Puuc, have produced almost noRegional monochrome pottery of any sort, althoughFormative pottery is relatively common on both. Theyalso have not produced any quantity of the Dzibilchal-tun and Puuc thin and fine wares nor of Holactun Slate-ware. The absence of Holactun Slateware is peculiarif our chronological reconstruction is correct, and ar-gues for strong cultural barriers; 9.16.0.0.0 datescome from both Holactun and Santa Rosa Xtampak,where they are presumably contemporaneous with siz-able sherd collections obtained from their immediatevicinities. The following chart tabulates this situation.

EarlyFine Holactun Florescent

9.16.0.0.0 Gray Slateware SlatewareDzibilchaltunHolactunSanta RosaXtampak

- +

+ + ++

+

I am inclined to believe all three sites were heavilyoccupied at about 9.16.0.0.0. For the two Chenes sitessampled we must either assume poor sherd sampling,occupations with Regional-stage chronological disjunc-tions, or that Medium Slateware began to be made inthis region immediately after the use of Formativeceramics stopped.

In the Puuc area the collection from Chac Caveshows that there was a Flaky Redware-using occupa-tion in the region. Slateware presumably followed thatperiod. At Oxkintok there are large collections of Ox-kintok monochromes which show not a sherd of slate-ware. In the Yucatan collections I do not find a singlestatistically significant lot of excavated pottery con-

taining either Flaky Redware or Regional Redwareswhich do not also contain Medium Slateware sherds,although in some cases the Medium Slateware is re-corded as one per cent or less and the slatewares are

noted as atypical. Although these collections all come

from sites which supported heavy slateware-using oc-

cupations, some of the collections, as at Acanceh forexample, came from under unbroken floors and thusintrusives are unlikely. I do not believe that the abovefinds constitute absolute proof that Medium Slatewarebegan to be made during early Regional times. It isconceivable that the few slateware sherds found mayhave fallen from trench edges, or been misidentified;some of the Late Formative monochromes, for exam-

ple, have a surface much like slateware. It is morelikely, however, that slateware was made from earlyRegional times onward in some areas on the Peninsula.This means that increasing attention must be given tovessel form and to technological work if close chron-ologies are to be built for Yucatan. As remarked else-where, there are variations which, with microscopicchecking, may allow reliable subgroups to be sortedon a sound basis. With larger and less mixed earlyslateware collections a good sequence can be built.The present work is of necessity merely a survey.

In summary, the slatewares are divided into ThinSlateware and Medium Slatewares, with the main sort-

ing criterion being that of vessel form; thinness ofwall and fineness of paste are subsidiary. The lack ofcomplete correlation among these criteria suggeststhat with temper and slip analyses and fuller archaeol-ogical sampling, more meaningful classification couldbe evolved, and the regional and chronological varia-tions could be separated. In chronological analysis ithas been possible both on ware and form criteria to

separate out from the mass of Florescent Thin andMedium Slatewares groups of characteristics which

76


indicate an "early" horizon, during which slatewareswere contemporaneous with Regional monochromewares in various regions of northern Yucata'n, froma "late" horizon best represented by the bulk of theceramics from the sites of the Puuc region. Theseearly characteristics are not at present of a sortwhich allow a definitive sorting of all slatewares intoearly and late, or Puuc, horizons. Because of this sit-uation, I have retained the earlier nomenclature ofclassifying all these slatewares as Florescent, andclassing sites which show slateware-monochrome col-lections as Regional-Florescent transition, although Ibelieve that the bulk of the Florescent slatewares ofMani, for example, are contemporaneous with Region-al and Regional-Florescent ceramics at Acanceh andDzibilchaltun.

The Early Mexican subphase is also characterizedby Medium Slateware as its major slipped pottery.Thin slateware does not last into this period. The maindiagnostic between Florescent and Early Mexican Me-dium Slatewares is again that of form and decoration,although change to an opaque white slip is also valu-able as a sorting criterion. The changes in style ofEarly Mexican Medium Slateware from the precedingFlorescent ware were more far-reaching as well asmore abrupt than those which took place during theFlorescent stage. They were caused by an influx ofpottery and general culture from the Mexican main-land. The main trade pottery, X or Chichen Itza FineOrange, comes from central Veracruz. Plumbate, ararer import, comes from the Guatemala highlands.Either trade pottery or strong stylistic interinfluencewith the Valley of Mexico is suggested by resemblancesto Tula Toltec ceramics. Resemblances between bothPuuc and Early Mexican slatewares and Xochicalcoceramics, although generalized, are pervasive enoughto suggest strong cultural connections of some sort.Some Early Mexican slateware ceramic forms notpresent in Puuc slatewares seem to point to southerninfluence. Among these are grater bowls, pedestals,and pestles.

The Thin and Medium Redwares, as described else-where, seem to be a color variant of their accompany-ing slatewares, probably dependent upon paste colorfor their distinctive appearance. The forms of thesewares are identical with those found among the slate-wares, but are restricted to the smaller, more elabo-rately decorated vessels. Thin Redware, like ThinSlateware, is not found in the Early Mexican substage,and Medium Redware forms of this period are quiteclosely restricted to copies of X Fine Orange, presum-ably because similarity to X Fine Orange in the colorof Medium Redware suggested such copying. Copyingof X Fine Orange vase forms is almost completelyrestricted to Medium Redware; X Fine Orange jar andbowl forms are copied in both Medium Red and MediumSlatewar e s.

Coarse Slateware is the diagnostic ware of the Mid-dle Mexican substage. Our collections of this ware are

relatively small; all evidence is that it follows quiteclosely in the stylistic tradition of the Medium Slate-ware which precedes it. Vessel walls average thicker,a result of the coarser paste used. Black-painted dec-oration increased in exuberance, doubtless inspiredby its visibility against the vivid white slip. The col-lections of this ware from Chichen Itza show mixtureswith both Medium Slateware and Coarse Redware (thediagnostic ware of the Late Mexican substage); Dzibil-chaltun collections show very slight Medium Slatewaremixture; Mayapan collections show Coarse Redwaremixture; although we cannot be sure on present evi-dence, we suspect that Coarse Slateware was manu-

factured as the exclusive slipped ware during a periodof time.

Coarse Redware succeeds Coarse Slateware and isthe diagnostic ware of the Late Mexican substage. Pos-sible forerunners of this ware, as small basal break,tripod-footed, unpolished bowls, occur in Early Mexi-can collections at Chiche'n Itza. The forms in this ware,although showing no break in tradition from CoarseSlateware, are usually distinctive from Early Mexicanwares through progressive style change, and perhapsthrough further foreign influence from Mayapan FineOrange, a ware imported from Campeche or perhapsTabasco. Mayapan Fine Orange shows culture inter-relations with Aztec III ceramics in the Valley of Mex-ico, and we have historical accounts of Aztec coloniesin its area of manufacture, as well as accounts of theimportation of troops from this area to Mayapan. Thusthe ceramics do not bear out earlier suppositions of aMaya resurgence in Yucatan during this period. Markedsimilarities between Yucatan ceramics of this substageand the scanty published collections from various siteson the east coast of the Peninsula suggest that the wholearea possessed a common culture at this horizon, andthat Mexican mainland influence was present over thewhole area.

Colonial Redwares show a marked continuation ofYucatan ceramic traditions in water jars, and this tra-dition has remained stable until now. Other vesselforms have shown changes, mainly impoverishmentin shape repertory, but macroscopic ware characteris-tics have remained relatively constant through the last500 years.

Ceramics of fine-textured, apparently untemperedpaste are found in collections ranging from the Region-al-Florescent transition until the Conquest. In the Re-gional-Florescent deposits at Dzibilchaltun are foundDzibilchaltun Fine Grayware and Dzibilchaltun FineOrangeware. From the Puuc Florescent stage colleo-tions comes Puuc or Z Fine Orangeware, from EarlyMexican substage Chichen Itza comes Chichen or XFine Orangeware, and from Mayapan Late Mexicandeposits comes Mayapan Fine Orangeware. Occasionalfragments of these wares have been found at othersites, in each case in deposits contemporaneous tothe type locations. The names given these wares in-clude the name of a type site, in each case the siteyielding the largest collection, because of the greaterusefulness of a specific name in discussion of thesewares when found elsewhere than in Yucatan. Sincethe wares are demonstrably foreign and specificallyrecognizable sherd by sherd, the reasons given else-where for the use of generalized names such as havebeen used for the always variable and usually inter-grading Yucatan wares do not apply here.

These wares, as well as certain of the thin waresdescribed here-plumbate pottery, and certain otherfragments illustrated and described but not classifiedfor want of a sufficient sample-have in common anotable thinness of wall, uniformity and clearness offiring color, fineness of paste, delicacy of forming,and elaboration of decoration. They occur only in re-latively small vessel forms, such as beakers of vari-ous forms, small bowls, cylindrical, pyriform, and oth-er vase shapes. The uniformity and virtuosity of theirworkmanship, the frequent presence of well-integratedform and design styles, and in many cases the unusualnature of their composition suggest that each ware wasmade by a group of specialists within a relativelysmall area, for the specific purpose of trade.76 Theyare unquestionably luxury wares, were widely traded,and are therefore of particular value for establishingsynchronism s and interrelationship s among cultur e s

77


of various areas, both through the study of the waresthemselves and through the study of the local copyingwhich frequently followed their spread.

Fine Grayware has remarkably close similaritiesin form and decoration to ceramics at Yoxiha, Chia-pas, and Piedras Negras, but the ware itself does notseem to be documented from these or adjacent areas.I have noted sherds of Fine Grayware in the CampecheMuseum collections from Jaina. The color of FineGrayware and certain general characteristics of formsuggest affiliations with the Oaxaca area, but thereare no really definite resemblances. If the gray colorof this and the Oaxaca wares can be assumed to be dueto smoke clouding (and such an assumption is reason-able), this technique, uncommon in Mesoamerica,somewhat strengthens the dubious case for Oaxacaninfluence on this ware. The paste of this ware and ofthe fine orangewares looks very similar in texture tothe naked eye. It is to be suspected that both are nateurally sandy transported clays, probably from riverflood plains. In X Fine Orange, one type of simpleprofile bowl is characteristically fired gray towardits bottom, suggesting that fine gray and fine orangepottery can be made from identical materials merelyby changing the firing technique. As for the dating ofDzibilchaltun Fine Grayware, only general estimatesare possible. Mine would be between 9.12.0.0.0 and9.18.0.0.0.

Dzibilchaltun Fine Orangeware is found constantlyassociated with Dzibilchaltun Fine Grayware at thetype site; the fine orange is much less frequent in thecollections than fine grayware, averaging under 5 percent as compared to a range of from 6 to 25 per centfor fine gray. There is no evidence for color gradationbetween the two wares nor for other similaritieswhich might suggest cultural interrelationships be-tween their makers. However, Dzibilchaltun FineOrange is without question culturally related to PuucFine Orange in such manner that I suspect that PuucFine Orange was made by the direct cultural descend-ants of Dzibilchaltun Fine Orange. I have elsewheredetailed the evidence that the beaker is a stylistichorizon marker of the luxury wares of the time periodof the Dzibilchaltun occupation. Both the fine graywareand fine orangeware occur principally in this vesselform at Dzibilchaltun.

Puuc, or Z type, Fine Orange is distinguished fromDzibilchaltun Fine Orange by several characteristics,the most definitive of which is that of vessel shape;the outcurved-sided beaker predominates in Dzibil-chaltun, the hemispheroid bowl in Puuc. Other differ-ences lie in the coarser technique of incision of Dzi-bilchaltun Fine Orange, and in the design repertory.Puuc Fine Orange includes a group of bowls decoratedwith crude scalloped designs in a dull gray or reddishpaint which does not intergrade into the rest of thegroup (see fig. 59, f, and thus may conceivably havea different region of origin. Although the region ofmanufacture of this ware is far from certain, theCampeche and Tabasco coastal regions seem the mostlikely at present. Its dating may on rough estimate beput from 9.18.0.0.0 to 10.8.0.0.0.

Chichen, or X type, Fine Orange is found in theEarly Mexican substage deposits at Chiche'n Itza insurprisingly large quantity considering the distancefrom its source. In general, the paste seems similarto the paste of all the other fine orangewares, andthere are several other similarities such as the use

of white and black slipped areas and the decoration inplano-relief carving. A new technique is found in thisware, that of decoration by elaborate painted designs.The painted designs, although used on the vessels

with plano-relief carving, are in a quite different tra-dition from that of the plano-relief (see Brainerd,1953), which in turn does not show close relationshipsto that of the earlier fine orangewares. Form, as wellas decoration, shows only general similarity to theearlier fine orangewares. With little question, X FineOrangeware comes from central Veracruz and can beassumed to have been introduced via coastal trafficby the Toltecs, the Mexican highlanders who are re-

sponsible for the surprisingly close resemblances be-tween Chiche'n Itza', Yucata'n, and Tula architecture.The plano-relief design of X Fine Orange is relatedto the 'entrelace" design of the yokes and palmas, andto the bas-relief sculpture of the Classic sites of theTotonac area. The Toltecs showed good taste in im-porting a ware much superior to the best made in theirhomeland. The effect of the copying of these wares bylocal Yucatecan potters is of particular interest as a

phenomenon of culture contact; of even greater interestis the use of elaborate foreign motifs by the bas-reliefsculptors of Chiche6n Itza, and of foreign architecturalplans and sculptured designs by Yucatecan buildersof this period.

The chronological placement of X Fine Orange inthe Yucatecan sequence can be made with reasonablecertainty. Its import began after the end of occupationof the Puuc sites (two sherds from Uxnal depositsprovide the only evidence contrary to this statement).However, sherds closely resembling the X Fine OrangeChiche'n Itza specimens are found at Mayapan in suchcollections as to indicate that its importation continuedinto the Middle Mexican substage (see fig. 28, caption).A chronologically significant breakdown of the varia-tion among these potteries should eventually be feasi-ble, but the Mayapan specimens stylistically belong toX or Chich6n Fine Orange as it has been defined (seeBrainerd, 1941).

The four-color fine orange polychrome found atChich6n Itza in Early Mexican stage deposits (Brain-erd, 1941, pp. 172-173) should be mentioned here be-cause of its abundance and dating importance in Vera-cruz archaeology. The five sherds found come fromdeposits which date from relatively late in the EarlyMexican substage. The lack of this ware at Mayapanand elsewhere gives negative (and therefore uncertain)evidence for its later persistence. I have been unableto discover evidence for any sequential position forthis ware in relation to X Fine Orange in Veracruzexcavation reports; they seem to be considered con-

temporaneous. The reversal in relative frequenciesbetween X Fine Orange and polychrome fine orange inthe Veracruz sites published thus far in comparisonwith Chiche'n Itza suggests that X Fine Orange comes

from a region not well sampled as yet. The rarity ofpolychrome and frequency of X Fine Orange traits inpottery from the Las Flores site, Tampico, Tamauli-pas (Ekholm, 1944, pp. 392-404), suggests that X FineOrange may have an origin not far from this area, butindividual designs and vessels there are not as closeto Chiche'n Itza specimens as are certain of thosefrom Isla de Sacrificios. This seeming anomaly isresolved when one remembers that Isla de Sacrificiospottery is almost certainly a collection of wares madeelsewhere-not the result of a single industry. The XFine Orange must have been carried in from the north.It is conceivable that a meticulous analysis of the dis-tributions of these wares might yield evidence as tothe route and other details of the "Toltec invasion"of Chiche'n Itza'. There are several distinct types offine orange which do not show the sort of interblendingof designs, techniques, and forms which serve as sure

evidence of their manufacture by contemporaneous

78


peoples in close contact; there may be several loci oforigin for X Fine Orange.

Plumbate pottery, associated with X Fine Orangeat Chiche'n Itza but in markedly lower frequencies(178 sherds as compared to 1,053 sherds of X FineOrange), has been so thoroughly discussed by MissShepard (1948) that no further discussion of the wareneed be made here. Evidence for some perdurance ofplumbate has elsewhere been cited: the unbroken jarfound among fallen stones from the Caracol tower.The jar may have been placed after the collapse of thetower, and it is difficult to imagine that this unsightlydebris could have been allowed to accumulate earlierthan the end of the Early Mexican substage, since ar-chitectural construction was continuing into MiddleMexican times at the nearby Monjas, and the Templeof the Wall Panels seems still to have been in use.The jar was more likely deposited during Early Mexi-can times.

Mayapan Fine Orangeware comes only from theLate Mexican substage deposits of Mayapan. Paste isin general similar to the other types. It may averagesomewhat darker in color; microscopic tests for pos-sible temper have not been made on it, and its namingas a "fine ware" is therefore tentative. In both formand decoration, its major vessel types are completelydistinct from those of Chichen Fine Orange. There isgood evidence that it comes from the Campeche-Ta-basco coastal strip. Various of its characteristicswere copied, and it exerted a noticeable influence uponthe local wares. It occurs in approximately the samefrequency as did Chiche'n Fine Orange in the Yucata'ncollections, and seems identical with pottery whichoccurs with major frequency in Campeche and Tabas-co collections. It, as well as the local Yucatecan ce-ramics, shows probably significant similarities toAztec III pottery. The presence in Mayapan collectionsof a number of fragments which closely fit the X FineOrange description demonstrates that this ware (orrather that types of fine orange belonging to the Chi-ch6n group) continued to be imported into Yucatan atleast as late as Middle Mexican times. There is cer-tainly no evidence here of the "Maya resurgence" sug-gested by earlier workers.

Forms

The forms of Yucatan vessels are of primary valuein chronological study of the pottery, in part becauseof the meagerness of decoration and the conservatismin use of slip color. As described elsewhere, formhas served as a subsidiary but at times a more defin-itive criterion than surface and paste qualities in sort-ing fragments into wares. Wares almost always in-clude several forms each, and forms seldom crossware boundaries in the ceramic repertories of singleregions (an exception is in the Florescent Slatewaresand Redwares); there is usually no question of theirrank in hierarchic classifications. And forms arecloser cultural diagnostics than wares; there is morevariety in forms, and they are therefore less likelyto be confused with similar but culturally unrelatedgroups. The disadvantage of the rarity of design inYucatecan pottery has been somewhat offset by thecomplexity of Maya ceramic shapes, which makestheir recognition from small fragments easier. Shapeshave been further investigated by comparative studyof whole vessels; every effort has been made here toreconstruct shapes when possible, and the classifica-tions have been made with the whole vessel as the de-termining consideration.

The validity of the use of similar shapes as evi-

dence of cultural connection has been demonstrated atnumerous stages of the work. Its value has been shownboth as a finer-cutting diagnostic through time thanare wares (as in the Medium Slatewares within theFlorescent stage, and between Florescent and EarlyMexidcan times), and in the demonstration of synchro-nism or near synchronism between potteries of simi-lar form but of more or less widely differing warecharacteristics (as among certain Formative and Re-gional collections). Although I believe that form, aswell as decoration, is generally inclined to rangethrough a short time span over a wide area, as opposedto wares, which, defined as they are by techniques andmaterials, are likely to hold a single, restricted re-gional position and last through longer time, it mustbe granted that this tendency is far from invariablein the Yucata'n material. A prime exception is docu-mented from the beginning of the Mexican stage; fromthis time on the vessel shapes show considerable simi-larity over the last 900 years, while wares shifted radi-cally three times during the first 400 years but changedlittle during the succeeding 500 years.

The close study of form has paid dividends in thedefining of synchronisms and interaction with areas

outside Yucatan, as well as in allowing finer subdivi-sion within that area. There are numerous illuminatingsimilarities between Yucata'n forms and those of otherareas, some at a considerable distance.

The uses of form described above are all primarilytechnical and analytic-they contribute to the buildingof the archaeological time-space framework. Theseuses are all made possible by the fact that ceramicvessel shapes were concepts in the culture of the Ma-ya. Vessels of varying form had uses, established bycustom, which evolved through history like other fac-ets of culture, modeled by changes in food preparationand usage, and by changes in those religious ceremo-

nies in which ceramics were used. Some of thesechanging fashions had local origins, others were copiesfrom a variety of outside sources. A great part of thisreport has been directed toward the analytic classifi-cation and documentation of these forms in time andspace. Some interpretation of their usage, and certainbroad generalizations as to their origins and develop-ment, will be attempted in the following sections.

Unslipped jars.-This vessel shape has the longest,least-changed tradition of any in Yucatan. Throughoutits history the form is globular, with round and un-

specialized bottom, thin and even wall, slightly con-

stricted neck, and texturing of the exterior surfacewith smooth interior. Unslipped vessels of this gener-al shape are very widespread among New World agri-culturalists and may perhaps be safely assumed tohave been used for boiling food wherever found. Thepots are adapted to this purpose by their globularshape which allows a maximum efficiency in heating,and by the uniformly thin wall and absence of special-ized base which prevents cracking due to thermalstrain. The characteristically wide but slightly con-

stricted mouth permits efficiency in heating with readyaccess to the contents for stirring and ladling. Thetexturing of the exterior surface, a widespread char-acteristic of cooking pots in many other areas as wellas in Yucatan, takes the form of striation. This tech-nique characterizes the unslipped jars at San Jose,Uaxactun, and Tres Zapotes during all their sequence,appears only from late Classic times on at Zacualpaand at Panuco, is sparsely represented in MotaguaValley collections, and is very scarce in Valley ofMexico ceramics. Thus it may be a Maya lowland area

trait which spread somewhat in late times. It is possi-ble that texturing aids in heat absorption, although I

79


know of no tests performed to check this hypothesis.Unslipped jars are the only cooking vessels iden-

tified among the Yucata'n pottery; the comal was nev-

er used there. This fact suggests that in this area

maize was eaten as a gruel throughout prehistorictimes. This custom still persists. Although the tor-tilla has become an essential part of the diet, maizeis most often taken as gruel.

The unslipped jar was replaced shortly before thetime of the Conquest by the cauldron, a more open-

mouthed vessel with an untextured exterior, althoughmost of the other local forms continued in use. Thecauldron is now universal in Yucatan as a potterycooking vessel. This change in the cooking vesselseems to have been one in form and likely in firingtechnique, but not in materials nor probably in form-ing practice. The cauldron of Late Mexican and Colo-nial times is buff in color as opposed to the gray ofthe earlier jars. This color change almost certainlymarks a change in technique which allowed more oxy-

gen in the firing. Modern Yucatan potters choose a

special tempering material for cauldrons and mix itin a different proportion, and there is some evidencethat this is an old custom.

Our earliest evidence for unslipped jars in Yucatanis in Late Formative collections. Information fromthe 1949 Campeche collections allow a few of the Yu-catan sherds from mixed collections from Yaxuna andChiche'n Itza to be placed definitely in that horizon(fig. 16, 1j 65, , 8-10). These jars have a gradu-ally curving rim-shoulder junction and a gently out-curving, round-lipped rim. Exterior striation, whichis of the "fine" variety, gradually fades out to the lip,and the rim interior bears fine diagonal to circumfer-ential striation.

Regional stage jars are also decorated in the mainby fine striation, and are globular. The most charac-teristic neck form has a sharp rim-shoulder angle,with short, flaring rim ending in a rounded, blunt, or

slightly thickened lip (this is called "notch neck" inthis report). Sometimes daubs or roughly drawn rimstripes of a pale gray (presumably organic) paint are

seen, and the interior of the rim is sometimes striatedhorizontally. Exterior body striation ends definitelya centimeter or so below the neck-shoulder angle.Several major variations in forms of Regional stagejars are apparent. The Oxkintok unslipped jars (fig.12,a) vary significantly in the presence of a steeplysloping zone bounding the notch neck as opposed to theglobular body contour of the jars described above.Some of these jars are striated, some are smooth.This jar form is close to that of slipped jars in thesame collections, and these slipped jars in turn are

probably related to other Regional stage slipped formsin Yucatan. Quite similar unslipped jars come fromthe Motagua Valley (Smith and Kidder, 1943, fig. 18).It seems reasonable to suggest that unstriated jarssuch as fig. 12, a, 6-12, 31, should be considered vari-ants of the slippedjars rather thanmembers of thegroup under discussion. In the Dzibilchaltun and Acan-ceh collections (Regional-Flore scent transition) anoth-er distinctive rim type is found (fig. 14, a). This typeis scantily represented at Acanceh (fig. 14, c,5-7).It is characterized by the same neck-shoulder junctionas on the notch neck rims, but the neck is gently out-curving and much higher, with an elaborated, thick-ened lip. A quite similar form is represented in my1949 collections from Xpuhil, Campeche, and there issome similarity to sherds from Early San Jose III(Thompson, 1939, fig. 46, a, b). The Dzibilchaltunform also comes in varying frequencies from Holac-tun, Uxmal, and Oxkintok (figs. 37, a, bj 38,f), sug-

gesting an early placement for the beginning of theFlorescent deposits at these three sites.

The Florescent-stage rims (figs. 38, -_e 68, e)are characterized by shorter, heavier necks than thelast-described group, and by heavier, more roundedlip thickenings, generally exterior in placement. Theneck-shoulder junctions are often rounded rather thanangular, and the neck profiles more markedly curvingand more nearly vertical in orientation. The striatedbody patterning is of the "coarse" type and usuallyends at or shortly below the neck-shoulder junction ina pronounced ridge, produced by smoothing the neckafter the body decoration had been completed. Thebody shape shows even curvature save for a faint an-gle bounding the convex bottom. This angle is followedby the striation patterning, which usually takes a sin-gle direction on the bottom and a chevron pattern onthe sides.

Data are scanty on unslipped jars of the Early andMiddle Mexican substages (fig. 68, a, bj 22, cL, 10-16,24-28), but are sufficient to suggest that the changesconsist of refinement and rounding of rim contour. Thethickened lip on these thinner, smaller rims is oftenin the form of an exterior bolster.

The jars of the Late Mexican substage (see fig. 93,a-i, for examples) show a more marked change fromtheir predecessors than do those of any other part ofthe time sequence. There is some evidence that thistype came into use shortly after the Coarse Redwarewhich has been taken as the Late Mexican substagediagnostic. The walls are distinctively thinner thanthose of preceding jars, the striation patterning moreirregular, the body shape probably more squat on theaverage. Rim-shoulder break is invariably sharp. Therim is short and straight, either flaring or rising ver-tically to a tapering, blunt or angular and slightlythickened lip. The mouth is somewhat wider in propor-tion to body diameter than in earlier jars. These jarsoccasionally bear simple zo6morphic lugs.

No unslipped jars are known for the Colonial period.As stated above, they were replaced by the cauldronswhich are first seen in the Late Mexican collections.

Unslipped bowls, basins, comales, cauldrons.-Un-slipped bowls, basins, and cauldrons do not show thelong enduring tradition to be seen in the jars. Theyrepresent fashions belonging to separate periods, eachof relatively short duration and each apparently repre-senting an innovation without direct antecedents inYucatan. A group of forms which are probably relatedto each other occurs in the early Oxkintok collections(fig. 12, b,c) and at Yaxuna (fig. 1, , -10). Theseare large, heavy, flare-sided basins which seem tohave had ring bases. Their form is close to that ofassociated slipped vessels. This shape is known fromSan Jose', British Honduras, and from the Motagua Val-ley, but seemingly not from Uaxactun. Smaller, thinner,unslipped bowls come from Yaxuna (fig. 11, b, 1125-27) and basins from Coba (fig. 1,f). The vegetaltempered basins from Coba and Yaxuna (fig. 1,e) areof an unslipped ware unique in Yucatan, and are thusfar uncorrelated with any known craft tradition, al-though rim and body form is well within the localrange. There is evidence that heavy, flare-sided basinscontinued to be made during the Regional-Florescentoccupation of Dzibilchaltun (fig. 14, d, 42-46), and iden-tical forms come from Uxmal (fig. 37, c) and Kabah(fig. 39, e,1-14 &,2-5), which occurrences documentthe form into the Florescent stage.

In Early Mexican times small unslipped bowls weremade, supported in some cases on long, ball-endedlegs of Mexican mainland "Mixtec" style (see fig. 70,jj. With little doubt, this form was introduced from the

80


Mexican mainland for ceremonial use. These vesselsbear postfiring paint, and were supplanted by the sim-ilarly formed Coarse Redware bowls of Late Mexicantimes which have been found, filled with balls of copal,in the Sacred Cenote (see figs. 95, 96).

Comales were exceedingly scarce in the Yucata'ncollections. The only two fragments found (figs. 66, g,2; 97, k) were probably of Early Mexican substagemanufacture. As stated elsewhere (fig. 97, caption),the custom of using pottery comales is documentedfrom the south, from whence these vessels may havebeen carried.

In either the Middle Mexican or, more probably,the Late Mexican substage, large basinlike unslippedvessels seem to have been first produced in smallquantity, and the same form, the cauldron, has per-sisted until now (figs. 22, c, 29-33; 33, 2 70, k). Theintroduction of this form may be considered evidenceof a major break in the cooking-ware tradition of thePeninsula, which only took final effect after the Con-quest, as evidenced by the complete absence of un-slipped striated jars from the Colonial Mani deposits.The cauldron may have come in with the Nahua troopswho by tradition were brought into Mayapan; other evi-dence of western influence during this period is evi-dent. The influence of European iron cooking vesselsof this form must also be considered as a possible in-fluence in molding the shape of modern cauldrons.

Unslipped ceremonial wares.-The ware assumedto have been used for ceremonial purposes in Yucata'nis characterized by several features: (1) it is all un-slipped and unstriated; (2) it is frequently decoratedby applique'd elements; (3) it usually shows evidenceof white postfiring paint (either lime or clay), whichoften bears additional overcolors. The form used dur-ing Regional and Florescent times was a flare-sidedbowl, commonly supported by a trumpet base; Mexicantimes were characterized by variations centeringabout the earlier type and the introduction of the in-cense ladle or blower, and rare "Mixtec braziers. "

During Late Mexican times the bowl-shaped incensariowas replaced by vessels bearing an attached humanfigurine with mold-pressed face and applique'd, hand-modeled body and ornaments. All of these vessel forms,as well as the ware and decoration, are established as

of ceremonial use by multiple evidences in pre-His-panic art, modern Lacandon usage, and archaeologicalassociations. I believe that this evidence is conclusiveand generally accepted, and therefore it need not bereviewed here.

We have no record of Formative incensarios fromYucatan. Although our collections are not large enoughto make the negative evidence conclusive, it seems

certain that the incensario could not have been an im-portant form. It seems to be a well-developed featurebeginning with the earliest phase (Las Charcas) andcontinuing through all the pre-Classic phases of theGuatemala highlands (Kidder and Shook, 1946; Borheg-yi, 1950, 1951a, 1951b). These tall, lidded, side-flangedincensarios are different from the traditional Yucata'nproduct, as are the Formative Monte Alban incensarios(present from Monte Alban I on) forerunners of thefigure type which continued this long tradition in Oaxa-ca and which must have inspired the form which sweptinto Yucatan in Late Mexican times. It seems likelythat both the vertical flanges and cylindrical vesselforms of these late Yucatan products have been drawnfrom the Guatemala highlands, and the attached figu-rines from Oaxacan traditions through the interveninglowlands of Tabasco and Campeche.

The "hourglass" type of incensario used during theRegional-Flor escent transition and the Flor escent

stage (fig. 104, a, b) is close to those of the MotaguaValley (see Smith and Kidder, 1943, fig. 22, 2,d,fig. 40, a-g), and shows similarity to specimens fromTres Zapotes (Drucker, 1943, fig. 41, Lh i, j, m), butless similarity to the forms from San Jose' and to thethus far published Uaxactun forms. This sort of distri-bution, covering most of the Maya area but missingthe 'Central Area," seems to apply to several of theFlorescent stage traits of Yucatan.

The Early Mexican incensarios (fig. 104, c) seem

to represent, for the major part, a continuation of theFlorescent tradition. The hourglass incensarios of thissubstage are larger and usually lack appliqu6d spikes;this form seems to have lasted into Late Mexican times.The ladle-formed incensario or censer is first definite-ly present in this period (see fig. 39 for possible Flo-rescent examples) and has a tubular handle; local cop-.ies of the 'Mixtec type" brazier also appear (fig. 104,d, 3; L 5).

The ladle is known in the Huastec area from periodII on (Ekholm, 1934, p. 346), from Monte Alban II on

(Bernal, 1949, p. 18), from the Esperanza phase atKaminaljuyu (Kidder, Jennings, and Shook, 1946, p.208), and from the Motagua Valley (Smith and Kidder,1943, fig. 41). The tubular handle begins in the lateClassic horizon at Zacualpa, Guatemala (Wauchope,1948, pp. 126, 150), and is found at the Toltec (plum-bate) horizon at Tula (Acosta, 1945, p. 37). In Yucatanthe ladle censer began in Early Mexican times andlasted through the Late Mexican substage.

"Mixtec type" braziers are characteristic of theplumbate horizon Toltec period in the Mexican andGuatemalan highlands (Dutton and Hobbs, 1943, pp.101-102; Wauchope, 1948, pp. 148-150; Woodbury,1948, p. 122; Acosta, 1945, p. 37). In Yucatan theyseem to have appeared at the same time and to havedisappeared before the Late Mexican substage, as evi-denced by their absence in the Mayapan collections. Ihave found no reference to their appearance in Aztec-period deposits in Mexico. Thus both the ladle censer

and the 'Mixtec type" brazier must have come to Yu-catan from the west, probably from the Mexican high-lands.

The high, bucket-shaped, flat-bottomed incensario(fig. 104, c, 1) is harder to place. A suggestion of sim-ilar form is found in a fragment from Dzibilchaltun ofthe Regional-Florescent transition (fig. 19, b, 2), andin Rio Bec region sherds which I have not yet published.There is a marked resemblance to published vesselsfrom Tula (Acosta, 1945, fig. 20, n) and to a vessel as

yet unpublished. A generally similar repertory offorms comes from plumbate horizon tombs at Nebaj(Smith and Kidder, 1951, figs. 79, a; 80, a-c) in thenorthern Guatemala highlands.

A smallish, globular to barrel-formed incensariowith flare lip and low trumpet base, thumbed filletsand appliqued disc ornaments (fig. 104, 4, 1, 2), appearsin the Late Mexican substage, during which time itwas replaced by figurine incensarios. An incensariovirtually identical with this Yucatan type is said tohave come from the Motagua Valley (Smith and Kidder,1943, fig. 23, i); others from Champoton, Campeche,are illustrated by Ruz (n.d., pl. 60, 51-56).

I have already pointed out that the figurine incen-sarios introduced into Yucatan toward the close of theLate Mexican substage seem to be of Mexican-Guate-malan highlands inspiration, as evidenced by theircylindric form, vertical side flanges, and applied,hand-modeled figurines. The documentation is pooron the eastward journey77 of this type through the area

south of the Gulf of Campeche. Specimens includedunder Drucker's Lirios-type figurines (Drucker, 1943a,

81


pp. 83-85) show close enough similarity in technique,as well as in scale, subject matter, and treatment ofdetail, to make me suspect that his type should datecontemporaneously with my Late Mexican substage,rather than 500 years earlier in Upper Tres Zapoteswhere he has placed them.78 Lirios-type figurines maythen be acceptable as a needed link in this eastwardsweep. There are many undocumented or only region-ally documented museum specimens made in the char-acteristic orange clay of the Campeche coastal area,and Ruz (n.d., pp. 150-151, 370-373) reports figurineincensario fragments mixed with sherds of the globu-lar to cylindrical type described above at Champoton.This situation is similar to that at Mayapan, exceptthat in the Mayapan deposits the replacement of thecylindrical to globular type by the later figurine in-censarios can be documented stratigraphically. A pos-sible historical placement for the introduction of figu-rine incensarios to Yucatan may be given by Landa'sdescriptions of the introduction of Mexican troops in-to Mayapan from the Aztec garrison at Xicalango (Toznzer, 1941, pp. 32-38, with notes).

There seems but little question that the modernLacandon incensarios (see fig. 100, 1) belong to thecontinuing tradition which is documented above. Theintervening stages in the Colonial sequence would notbe expected in the Colonial church sites which we ex-cavated, but may be represented by specimens in theMe'rida collections (see caption, fig. 100). It is possi-ble that the modern Maxcanu whistle-making industrymay also qualify as a continuation of the incensario-making craft. hand modeling with applique and post-firing overpainting are still practiced there (see fig.93, bb, cc).

Study of the distribution and dating of incensariosin the Maya area should be of particular interest inthe tracing of Maya folk religion. This type of vesselis nearly universal in the Maya area, and shapes, tech-niques, and styles are unusually coherent, widespread,and easy to trace. This is the pottery ware which pre-sents the best possibilities for correlating archaeologywith Maya religion.

Unslipped bottles. -The unslipped, narrow-mouthed,bulging-lipped, pointed-base jars of the Early Forma-tive deposit at the Mani Cenote (see fig. 30, c) are

unique thus far in the Maya area, save for a few some-

what similar neck sherds found in Middle Formativedeposits.79 This vessel must unquestionably have beenused as a portable water container; the small mouth,diameter about 255 mm., would make other use diffi-cult. Capacity is estimated at about 13 liters (3 gal-lons) and weight full about 13 kilograms (30 pounds),a one day's allowance of water for a family. The onlyother narrow-necked bottles I know of in the Mesoam-erican ceramic repertory come from the Upper Mid-dle Period Valley of Mexico site at Tlatilco, and fromWestern Mexico. These bottles, however, are muchsmaller, slipped, and markedly different in form. Thepresence of the bottle shape in the Mani deposit sug-gests a marked cultural break between this subphaseand all those which followed it. The tradition of jarforms is quite conservative for all later periods upto the Conquest. The paste and finish of this ware alsosuggests a disjunction or marked cultural change, atleast between this horizon and the Late Formativesubstage. Pattern burnishing is unique to these ves-

sels, and the paste texture is finer than in the suc-

ceeding unslipped wares of Yucatan.Slipped,iars.-This vessel is known from Forma-

tive times through the whole chronologic sequence,and is still manufactured in Yucatan. The Late Form-ative jar shape is not completely reconstructable

from our samples. The Holactun collection gives usour most certainly defined material (fig. 16, c b, 3,I-lo), and the Yaxuna (fig. 5, c) and Mani collections(fig. 31, d, e) add detail. The forms are in generalquite close to the Chicanel jars from Uaxactun (RicketVson, 1937, fig. 156, a, b Smith, 1936, fig. 2, 5-7) andto those of San Jose' I (Thompson, 1939, fig. 33). Theyare also quite similar to the Regional Redware jars(fig. 105, d). Differences (most clearly seen in theMani collection-fig. 31, d) are in a more graduallycurved rim-shoulder junction, less shouldering of thebody, and undifferentiated or simply flattened base.The incised shoulder designs of the Mani jars (fig. 31,e) are very close to those from Uaxactun (Smith, 1936,fig. !-6j Ricketson, 1937, fig. 150). Other variationsin Formative Monochrome jars among the Yucatansites should, with more material, almost certainlyshow subdivision of the period, and a closer chrono-logical equation with the Chicanel phase.

Regional Flaky Redware jar forms (see fig. 105, a,_1, .j) show a striking, and seemingly rather ephemeral,break in the tradition of jar forms in Yucatan. Thecylindrical neck form with small exterior bolster lipis distinctive, as is the jar base with its discurvateexterior zone and dimpled bottom. A further distinc-tive feature is the presence of interior striation on

these jars. Regional Polychrome (see fig. 63) suggestslater changes in the jar form, those shown being theuse of a flat bottom and the addition of three handles.

Early Oxkintok Monochrome jars (see fig. 105, i.)are of uncertain derivation. The sharp rim-neck break,with short flaring rim, is not far from some LateFormative jars, but the rim is shorter and the breaksharper. The greatest distinction of Oxkintok jars,however, is in the presence of the long conoid neckconnected by an angular break to the low, stronglycurved shoulder.

The Regional Redware jars (fig. 105, c) with theirstrongly carinated shoulders look like a refinement ofthe Oxkintok Monochrome jars. Walls are thinner,dimpled bottom is mor e pronounced, and the conoidneck is convex in profile. The dimple bottom may bederived from the Flaky Redware forms; the other fea-tures seem to reach back to the Late Formative tradi-tion, perhaps via Early Oxkintok. Red on Thin Gray-ware of the Regional-Florescent transition (fig. 105,d) follows the earlier Regional forms, but with longer,sometimes outcurved lip.

Medium Slateware jars show a marked break intradition from the various Regional jars in form as

well as ware characteristics. In Dzibilchaltun, whichproduced our best Regional-Florescent sample, thejars are Red on Thin Grayware, suggesting that slate-ware jars were not used until later than other slate-ware forms in this area. The slateware jars were

made in two size ranges, one larger, one smaller thanthe Regional and transitional jars. The larger jars(fig. 105, f) seem to have been made for carrying wa-

ter on the back and, as discussed elsewhere, were

more widespread regionally, and probably originatedearlier, than the small jar form. Capacity of the largejar ranges from about 30 to 60 liters (8 to 16 gallons),weight full of water about 35-65 kilograms (75-145pounds), a very heavy load. The small jars (fig. 105,g) hold about two quarts each. They have thickenedlips and each has a pair of opposed vertical loop han-dles of such small size as not to admit a finger intothe loop. These jars, by evidence of balance, construc-tion, size, and association, seem to have been used todraw water from the wells and cisterns of the siteswhere they are found.

The Florescent Medium Slateware and Medium Red-

8Z


ware jar forms were superseded by a new repertoryearly in the Mexican stage (fig. 106, a, 1-9). Of thegroup illustrated, fig. 106, a, 5, alone may owe someof its form characteristics to the Florescent ceramictradition. The most distinctive of the new characteris-tics is the tall, thin-walled, concave cylindrical neck,a form which has persisted until the present. The jarbodies are somewhat more squat than Florescentforms, strongly rounded, with an angular break to aflat or faintly concave bottom. Another common neckform is thicker-walled, wider and shorter, with ex-terior encircling grooves (fig. 106, a, 1). Jar size iswide in variation, without the marked division intosize ranges which characterize the Florescent stage.Handles appear on the larger jars, and are horizontalloops, of good size for hand holds, placed just abovethe center of gravity of the vessel. Without question,these new forms are copies of the X Fine Orange (seefig. 106, a, 3) and, to a lesser degree, of plumbatetrade pottery which is recovered from Chich6n Itzain association with the native forms. This markedtaking over of foreign forms is particularly interest-ing since the ware characteristics, as well as certaintechniques of forming and decoration, change but slight-ly from those of the Florescent stage.

Slipped platters.-This form is limited almost com-pletely to the Regional stage in Yucata'n; a few frag-ments come from Puuc sites. The vessels are large,ranging from 35 to 55 cm. in diameter, shallow, withrounded profile and outsloping rim. Most of them borelow ring bases, some may have had hollow tripod legs,some may have had flattened bottoms. The unslippedbasins at Oxkintok (see fig. 12, b, c) may be related tothese forms.

Chief collection is from Coba (fig. 2, a, b, and cap-tion); a few rims from Yaxuna may belong to the samegroup (see fig. 8) but may more logically belong withbasins and bowls. There are occasional exterior sub-labial flanges and interior encircling grooves. Theonly close similarities to these forms which I havelocated are to San Jose', where synchronisms aremainly with San Jose III, scarcer with San Jose II.This dating is supported by evidence from othersources. Smith and Kidder (1943, figs. 20, c; 21, a)illustrate vessels from the Motagua Valley which mayalso be related. Vessels of this size and shape inslateware were found at Xpuhil in the Rlo Bec area in1949, but never occur among the Yucatan slatewares.

At San Jose' the platter, or dish, as Thompson callsit, has a long developmental history (Thompson, 1939,fig. 93) ranging San Jose' II-IV. The Yucata'n occur-rences must represent an evanescent fashion at Coba,introduced, like its architecture and glyphics, fromthe south.

Slipped basins.-The basin form, in contrast to theplatter, has a long history in Yucatan, and may haveoriginated locally. The form is hemispherical to glob-ular with erect or incurving rim, which usually bearsan exterior bolster lip. Bottom is flattened in mosttypes. The vessels are rather heavy-walled; diameterranges 20 to 60 cm. and height from one-half to fullmeasurement of diameter. Interior and lip are alwaysslipped, exterior is slipped in some types.

Earliest basins are of Regional Flaky Redware(figs. 4, a, c; 6, c), and of early Trickle on Flaky Red-ware (fig. 7, b). These basins have rounded exteriorbolsters and exterior as well as interior slip. Theyare globular in form, with markedly incurving rim.The early Oxkintok Coarse Monochrome basins (fig.11, d-f;j 13, a) are of two types, one globular with anexternal bolster thinner than hemispheroid in shape,with a highly diagnostic rim which consists of an ex-

terior thickening or bolster just above an externalridge. The exterior of the vessel is unslipped andstriated below this flange, and usually bears a ringbase. From Yaxuna come Regional Redware basinsquite similar to the latter shape from Oxkintok (fig.8, e-g.), and from Coba' (fig. 2, c-e) comes an assort-ment, all of Regional Redware, including some of Ox-kinitok form but witl the sublabial ridge thumbed, andothers with bolsters but no sublabial ridge. The lateTrickle on Flaky Red basins from Yaxuna (fig. 7, c)bear a bolster much like that on slateware basins. TheOxkintok basins are in collections which show resem-blances to Tzakol Peten pottery, the Yaxuna forms as-sociated with ceramics identified as similar to Tzakoland Tepeu 1. A few similar basin rims from San Jose6,all thumbed, are dated later by Thompson, at San Jose'III-IV and San Jose' IV (1939, fig. 59, b, e), others fromUaxactun as Tepeu (Smith, 1936a, pl. X, 5, 6, 8). Thistype of rim shape also occurs sparingly in the Chenesand Rlo Bec collections. Of all these shapes, the clos-est to the slateware basins which follow are certainRegional Redware rims from Coba' (fig. 2, c, 4-10),Regional Flaky Redware and Trickle on Flaky Redwarefrom Yaxuna (figs. 4, a, 6, c 1-7), and OxkintokCoarse Monochromes (fig. 13, a). Of these, the Oxkin-tok form, although in collections which are earlierthan some of the others, is considerably the closest.This fact, like the basal break bowl shapes, supportsthe early Oxkintok ceramics as close relatives, mnorelikely ancestral than collateral, of the slatewares.

The earliest Medium Slateware basin forms canmost logically be sought at Dzibilchaltun. The collec-tions from there, although small and mixed with Mid-dle Mexican substage material (see fig. 15, e, and cap-tion), show several sherds (fig. 15, , 2, 4, 1) of atype rare in the Puuc collections but predominant inthe Chenes-Rlo Bec collections of Medium Slatewareand found in other Yucatan collections which show sep-arate evidence of early date. These are Acanceh (fig.21, , 8-15), Mani (fig. 32, g7,2-, 12), and Oxkintok(fig 43, d, 1, 4-6, 14-21, 28, 3- 3-_ 45-50).These rims, which I have called Chenes style, are ofexterior bolster form. The inner wall, just below thelip, is concave in profile, ending in an angular to sub-angular lip. There is sometimes a pronounced inwardbend of profile just under the lip. Slip sometimes ex-

tends exteriorly no farther than the bottom of the bol-ster, and usually does not cover the whole exterior.The bolsters average higher and thinner than typicalPuuc-style bolster rims; the greatest thickness is com-

monly near the top of the bolster.The Puuc-style Medium Slateware basin forms dif-

fer from the above by having the inner vessel wallstraight to convex below the lip. Lip is angular to sub-angular. Exterior is entirely slipped. Vertical loophandles, set just under the bolster, are common,whereas they are rare or nonexistent in the earlyform. The bases of both these and the early form are

flat to slightly concave, thick-walled at the basal edge,with vessel floor a continuous curve into the sidewalls(see figs. 41, d-Lf 43, 44). Although all rims are ex-

teriorly thickened, some are of more elaborate formthan the simple bolster. Of these, the wedge bolsteris the commonest (see fig. 44, b, 33-40i). Occasionallythe rim is outbent (fig. 44, b, 43, d, 1. 3L 6). Other bol-sters are grooved or faceted (fig. 44, d, 16-23, il2-32)and a smaller number are decorated by incision, ap-plique', and specular red paint (fig. 44, a). In general,the elaborate rims belong to small, relatively thinwalled basins.

The Holactun slateware basins (fig. 53, c, d) whichare often found associated with the type described at

83


the Puuc sites above, are in general similar to it invessel form. Form is globular to hemispheroid; bothexterior and interior are slipped. However, the rimform is distinctive, as are the ware characteristics,and strap handles are mounted horizontally ratherthan vertically. The rim, which I have called T form,has a more pronouncedly concave interior profile thanhave Chenes-style rims, and the exterior bolster pro-jects horizontally and is shallow in depth. The top ofthe rim is flat, so that the rim, with its interior andexterior projections, takes on the form of the cross-bar of a T. The segregation of these basin rim formsto Holactun slateware is complete, in contrast to jarswhere the rim styles cross ware boundaries.

Early Mexican Medium Slateware basins are ingeneral similar in form to those of the Florescent pe-riod. There seems to be a tendency toward a higher,globular rather than hemispherical, form, with rimsmarkedly constricted (figs. 15, e, 9, 11, l2; 73). Bot-toms are flat to slightly concave, slip covers both in-terior and exterior. The form of the rims is diagnos-tic from Florescent forms; the lip angle is missing.The external bolster is evenly rounded at the lip andof nearly uniform thickness, and joins the exteriorvessel surface in an angle. It seems to have beenformed by folding over the top section of the lip ontothe exterior surface of the vessel. The two opposingstrap handles are vertically placed. Some basins lackbolsters. This characteristic is commoner in smallbasins. The rim is direct, lip usually blunt. Both bol-ster and direct rims are often ornamented by a seriesof encircling incised lines on their exterior face. Theeven spacing of these suggests the use of the kabal.

Middle Mexican Coarse Slateware basins (figs. 19,L, 6-9; 20, b., 14-L; 24, e) follow quite closely theform of the Early Mexican basins, but are readily dis-tinguishable from them on ware characteristics. Bol-ster rims are often squared at the upper edges. Oursample of this type is small; one sherd has a horizon-tal strap handle, a break from the Early Mexican style.

Late Mexican Coarse Redware basins (figs. 27, ,k; 94, g, 5-10) continue the earlier form tradition. TheMayapan sample shows no incised rims; the Chiche'nsample has one. One Mayapan sherd has a horizontalstrap handle. The slipped basin, unlike the slipped jar,seems not to have been made during Colonial times.It is conceivable that the cauldron took over its uses.

The place of the vessel form described above in thedomestic life of the Maya can, I think, be treated as asingle problem; the form persisted with but minorvariation for a thousand years or more. I suspect thatbasins were used for storing masa, the wet-groundcorn still used as their main staple by the modernMaya,80 as well as perhaps for the soaking of corn andlye which usually precedes the grinding. The shape,with its wide mouth, rounded bottom, smooth innersurface, and durable lip and walls, should make itvery suitable to this purpose. Capacity of these basinsranges from about 2 liters to 50, with an average ofabout 10 liters (10.57 quarts). Chart 24, showing thediameter range of these basins, gives an idea of thewide variation.

A modern estimate81 allows about 21 kilos a day ofcorn per Maya family, and the corn is normally grounddaily after overnight steeping. Although there are sev-

eral uncertain factors here, such as the modern use

of corn for domestic animals, and the degree of swell-ing of corn through absorption of water, it would seem

that the average basin would hold masa for a familyof five. The commonest modern receptacle for theday's masa in Yucatecan towns is an enameled washba sin.

Slipped composite profile bowls.82 -Relatively shal-low, broad-mouthed bowls having an abrupt change ofcurvature between side and bottom seem to be charac-teristic of the high-culture area of Mesoamerica fromthe Formative stage on, with the exception of the Ta-rascan and neighboring areas in western Mexico. Inother areas such as Central America, the northernAndes, the lower Amnazon, and the Hohokam area, theform is only occasional.

In late Formative collections the bulk of the ceram-ics recovered belong to flat-bottomed bowls with angu-lar to subangular base-wall junction and heavy walls,often with outbent (see fig. 107, a) or exteriorly thick-ened rim. One large hollow leg may have come fromone of these bowls (fig. 5, b, 41). These vessels, withtheir massive walls and hand-burnished, often groovedor incised finish, are the most easily identifiable oftheir horizon. Size varies far more than in any otherperiod. The finish is handsome, underside of base isusually slipped. Occasional rims bear plastic decora-tion. Contemporaneous with the above bowl form isevidence of one which may be the predecessor of theearly Regional Flaky Redware bowls: a few sherdsoccur with sublabial flange or ridges overlying a breakin wall contour (figs. 5, , i i 7; 31, c 44).

The Flaky Redware bowls (fig. 107, B) are shoul-dered. The shoulder in Flaky Redware is a slightlythickened angular break (see fig. 6, c); in incised di-chrome it often seems to represent an overlappedjoint between bottom, and the sides are built at a ver-tical or insloping angle upon it (figs. 6, g, ; 63, a, 1

or the perimeter of the bottom may curve sharplyinward and the wall rise flaring or outcurving fromit. The Regional Polychrome bowls have either breaksof the last type or "basal flanges" rising from a sim-ple angular bottom-wall junction (figs. 9, f 8- 30, a;63, a). The sidewall is short relative to those of theother types of forms considered, straight or outcurvedtoward the rim, slightly insloping to markedly outcurv-ing in orientation. Lips are rounded or, on Flaky Red-ware, exteriorly thickened (fig. 6, O). No certain infor-mation on bases is available for Flaky Red or FlakyDichrome; they seem to have been slightly thickenedand flattened on the bottom. All of the Regional Poly-chrome found shows low ring bases. The bottoms are

far more deeply curved than those of any other of thebowls of the composite silhouette series.

Regional Redware composite silhouette bowls come

from two widely separated regions: Coba' and Yaxunain the east, and Acanceh in the west of Yucatan (seefigs. 2, g; 8, jS 18, h). Although the eastern and west-ern wares differ somewhat, the shapes are markedlysimilar in the form of the basal ridge, the roundedprofile with faint S curve to wall, and the conoid legs(fig. 107, f). This east-west similarity is also marked-ly apparent in the jars of the same ware and sites,and both ware and forms are similar to Tepeu 1 pot-tery from Uaxactun. Miss Shepard remarks that some

of the Regional Redware from Coba' contains ash tem-per probably of foreign origin. All evidence, includingstelae and architecture at Coba', points to interinflu-ence between Yucatan and the Peten in these times(roughly 9.8.0.0.0-9.13.0.0.0 Maya).

Oxkintok monochrome bowl forms (fig. 107, , 15),although they presumably date earlier than the lastmentioned group, are closer in form to Florescentslateware. Bottoms are nearly flat, basal breaks are

angular; the basal projections, when present, are

usually a projection of the base, side walls are straightto slightly outcurving; hollow as well as solid conoidtripod legs occur. As has been said elsewhere, theOxkintok repertory is not closely related to other

84


known ceramic assemblages of its period. If we as-sume that Florescent slateware forms derive from it,we must assume that the close relationships shownamong Acanceh, Coba', Yaxuna, and the Peten werenot characteristic of all of the area, that striking geo-graphic variation, probably only to be explained bysharply separated political entities, also existed dur-ing regional times.83

Red on Thin Grayware bowls (fig. 107, 4, 1 2) areindistinguishable from the associated Medium Slate-ware basal break bowls in shape, although jar formsin the two wares are different. This suggests that thebowls in the two wares may have been used interchange-ably, the jars for different uses.

Medium Slateware basal break bowls are a form ofmajor importance during the Florescent stage. Theyare found in burials and caches as well as being plen-tifully represented in refuse. They were presumablyused to serve food, and probably for the offering ofsacrifices as well.

The typology of these forms (figs. 107, e; 108, a)has been discussed elsewhere (see caption of fig. 45).Although in the Puuc collections two more ornatestyles can be partially isolated from a more wide-spread, simpler style, the chronologic-regional affili-ations of the ornate styles are not yet clear. A fourthstyle (for discussion, see caption of fig. 67) is con-fined to north and east Yucata'n and is suspected ofhaving earlier beginnings than the Puuc styles, but islikely synchronic with them through a considerabletime span. The shape, with its nearly flat bottom sup-porting a rather heavy wall usually luted onto its up-per face, is distinctively Yucatecan of this horizon.Tripod feet are nearly universal, and a wide varietyis used, including solid slabs, solid and hollow conoids,spheroids, and cylinders, some mold pressed. Compli-cated rims and low basal ridges with thumb-pressedor stamped surfaces, and pendant skirts with cut ter-racing, are used. Bowl exteriors are sometimes in-cised. Organic paint is often used, red mineral paintis limited to elaborate forms. The only probable di-rect antecedent found in the Regional phase is the ear-ly Oxkintok style.

The basal break bowls of the Early Mexican sub-stage show influencing from imported X Fine Orangebowls from Veracruz, although influence in form hasbeen less than in decoration. These imported bowlsare of two principal types. The one has a deeply curvedbottom, often with an incised pattern on the floor, andthree spherical rattle legs set on constricted necks.The walls of this type are uniformly thin, without pro-nounced thickening at either the basal angle or the lip(fig. 108, b, 1). The other type is flat-bottomed withflaring, often outcurving sidewalls (fig. 108, b, 2, 3).Basal angle is usually markedly thickened and rimoutbent and thickened. Hollow tripod legs are cylin-droid with some flaring to the closed bottom. They arehollow with a pellet inside; the wall of each leg is of-ten pierced by a diagonal gash or slot. The influenceof the X Fine Orange forms is of a generalized nature,whereas the decoration was frankly copied. Influencewas concentrated on, though not restricted to, MediumRedware. An evidence of the generalized nature of theinfluence is the fact that the two above-described XFine Orange shapes cannot be isolated in the nativeYucatecan pottery, although vessels without thickeningat basal angle and lip (a characteristic absent in Yu-catan Florescent) are common (see fig. 108, c, 4, .)and many flat-bottomed bowls show outbent rims andother suggestions of X Fine Orange influence (see fig.108, c, 1-3). The cylindroid leg seems never to havebeen copied locally. The constricted neck of the X

Fine Orange spherical rattle legs is either absent ormodified. It seems likely that the Yucatecan potterscontinued many of their old practices of forming; anexample of this may be the method of attaching legs.X Fine Orange legs seem always to have been attachedby simple luting. Early Mexican subphase legs on na-tive wares were often, although not always, tenonedinto a scored circle on the bowl bottom, as they werein Florescent times.

There is but little information on composite silhou-ette bowls of the Middle Mexican substage. The samplefrom Mayapan shows no leg fragments from this form(fig. 24, a, b), but we suspect that two unslipped, appli-que'-modeled effigy-head legs from Chichen Itza (fig.92, e, 1, 2, and caption) may come from Coarse Slate-ware basal angle bowls of this period. In our restora-tions (fig. 108, d) we have shown no legs. Little com-ment can be made on this small number of sherds,save that they document the continuation of the vesselform through the Middle Mexican substage.

Late Mexican substage composite silhouette bowlsare smaller in diameter and much taller in proportionthan the bowls of earlier periods. Comparison withAztec bowl sizes (Griffin y Espejo, 1950) indicatesthat size and proportion are very similar to those ofAztec II and IV bowls, although Aztec III bowls arelisted as somewhat larger than bowls of succeedingand following periods. It is conceivable that if, assuspected, the composite silhouette bowls were usedas "service" vessels, this decrease in size may indi-cate a change from the custom of a group eating outof a common bowl to individual service. Form is notin general characterized by delicacy, although someof the bowls have a high lustre and burnished finish.The leg forms are more diagnostic of period than isthe body. Mold-pressed effigy-head legs are the mostelaborate form (figs. 27, g; 96, bj see fig. 95 captionfor a discussion of the distribution of this type). Highterraced slab legs also occur rarely in Yucata'n, andat Cintla, Tabasco, supplying a probable synchronismwith Aztec III in the Valley of Mexico (see fig. 94 cap-tion). A third type with wide distribution is hollowtruncated conical, with two or more perforations,placed in a vertical line on the external face of theleg (see fig. 94, h, 19, 20). This type is also found atTulum on the east coast of the Peninsula. Skirts arecommon, some of them notched, some in additionbearing geometric incised patterns (see fig. 26). Theskirts, animal effigy legs, terraced slab legs, and in-cised patterns are found also on Mayapan Fine Orangeand on other wares which, like it, seem to center onthe Tabasco coast (see fig. 28, a, b, c, and caption,also fig. 103, k), and both typologically and from his-torical accounts have Aztec cultural affiliations. Theirimportation and copying support the thesis that inLate Mexican, as well as in Early Mexican times,Yucatan was influenced in major degree from theMexican mainland.

No composite silhouette bowls seem to have beenmade in Yucatan since the Conquest, nor has any ves-sel of similar form or capacity replaced them. If itcan be supposed that this form was used for servingfood, the size of the vessels during various periodsmay be taken as clues to the number of people diningfrom a single vessel. Formative bowls range widelyin size, but those of following periods do not. This wasperhaps due to their manufacture for a specific familysize as opposed to standardized production by com-munity specialists in later times. Early Regionalbowls have a capacity of about 3 liters. Late Regionaland Florescent average about 1 1/2 liters capacityand Late Mexican bowls average well under a liter,

85


with Early Mexican transitional in size between Flo-rescent and Late Mexican. If we assume the eating ofsome sort of corn gruel as a staple, and this seems

likely in the absence of evidence for the tortilla, theMexican stage may have seen the vanishing of group

use of a single eating vessel, and the Regional stagea decrease in size of communal eating groups. Itshould be said that both modern custom and Conquest-period custom among the Yucatecan Maya is for menand women to eat separately, and that extended fami-lies of 50 people to a single group were common at

the Conquest, but a single couple with children are

the local unit today. As may be seen, there are sever-

al unknown and dubious factors to be considered inany reconstruction which might be attempted.

Slipped simple profile bowls.,-This form is definedas a vessel with profile in form of a simple, uninter-rupted curve to give a hemispheroid or lower shape,and of size which would permit use with one hand. Theform is very simple and nearly universal among ce-ramics, and thus its presence cannot be taken to indi-cate cultural continuity. Such cultural connections andtraditions as seem suggested will be discussed indi-vidually.

This form is present as a minor constituent in theLate Formative collections (figs. 5, a, 6-10; 16, b41,

j 65,c, 15), and Middle Formative collec-tions obtained in1949 suggest that it was commoner

in still earlier times. Shapes range hemispheroid toshallow. Bases to these vessels are not positivelyidentifiable but, from the absence of specialized basesin the collections, must simply have been flattened.Interiors were always slipped, exteriors often slipped.Simple profile bowls were not certainly made in FlakyRedware; if made, they were rare.

The form is common in Oxkintok monochrome (figs.12, d.,e,h;ky 13,

_ z__-; 103, 4) where it seems tobe divisible into two types, confined in major degreeto separate wares. Shallow, ring base, rounded pro-

file bowls, usually with blunt or rounded rim (figs. 12,4,

e h13, j, m) are of Oxkintok Coarse Monochrome;while flat-based, vertical-sided bowls, with a heavilycurved base-wall junction giving a somewhat squarishprofile, and with tapered rim (figs. 12, j k- 13,k, butof Coarse Monochrome, and 103, d), are almost exclu-sively of Oxkintok Thin Monochrome. These two formsare so distinctive and so definitely restricted to theirrespective wares that it is suspected either that sep-

arate groups of people made them, or that their uses

were quite different.Regional Redware simple profile bowls are nearly

as common as the more diagnostic basal-ridged bowls(figs. 2, g 8, f-h, k 18, e, g 19, h). Their body shapeis not far from that of the ridged bowls, but they lackfeet, the ridge, and the wall outcurve commonly foundjust under the lip. For the most part they show thesame tendency seen on the thin Oxkintok Monochromebowls toward a strong curve at base-wall junction,and nearly vertical walls. The bottoms are of constantthickness, and often bulge upward slightly at the cen-

ter (see figs.2, ., j7 8,h

,L3 ). Rims usually taper

toward the exterior. Slip normally covers the wholevessel, including the bottom. The forms of the poly-chrome bowls which accompany Regional Redware atCoba should be here mentioned. They have consider-ably thinner walls and are higher and more narrow,

approximating the Peten shapes which their decorationalso copies (see figs. 2, 3, aj4, , and Smith,1936a, pl. 13,14 1). No simple profile bowls of Redon Thin Gray are known.

The Florescent simple profile bowls can be sub-divided, the shape differences correlating quite close-

ly with differences among wares. The Florescent wares

in which this shape was made are: Thin Slateware, ThinRedware, Medium Slateware, and Z Fine Orangeware.The thinwares and Fine Orangeware may be related toforeign traditions, and their shapes are discussed in

the section on Thin and Fine ware shapes. MediumSlateware shows, among the earliest collections, whatseems to be an evolution from the Regional Redwaresimple profile bowl shape. This is best seen in theChenes and Rio Bec forms, but for illustration herethe Dzebtun collection and other scattered finds maybe used. The Dzibilchaltun Medium Slateware rims(fig. 15, i?) are valuable because of the lack of mixingof deposits. They show the profile and rim taper char-acteristics of the Regional Redware vessels. The two

Dzebtun atypical Florescent Redware bowls (fig. 35,

4,fLh) show the same features: f is as deep as theCoba polychrome bowls, and the Medium Slatewarebowl d is also of this general form. All show flattenedbottoms. Sherds belonging to this horizon at Acanceh(fig. 18, k) and Mani (fig. 32, 4) again show enough sim-ilarity to assume cultural relationship to Regional Red-ware forms. Also, certain sherds of Medium Slatewarefrom the Puuc sites seem to belong to this form (fig.49,c.9-15). This evidence, as well as the thus far un-published Chenes-Rfo Bec evidence, suggests thatslatewares must have been made contemporaneouslywith or not long following the Regional Redwares. Asomewhat dubious continuity between Regional formsand Thin Slateware is suggested by a few specimens.b2..., , from Labna, classed as ThinSlateware, belong to the form group discussed above,as do figs. 73, e, and 87,w, from Chichen Itza. All ofthese vessels lie in the range of Thin Slatewareswhich resemble the Medium Slatewares in slip andpaste, a poorly defined group.

The hemispheroid Medium Slateware bowls fromthe Puuc sites (fig. 49,a-i) are the next group in thisseries. They seem to have gradually superseded thetype described above; both shapes are found in theChenes and Puuc sites. The form style preponderantin the Puuc sites is distinguished by ring base, some-

times terrace cut, and by the presence of a thickenedlip with exterior sloping facet, which rises from a

gradually thickening wall. The form is evenly curving,

almost an exact hemisphere.This form is highly distinctive and, to my mind,

shows no close stylistic derivation from any whichprecede it in Yucatan, nor any descendants; the clos-est possible antecedent would be the hemispheroidbowls in Oxkintok Coarse Monochrome which showthe ring base, but no closer similarity, and which dateconsiderably earlier. The closest exterior relation-ships I know to the shape are the SanJose' IV and V

Redware incurving bowls (Thompson, 1939, fig. 71)which, although markedly close in form, lack the dis-tinctive faceted rim of the Puuc specimens. This formseems to be approximately contemporaneous (ibid.,p. 240) to the time at which we would date the Puucsites. The absence of the form from those of our 1949Chenes samples which are tentatively dated at

9.16.0.0.0 also corresponds to Thompson's San Jos6placement.

The form and size of this bowl are close to that ofthe gourd vessel used at Conquest times and still inuse for mixing and serving pozole and other sorts ofcorn gruels. The gourd vessel is used by the individ-ual as a drinking cup. The size of this vessel is aboutright for such use. Diameters range 12-20 cm., aver-

age volume is somewhat over a liter. It would seem

that liquid foods could conveniently be drunk from a

vessel of this depth, whereas shallower vessels would

86


be useful only for solid foods unless a spoon (unknownin this area) were used. If the identification of thisvessel as a cup is correct, it seems dubious whetherits restriction to the Florescent stage is a significantmeasure of a change in eating habits. Gourd bowls orcups were likely in use throughout Maya history.

The Early Mexican forms include vessels whichare classifiable as simple profile bowls, but they donot form a single, well-defined group. There are sim-ple profile bowls in Medium Slateware (fig. 74, a, 3-2, c) and in Medium Redware (fig. 87, u, x-cc; 88, a,c, 3). Most of these probably had flattened bottoms,some perhaps had hollow tripod legs; ring bases wereabsent or nearly so. In general, these vessels areshallow, without sloping rims, and with considerablerange of shape and form. There is a noticeable inter-grading between them and the associated basal breakbowls both in shape and size, which suggests that thetwo forms were probably interchangeable in use. Theyare suspected to have been influenced by Mexicanmainland forms, but are so generalized that searchfor antecedents is difficult.

A second group of the Early Mexican substage ismuch more tightly drawn and coherent. These are theso-called 'chile graters" (fig. 74, h-j). The shape islow with increasingly abrupt curvature toward therim. The rim is blunt to rounded, and incurved toslightly beyond the vertical. Tripod legs are spherical,with a pellet and slashed side. Covering the floor is aroughly geometric incised pattern, either done beforeslipping or on an unslipped area. The occurrence ofthis vessel type in the Huastec area during Formativeand all subsequent periods, and its occurrence in theChenes-Rio Bec area during the Initial Series period(see caption of fig. 74 for data on distribution), sug-gest that this form came to Chichen Itza from thesouth and probably to there from the Mexican main-land-a long and devious history. The chile grater con-tinued in use through the Mexican stage, but is notdocumented in Yucata'n after the Conquest, and thusits use remains unestablished.

Simple profile bowls continued to be made duringMiddle Mexican times (figs. 20, b, 24, b; 92, ., m),and during the Late Mexican substage (figs. 26; 94, h;95, 96). They seem to have dropped in popularity inLate Mexican times in favor of composite profilebowls which, by a reduction in size since Florescenttimes, had become functionally interchangeable withthem. Our Colonial samples (fig. 34, e) show a fewfragments (fig. 34, e, 9-12) which may have belongedto small shallow bowls, and a modern hemispheroid,flattened-bottomed bowl is made over Yucatan today(fig. 34, h).

Beakers and cylinders.-These vessel forms were

never important in the utilitarian ceramic repertoryof Yucatan. They occur mainly in the fine and thinwares and are much commoner in collections of wholevessels, presumably from caches and tombs, than inrefuse deposits.

There is some fragmentary evidence of beaker or

cylindric vessels of Flaky Redware, but not sufficientto give vessel form proportions (fig. 109, a). Severalof these pieces show more elaborate decoration thanaverage, indicating a tendency evident through the his-tory of this vessel shape. The Oxkintok Monochromesare singularly free of cylindrical or beaker forms.Only one sherd (fig. 12, j 1) may have belonged tosuch a vessel.

Fine Grayware and Thin Blackware (fig. 109, b),which are associated at Dzibilchaltun and, in lowerfrequencies, at Holactun, occur characteristically inbeaker form (figs. 28, e-g.; 35, b; 36, b,_ -g, ij 53,

e-X; 103, h, i). This shape, as it appears in Fine Gray-ware, has a height of somewhat more than half the di-ameter, an angular base-wall junction from which thewall rises vertically, gently outturning to a taper rimwhich occasionally is incised or hoop form on the ex-terior. Flat bottoms are commonest; another bottomform is convex exteriorly, with three spheroid rattlefeet; a third is double-bottomed, with pellets between.Blackware also occurs in beaker-shaped vessels, andboth wares occur in flat-bottomed basal angle bowls.These faintly outcurving walled bowls and beakers are

somewhat similar to Teotihuacan II and III forms, butother vessel details are so unlike as to make thechances of cultural interinfluence highly conjectural.The relationships of this form quite likely stretch toTabasco and lower Veracruz where the Yoxiha, Pie-dras Negras, and Tres Zapotes gray and blackwareforms show similarities. Unfortunately, the existingreports from this area do not allow chronologic place-ment. Dzibilchaltun Fine Orangeware is also relatedby use of a very similar group of vessel shapes (fig.59, g, h cf. fig. 36, h), as well as by a barrel-shaped,flat-bottomed vase (see fig. 59, a; 60, d, z; 103, m).

Z Fine Orangeware (figs. 36, .i 59, b-f; 89, e, 4which seems to have succeeded Dzibilchaltun FineGray and Fine Orangewares as a luxury tradeware,shows a marked difference in shape repertory. Thecommonest shape is a simple profile bowl, hemisphe-roid to parabaloid in shape, with shallow ring base andthin taper lip. The barrel-shaped vase has disappeared.Thin Slateware and Thin Redware from the Puuc sites(figs. 50, 51) are also made principally in simple pro-file bowl forms, but a few specimens of Thin Slatewarecome from flare-sided, flat-bottomed beakers (fig.50, ; 51, a, , -15) which sometimes have a hooprim. The atypical Thin Slateware sherds from Chiche'nItza (fig. 4, p) show a much higher incidence of flare-sided flat-bottomed beakers, and the Dzebtun collec-tion, believed to antedate the Puuc collections, showsan orange-colored Thin Slateware faintly flaring sidedbeaker which, however, has a rounded bottom (fig. 35,e). This shape of vessel also occurs in the Chenessites, notably at Santa Rosa Xtampak. This beakershape never intergrades into the outcurved wall typecharacteristic of Fine Grayware, nor into the associ-ated Thin Slateware simple profile bowl. Both on sty-listic and distributional grounds, it seems likely thatthis shape was in use earlier than the simple profilebowl, and there is probably some chronological over-

lap between them.The collections of elaborate Yucatan whole vessels

seen in Me'rida and elsewhere contain (see fig. 109,, d) a considerable number of beakers and cylindri-cal vessels decorated with paint, incision, carving,mold pressing, etc., which Vaillant (1927, pp. 76-87)includes in his carved, modeled, and fine slatewarecategories. These vessels in great part belong to myMedium and Thin Slatewares, but, as might be ex-

pected, show surprisingly few correspondences in mysherd material; they were used for ceremonial andmortuary purposes, and do not commonly occur inoccupational debris. Beaker forms are frequent inthis group, the diameter ranging slightly greater thanthe height, as it does in the above-described groups.There is considerable form range. Flat-bottomedvessels with angular bottom-side junction may bebarrel formed, or have cylindric or faintly flaringwalls. Lip may be straight or outcurving tapered, andis often exteriorly thickened. 'Collared beakers" oc-

cur, on which a wide, shallow neck section surmountsa globular body. Bell-shaped beakers are found. Someof these vessels have small, solid, tripod legs.

87


The forms of these vessels can in general be placedstylistically as early among the slatewares on the ba-sis of the Thin Blackware and Fine Grayware place-ments given above, with additional suggestive evidencefrom the simple profile Medium Slateware bowls, andfrom the fact that certain of these shapes are alsofound on monochrome vessels in the Merida collectionswhich, although they cannot be classified closely fromour excavated collections, are probably late Regionalin date. A further bit of evidence for this placementlies in a vessel (fig. 103, D which belongs in this groupon basis of form and decoration, and is reported tohave come from Dzibilchaltun. In seeking foreign inter-relations for these vessels, the Motagua Valley collec-tions of Smith and Kidder are by far the most sugges-tive. A considerable proportion of their Magdalenaphase tomb-offering ceramics are close to or identicalwith certain of the more elaborate Yucata'n vessels ofthe Regional-Florescent and Florescent horizons (seeSmith and Kidder, 1943, figs. 19, e-v 21, c-_, ., h25, a-d, g, h; 27; 284 b, c; 34, h; 42; 44, d; 47, d; 48,a; 52; 55, b). The polychrome of the Motagua collec-tions may be compared with certain Yucatan speci-mens in the Merida collections, and with the Labnaspecimens, fig. 3, b, c. The clue to the placement ofseveral of these shapes may be found in the Uaxactunceramics illustrated by Smith (1936b, pl. 13, 1-9),which he tells me are Tepeu 2 in date. The San Jose6IV blackwares also show the beaker form (Thompson,1939, fig. 73).

The pedestal base appears on a few Florescent ves-sels in Thin Slateware (fig. 50, c, , g) and in Z FineOrangeware (fig. 59, d, 4). The tumbler vessel form,and perhaps a rather awkward pyriform body shape(fig. 79, o), are found in combination with this base.These may be forerunners of the Early Mexican trum-pet-base cylinders and trumpet-base plain and shoul-dered pyriform vessels, which are prevalent in X FineOrange and plumbate (figs. 77, 78, 79) and were local-

ly copied (fig. 86). The distribution in time and spaceof tall, vaselike vessels with pedestal bases has notbeen worked out; I suspect that they came to Yucata'nfrom the southwest. This form of vessel may havelasted in much altered form until now, or, conversely,the Colonial and modern forms may have been reintro-duced (see fig. 34, e, g, for Colonial and modern pedes-tal vases). In modern Yucatan the wooden vessel usedfor mixing chocolate is in the form of a pedestal-basecylinder with small side handles.

The restriction of the vessel forms discussed aboveto the finer-textured wares, the small vessel size, theelaborate decoration, the evidence of trade, the con-stancy of use as tomb furniture, and rare presence inrefuse, all suggest luxury and ceremonial usage. Theextensive trade which must have been carried on invessels of these forms should make them more thanusually valuable in study of intercultural connectionsand discovery of synchronisms.

Unusual ceramic forms.-A considerable numberof potsherds in the collections do not belong in any ofthe form categories discussed above. These specimenshave in every case been illustrated, and in many casesare discussed individually in the figure captions.

It perhaps needs to be pointed out that the organiza-tion of this section on forms best fits the Florescentceramic assemblage, which is the assemblage we knowbest, and that other stages have been fitted to it whenfeasible. It is this fact, as well as our lack of know-ledge, that causes the paucity of description of Forma-tive vessel shapes. These shapes are in several in-stances radically different from the remainder of theYucatan sequence; the use of slip also contrasts inseveral respects, most striking among which is itsuse on only the exterior of wide-mouthed basinlikeforms, and on the bottoms of flat-bottomed bowls.More excavated material is needed for its understand-ing.

88

VI. YUCATAN CULTURE HISTORY IN VIEWOF THE PRESENT STUDY

This short section is an effort to turn away fromthe mass of potsherds and take some bearings on theultimate goal, the understanding of Maya history inYucatan. Most of the statements to be made here havebeen discussed at length at scattered points in thetext. I shall take advantage of this by minimizing qual-ifications in the hope of giving the account more con-tinuity and coherence.

Yucatan was inhabited by pottery-making peopleslong before the elaborate Classic Maya civilizationhad developed in the Peten lowlands. Although we haveno direct evidence of the beginning date of settlement,a rough comparison with Valley of Mexico ceramicstyles and radiocarbon dates suggests 1500 B.C. oreven earlier as a likely date for the earliest depositswe have found. It must be pointed out first that our ex-cavations have been limited to the vicinity of areasshowing masonry structures of obvious Maya religiouscenters-a highly selective sampling technique-andthat we have found no evidence of communities makingonly crude ceramics. We must suppose that such cul-tures as we have sampled have been preceded by pre-agriculturalists, followed by settlements characterizedby the beginning of agriculture and by primitive ceram-ics. More adequate sampling techniques may in thefuture produce evidence of such peoples.

FORMATIVE STAGE

Our earliest evidence of man, the Early Formativesubstage collections from the Mani Cenote, tell usthat these people hauled water in 3-gallon 30-poundloads, in pointed-bottomed, narrow-mouthed vesselsdecorated by pattern burnishing, a decorative techniquealso used by contemporaneous Highland Guatemalanpeoples (Brainerd, 1951, pp. 77-78), but very uncon-mon later in the Maya area. This pottery is smoothlyfinished, in a difficult shape to form, with delicate,restrained decoration-not the sort of craftsmanshipto be expected from beginners in pottery-making. Oursample consists entirely of water bottles, and thus wecannot be certain that slip and other decorative tech-niques found in the next substage, Middle Formative,were not also used earlier, but the very fact that wa-ter jars alone were found stands in marked distinctionto the later deposits. The Mani Cenote Late Formativedeposits include a wide variety of vessel shapes; radi-cal changes in uses of vessels seem indicated throughthis time lapse.

The Middle Formative substage was found at thetwo Chenes sites of Dzibilnocac and Santa Rosa Xtam-pak in 1949. This work has not yet been reported indetail.84 The two appearances of this horizon are inquite different contexts, and each gives information ofinterest. At Dzibilnocac, a deep refuse deposit wastested in a flat area at some distance from the near-est masonry ruins. The deposit contained leached ashin quantity, and randomly distributed and badly brokenbut unburned human bone. The impression given fromthe single sounding-excavation alone would confirmor deny this-is of a habitation site and suggests thatvillages, unknown during the Classic stage in the Ma-ya area, may have existed earlier. At Santa Rosa

[89]

Xtampak, under the eight stelae found in a group, therewas a plaster floor underlaid by an estimated threemeters of rough stone fill containing only Middle Form-ative pottery. This fill must be part of a masonry plat-form. Its size is unknown but may be large, since it issituated in an artificially raised terrace about 100 me-ters square, flanked by the largest pyramid of the site.Thus the building of substructures is documented forthis period. This activity implies co6perative laborfor nonutilitarian-probably religious-ends by a con-siderable number of people. This sort of activity is ahallmark of the Classic Mesoamerican civilizations,but has not been assumed to have begun at such anearly date. Throughout Mesoamerica, substructureswere used as temple bases, and in the Guatemala High-lands as coverings for elaborate tombs which are richin offerings of pottery. I have examined about a thou-sand whole pottery vessels from Yucatan, most of themfrom tombs or caches. Since not one Formative vesseloccurs among them, it seems safe to assume that richFormative tombs are rare or nonexistent in Yucatan,and that the Formative substructures there were usedonly for altars or temples.

The ceramic repertory of this period is extensive,including rare bottle forms related to Early Formative,and a host of slipped vessels including some formsand techniques in common with Late Formative pottery.

Late Formative pottery comes from nearly all sitessampled, from 14 in all. Exceptions among well-sam-pled sites are Labna and Uxmal, both in the Puuc area.Kabah and Sayil gave very light samples. Since few ofthese sites are so desirable geographically as to in-sure their occupation throughout all the time of humanoccupation in the area, it seems certain that singlelocations for religious centers continued in use overlong periods, and that formal Maya religion, which haslong been documented through the Classic stage, beganas early as Middle Formative times. Many Formativestructures must lie buried within the later structuresof Yucatan, entombed by the Maya custom of renovatingand enlarging religious buildings.

The only Formative substructure so far known fromthe central Maya area is E VII sub, Uaxactun, and thisis believed by some to date early in the Initial Seriesperiod. The pyramid is small (volume under 200 cubicmeters) and ornate, and if a temple was ever on itssummit it was made of perishable material. At Yaxunaa large mound (map 4, mound 8) was tested by fourtrenches at the base of its slopes. These trenchesproduced almost exclusively Late Formative pottery.The shape of this mound suggests that several mason-ry buildings had been built on top of it. The volume ofthe mound is approximately 50,000 cubic meters, aboutas large as any Classic stage Maya substructure savefor a few of the huge acropolis platforms. The size ofthis mound, if its dating may be assumed to be correct,attests an early peak of priestly control and religiousdevotion not unexpected if Dr. Kidder's theory as tothe pre-Classic dating of the huge Pyramid of the Sunat Teotihuacan be accepted. The presence of largesubstructures at this approximate time period is fur-ther demonstrated by the large mound of the Verbenasubphase, Miraflores phase, at Kaminaljuyu in theGuatemala Highlands (Shook, 1951, p. 98). This mound,


to which many structural additions were made, waspresumably used to support a temple or temples madeof perishable materials; large tomb chambers hadbeen cut into it.

Further generalizations on Late Formative culturecan be tried from the ceramics themselves. First,there is marked variation in wares, forms, and deco-ration among the various sites in Yucatan. This arguesfor either marked regional variation, or for a seriesof time horizons each with its own ceramic repertory.Of these two causes of variation I am inclined tochoose time as the major factor, my reason being thesurprising closeness of similarity between the major-ity of Formative Yucatan pottery and that from otherareas.85 Such homogeneity over the whole area is notconsistent with regional isolation within Yucata'n.

I believe that the workmanship of Late Formativepottery, and its range of style in shape and decoration,shows a greater inventiveness and attention in thecraftsmen than do the ceramics of later periods. Manyvessels are laboriously burnished to a high lustre. Therange of form and of size in each form shows morevariation; decoration is frequent and of sorts such aschanneling, incising, and grooving, which requiredskill. The wall profiles suggest that the pottery washand-modeled, often with great care. The vessel wallswith thick horizontal lips necessitate a firm clay andmuch patience in their forming. Although it is hard todocument this statement objectively, my present im-pression is that Late Formative pottery shows a ma-jor interest in the craft, probably by the population asa whole rather than by a group of specialists. Laterpottery is more refined, some of it is more elaborate,there is more evidence of individual technical virtu-osity; but this early pottery reaches an exceptionallyeven high standard. suggesting a wide appreciation of,and proficiency in, ceramic craftsmanship. Pottery-making must have been a major, not a minor, art tothe Late Formative Maya. The presence of MiddleFormative deposits in several, and Late Formativedeposits in most of the Maya religious centers sam-pled in Yucata'n, supports the architectural evidencegiven above that Maya religion dates from far earlierthan the Classic stage.

REGIONAL STAGE

The Yucatan Regional stage is composed of potteryassemblages from several widely distributed sites.The major slipped wares of the stage are aUl mono-chromes, most of them redwares, but they show a

striking variety in ware characteristics and thus incraft traditions; hence the name selected for the stage.At Acanceh, Flaky Redware is found in Late Forma-tive vessel shapes underlying the same ware with ear-

ly Regional vessel shapes. At both Acanceh and Yaxu-na, Early Regional Flaky Redware lies stratigraphi-cally below Regional Redwares which, although theyshow distinct ware differences, are closely compara-ble in form repertory. At Oxkintok, from under Fl-rescent deposits, comes the Early Oxkintok Mono-chrome assemblage, with unproved but likely associa-tion with the Oxkintok lintel date of 9.2.0.0.0. Thislikelihood is strengthened by contributory evidence,some positive and some negative, from other sitessuch as Acanceh. Relatively unnixed collections ofFlaky Redware and of seemingly related polychromesfrom Balam Canche, Yaxuna Cenote, Chac Cave, theHacienda Cenote at Chiche'n Itza', Cenote Ch'en Mul atMayapan, Mani, and E. H. Thompson's "Mound near

Merida, " add substance and suggest developmental

theories on the Early Regional substage. The Regionalsequences are further confirmed and fixed to some de-gree in the Maya calendar by Robert Smith's identifi-cation of sherds from the above-mentioned sites whichwere either traded from the Peten area or show closestylistic and ware similarities to Uaxactun pottery.These determinations equate Regional Flaky Redwarewith Tzakol 1 and Regional Redwares with later Tzakoland Tepeu 1 at Yaxuna. Red on Thin Grayware andearly slatewares equate with Tepeu 2 at Acanceh; theEarly Oxkintok assemblage equates with Tzakol andthus with Tzakol-like collections from Mayapan andMani.

The samples of the Regional stage are not large,and are from widely scattered areas. Three widespreadhorizons are in general recognizable, the first twoshowing developmental relations, the third seeminglyremoved from the local sequence and inspired fromthe south. Then there is the Early Oxkintok assemblageof thus far quite localized occurrence. The earliest ofthese Regional assemblages is almost certainly thatfrom the Yaxuna Cenote, where Flaky Redware isfound with Incised Dichrome and Trickle on Flaky Redas decorated variants. A number of stylistic traits andtransitional specimens link the dichrome to Tzakol 1-like polychrome (our Regional Polychrome), into whichit would seem to have evolved, and there is suggestiveevidence that in various other parts of Mesoamerica,as well as perhaps in Peru, a similar sequence in pot-tery decoration occurred (see Brainerd, 1948). Thusthe forms and ware variants of the group which we havecalled Flaky Redware picture a gradual cultural devel-opment from the Late Formative stage into a terminalhorizon which, from the close similarity between theYucata'n and Peten polychrome bowl design and shapes,can be dated in the Maya calendar as Tzakol 2 or 3,Maya date ca. 8. 16.0.0.0-9.8.0.0.0. At Uaxactun andSan Jose6, a cultural disjunction is thought to have takenplace during the horizon of this sequence, which ismarked by incised dichrome. Holmul 1 and MountainCow 2 have been suggested as examples of this hori-zon (Thompson, 1939, table 17), but both of thesephases have produced very small samples. This hori-zon may be placed inferentially in the Maya calendar.The Tzakol phase is estimated to have ended some-what before 9.8.0.0.0 and to have begun sometime be-fore 8.16.0.0.0; A. L. Smith (1950, pp. 15, 87) estimatesthe beginning of the Tzakol phase at 8.12.0.0.0. Sincethe incised dichrome horizon in Yucatan is believedto precede the polychrome horizon, which in turn isso similar as to assure its contemporaneity with Tza-kol at Uaxactun, then the Yucatan Regional stage, ush-ered in by the incised dichrome horizon, must havebegun well before the first Maya stela was erected-perhaps as early as 8.8.0.0.0.86

No descriptions of the architectural or other activ-ities of the makers of Flaky Redware can be givenwith certainty at present. The north building groupat Yaxuna was occupied during these periods, butwould need a careful dissection since its occupationseems to have been continuous from Late Formativethrough Florescent times. It seems likely, though atpresent unproved, that the north Pyramid at Acanceh,with its huge stucco masks, may also belong to thisperiod. The mounds near the mouth of Balam Canchecave (near Chich6n Itza) would also repay investiga-tion, although here also there was probably a slate-ware overlay. Another possible source for informationon this period is the plaza of Chuburna, a suburb ofMerida, where I have made a small surface collectionof sherds almost exclusively of Flaky Redware of theFormative-Regional transition. Although this site has

90


been extensively robbed of materials for local con-struction, masonry wall bases were observed in situ,seemingly about at ground level. This may conceiva-bly be E. H. Thompson's "Mound near Merida' (seefig. 63, caption). The evidence of the beginnings ofpolychrome in Yucatan, not thus far found elsewherein the lowland Maya area, makes this period in Yuca-tan of particular interest, since this would seem tobe the precise time during which the Maya Classicculture pattern was becoming established.

The Regional pottery from Oxkintok, as mentionedearlier, shows but little connection with other Yucata'nceramic assemblages. At the same time, the crafts-manship is so refined in form, firing, finish, and dec-oration as to suggest the product of a long, widespreadtradition. This tradition shows traits markedly dis-tinct in nearly every respect from the sequence abovedescribed, although the two (Oxkintok Monochrome andFlaky Redware) must have been contemporaneous overat least part of their spans. The attention at Oxkintokto surface finish may well stem from the Formativemonochromes, but transitional stages are lacking atpresent. There are a few similarities in form whichpoint to Regional Redware jars and basins; more posi-tive are the resemblances between Early Oxkintokbasal break bowls and this form in early slateware.The Oxkintok forms may well be those from which theslateware forms derive; they are certainly more like-ly candidates for that position than are the RegionalRedware composite profile bowls.

There is evidence that the Flaky Redwares wereused at Me'rida, Mayapan, Mani, and at Chac Cave inthe Puuc area. West and south of these areas the Ox-kintok Monochromes may have predominated. At Acan-ceh, although Late Formative Flaky Redware is prom-inent, Regional Flaky Redware and Tzakol-style poly-chrome are rare, and the trenches dug in front of theStucco Facade Building gave collections which, al-though very small, seem to document the presence ofthe Oxkintok assemblages there, perhaps contempora-neously with Regional Flaky Redwares elsewhere. Itis tempting to theorize that Oxkintok connectionsstretch to the southwest into Campeche and Tabasco,an area of little-known ceramics, although Ruz' recon-

naissance (examined by his kind permission from hisunpublished thesis, 1946) shows no Oxkintok-like pot-tery, but at the same time little evidence of ceramicsfrom the proper time horizons. We know too little ofthe place of Oxkintok ceramics in relation to the oth-ers of Yucatan. The sites of Bakna and Acanmul alongthe Yucatan-Campeche western coastal strip (see Pol-lock, 1940, pp. 266-267) might yield archaeologicalresults pertinent to these problems.

The architectural characteristics of the RegionalOxkintok horizon are somewhat uncertain. Buildingsof heavily plastered block masonry with nonspecial-ized vault stones underlie Puuc-style masonry at thesite (Shook, 1940, pp. 168-169). The Initial Serieslintel, dating 9.2.0.0.0, was found in situ in a buildingof this style, although it may conceivably have beenplaced there secondarily since the inscription is a

trifle too large for the exposed part of the lintel.87Trenches dug in the courtyard south of this buildingshowed mixtures of slateware under the only court-yard floor found, indicating that this floor was placedlater than the date of the Regional (slateware free)pottery samples. Thus it is quite possible that theInitial Series lintel building dates somewhat later thanthe early monochromes and the lintel but, accordingto the Uaxactun dated ceramic sequence, the potteryand lintel may well be contemporaneous and all threeare definitely pre-Florescent. The famous stucco

fa,ade at Acanceh is probably associated with OxkintokRegional ceramics. The small collection excavatedfrom an architectural fill of midden earth in front ofthe facade, and almost certainly postdating it, seems

to belong to this assemblage. The technique of execu-

tion of the stucco fagade-heavy plaster modeling-fitsthe technique of this period as found in the Peten.

The Regional Redwares, found at Yaxuna and Coba'in association with predominantly Tepeu 1-style poly-chrome, and at Acanceh progressing into a Regional-Florescent assemblage, definitely belong to closelyrelated cultural phases. Stylistic resemblances are

numerous and close, although the wares of the two re-

gions are distinct. This body of ceramics stands rela-tively distinct in its traits from either of the two tra-ditions described above; indeed, it stands closer tothe central Maya region than to anything else in Yuca-tan. The eastern manifestation of this horizon is withreasonable certainty associated with the major Clas-sic-stage architecture and glyphics of the site (Thomp-son, Pollock, and Charlot, 1932), which has been de-scribed by Thompson as a remarkably advanced out-post of the central Maya culture. The high frequencyof elaborately designed polychrome pottery from thesedeposits reinforces that judgment. Yaxuna has no

standing architecture of this period, perhaps becauseof depredations caused by Florescent-stage occupationand by robbing of stone which is still going on, but theceramics strengthen the case for communication docu-mented by the 60-mile causeway between the two sites.The Acanceh Regional Redware sample is much small-er and lacks the high frequency of polychrome potteryof Yaxuna and Coba. It may be that the spread of theClassic Maya culture88 was confined to the high forestzone which bisects the Peninsula along a northeast tosouthwest boundary line, although the presence of theChenes-Rifo Bec sites, which are definitely of a non-Clas sic pattern, within the fringes of the high forestzone show that the hypothesis in so simple a form doesnot fit.

The Coba architecture adds substance to our know-ledge of the eastern area for this horizon. A few ob-servations could be made at Yaxuna. A stela was found(fig. A), and heavy stucco fragments with carving andpaint were seen in trenches. Floor and platform sur-

faces were painted with what seemed to be a hematiteundercoat with a specular hematite finishing coat. TheAcanceh stucco fagade probably dates earlier than thishorizon (see above), thus leaving no architecture ofthe horizon documented in the western area.

The horizon for which the Dzibilchaltun collectionsserve as the most unmixed examples has here beencalled Regional-Florescent, because a monochromeslipped ware (Red on Thin Grayware) occurs in con-stant mixture with Medium Slateware, our best Flo-rescent-stage diagnostic. There is no sure evidenceat the site that a "transition" to slateware was takingplace through time, although it is reasonably certainthat slatewares succeeded Red on Thin Grayware gen-erally in northern Yucatan. It is quite likely that duringa considerable period of time in Yucatgn, both Red on

Thin Grayware and slateware were in use simultane-ously; perhaps each made in different areas or sites,and each preferred for certain uses and in certainvessel shapes.

The deposits at Acanceh allow a crossdating of theDzibilchaltun assemblage. There the Regional Redwareassemblage (mentioned above) is gradually replacedby a mixed Medium Slateware-Red on Thin Graywareassemblage which stylistically dates later than theDzibilchaltun assemblage. These assemblages showsigns of increasing popularity of slateware toward the

91


tops of deposits, but no deposits were completely freeof Red on Thin Gray. Robert Smith's identification ofthe polychrome sherds in these "transitional" depositsas mainly Tepeu 2 with occasional Tepeu 1 sherdsadds a welcome, though rather loose, anchorage to theMaya calendar. (Tepeu 2 dates from 9. 12.10.0.0 to9.19.0.0.0 as outside dates; 9.12.10.0.0 is Tepeu 1,9. 19.0.0.0 is Tepeu 3). Perhaps somewhere between9.12.0.0.0 and 9.16.0.0.0 is the most likely dating forthe Dzibilchaltun deposits. These deposits, with theirfine orangeware and fine grayware tradewares, shouldeventually be easy to crosstie over a considerablearea. Fine grayware is also found at Holactun, wherethe 9.16.0.0.0 date gives a clue, but not in the Chenes,where stela dates suggest that the same horizon shouldbe represented at Santa Rosa Xtampak. At Mani, whereslatewares very similar to those from the Santa RosaXtampak stela platform (date 9.16.0.0.0) are found,there is no fine grayware nor Dzibilchaltun FineOrangeware. Nor at these two latter sites is therefound Holactun Slateware, although it is common atUxmal. The dating and sequence given for this part ofthe chronology is thus still provisional because of un-certainty of definition between regional and chronolog-ic change in the ceramics.

Dzibilchaltun provides consider able information onthe nonceramic characteristics of this horizon in Yu-catan. The masonry is of block type, with undifferen-tiated vault stones. Numerous carved and painted plas-ter fragments in the pottery trenches attest elaboratesurface treatment in stucco. A surprising number ofstone monuments were located, 22 in all, of which sixwere carved but badly eroded and 16 were plain. Ac-cording to E. W. Andrews, to whom I showed them,the carved monuments bear no Initial Series dates.Andrews also took notes on the standing buildings, andhas made a brief statement (Andrews, 1942, p. 259).The architecture and glyphics merit further detaileddescription. The site plan (map 7) is notable for sev-eral quadrangular plazas with long ranges of buildingssurrounding them (note group 3, group 7, and two quad-rangles along the west side of group 5), and for thecauseways which connect three separate buildinggroups to the two central plaza clusters.

FLORESCENT STAGE

The Florescent-stage ceramics in this report are

defined as those in which the pre-Mexican slatewaresand closely related Florescent Redware constitute thewhole of the locally made, slipped pottery. The assem-

blages can be divided into two major groups whichwith reasonable certainty represent separate timehorizons, although further excavation is need to makepositive that we have not underestimated regional var-

iation-the other component in our analysis.The criteria for recognition of early slatewares

are discussed elsewhere. They have been recognizedin collections satisfying one or more of the followingcriteria: (1) they show a number of types either ab-sent or rare in collections from the Puuc sites; (2)the collections show other slipped wares believed ear-

lier than Florescent ceramics; (3) they are dated bymore or less likely association with early Initial Se-ries Maya dates; (4) the vessel shapes of the slate-wares show similarities to those of Regional mono-

chrome pottery. To these four criteria we are at pres-ent endeavoring to add determinations made by ana-

lyzing the characteristics of ceramics which can beplaced as early by means of statistical ordering tech-niques (see Robinson, 1951, and Brainerd, 1951). Al-

though there is evidence that the statistical orderingtechniques will provide more finely graded as well asmore objective placements for these early-late cri-teria in the Florescent stage, the work is not far enoughadvanced to be given in this report. There is much evi-dence that a collection wgich fits one of the four abovecriteria is likely to fit others, but also clear evidencethat the four criteria do not constantly apply to thesame collections.' This situation suggests that (a) morethan two time horizons showing sharp distinctions inthe ceramic traits used as diagnostic are representedin our material, or (b) regional variation with two hor-izons, or possibly that (c) regional variation with onlyone horizon is the causative agent.

Hypothesis (c) seems unlikely because of the asso-ciation of 9.16.0.0.0 dates with Early Florescent slate-ware at Santa Rosa Xtampak, and with considerablylater dates read for the Puuc, by stratigraphic evidencein the Chenes (see Brainerd, 1949), and by Tepeu 2-style polychromes at Acanceh contrasted with a fairnumber of Tepeu 3 stylistic resemblances with Puucceramics. I believe hypothesis (a) unlikely since cer-

tain of the traits show regional distributions which Ibelieve our samples are large enough to validate (highterraced slab legs are an example of such a trait).Therefore, my best hypothesis at present is (b), thatwe have two recognizable horizons with some regionalvariation.

The chronological placement of the beginnings ofthe replacement of Regional wares by slateware isearlier than 9.16.0.0.0 at Santa Rosa Xtampak, accord-ing to collections found under the flooring of the stelaplatform, and is well after 9.13.0.0.0 at Acanceh on

the basis of the Tepeu 2-style polychromes found therein a Regional-Florescent assemblage. If the Florescent-stage ceramic assemblage be assumed to have arrivedsimultaneously in the Chenes and at Acanceh, thisleaves a maximum of 60 years for the disappearanceof Red on Thin Gray, too short a time period. Thelikelihood of a simultaneous appearance of Florescent-stage assemblages over the whole of Yucatan is fur-ther lessened by the absence of Red on Thin Graywareat Mani, at the Chenes sites, at Oxkintok, Holactun,and at the Puuc sites where early styles of slatewareare recognizable. Also, the presence in Regional-Flo-rescent deposits of slatewares which show marked re-semblances to slatewares from Florescent stage de-posits in nearby sites, for example at Acanceh andMani, suggests that regional rather than chronologicalvariations distinguish them.

If it be assumed that the Florescent pottery assem-blage first arrived in the northern Yucatin plainthrough expansion from the Puuc area (a theory I be-lieve highly probable but not unassailable on presentevidence) and that Flore scent ceramics became uni-versal over Yucatan before the Mexican stage (againhighly probable but not certain without more testing),the dates and geographic spread of these first slate-wares become of peculiar interest, documenting, as

they must, the spread of a whole craft process withspecialized use of materials, techniques, and stylerepertory into new areas. The close correlations be-tween Regional wares and block masonry and betweenslatewares and veneer masonry further suggest thatthe appearance of slateware documents a general cul-tural replacement rather than a simple change in crafttechnique.

The Florescent collections showing early slatewaretraits are, first, those from the two Chenes sites,Santa Rosa Xtampak and Dzibilnocac, excavated in1949, which are not described in this report. The Manicollections show pottery strikingly similar in several

92


particulars to these Chenes sites, as do the OxkintokFlorescent collections, the Dzibilchaltun pottery, andsome of the Uxmal collections. The Merida State Mu-seum collection from tombs at Dzebtun, near Ceno-tillo in northeast Yucatan, is early according to thesecriteria, as are various of the ceramics from pre-Mexican collections from Chich6n Itza and Yaxuna.

Any statement on the architecture of this horizonmust of necessity be a poor makeshift until the resultsof Dr. H. E. D. Pollock's architectural survey, whichhas covered a major part of the northern Maya area,are available. The northern Maya area boasts what isunquestionably the most extensive as well as the mostelaborate mass of pre-Columbian architecture in exis-tence. The lion's share of it is Florescent in ceramicassociation and, because of its elaboration of structureand decoration, holds the same order of value for ar-chaeological analysis as do the ceramics, and a stra-tigraphy in many cases more certain and easier toanalyze.

Thus far in our study the architecture associatedwith Regional ceramic assemblages appears to be sim-ilar to that of the Peten area during the first part ofthe ninth cycle Maya calendar. Florescent pottery as-semblages, on the other hand, have been found only inarchitectural contexts of the Rfo Bec, Chenes, Puuc,Maya-Chiche'n types, and in the thus far unnamed typedifferentiated by Pollock west of the Puuc and Chenesareas in southwest Yucatan and northwest Campeche.89Although each shows distinctive tendencies in suchcharacteristics as masonry techniques, building form,decorative technique and style, and site planning, thesetypes seem to show a general cohesiveness as docu-mented by numerous traits common to two or morestyles of the above five. The relationships of Petenarchitecture of the latter half of the ninth cycle to theabove-mentioned northern styles which are contempo-raneous to it are quite generalized. It would seem thatLedyard Smith's substructure, wall, and vault masonrytypes (Smith, 1950, pp. 70-71) from the beginning ofhis Late Classic period (9.8.0.0.0) correspond in gen-eral outline with the northern types, but that the north-ern sites progressed distinctively farther in the devel-opment of thin, finely fitted veneer surfaces and wereunique in the development of beveled facing stones andtapered vault stones. Northern decorative techniquesand styles, building form, and site arrangement arealso distinctively different.90 This difference is rein-forced by Ruppert's finding, on the basis of architec-ture and glyphics, of what would seem to be at least apartial "cultural frontier " between Peten and Rlo Becstyle sites (see Ruppert and Denison, 1943, pp. 5-10),and by the fact that although the Xpuhil (Rlo Bec) pot-tery collections contain much slateware, the Uaxactunand Calakmul (Peten) collections which are contempo-raneous do not.

These findings lead me to the following hypothesiswhich is presented as a working tool, to be proved or

refuted by future finds: sometime before the midpointof the ninth Maya calendric cycle, within the centralpart of the Yucatan Peninsula, north of the Guatema-la-Mexico boundary but south of the Puuc, a cultureevolved which was characterized by slateware pottery,by a group of distinctive architectural traits, and bythe absence of the custom of erecting Initial Seriesstelae. This culture gradually spread and evolved un-til it engulfed the northern Maya area and influencedthe Peten as well as perhaps a larger area, becomingthe dominant culture of the Maya area until over-

whelmed by the incoming Toltecs in Yucatan. Such an

hypothesis would explain the ceramic finds in Yucatanmuch better than can the hypothesis of which Dr. Mor-

ley was long a staunch defender, of the movement ofcultures and peoples north from the Peten to developa "renaissance" in Yucatan. The usefulness of thishypothesis will depend in large measure upon the anal-ysis of the architecture within the northern Maya area;a "culture" cannot be reliably reconstructed from onlyone branch of human activity, such as ceramics. Theculture I hypothesize might be called Yucatec Mayain contrast to Classic Maya in the particular senseused by Proskouriakoff (1950, p. 1). Toynbee's previ-ous use of this coined word should not confuse the termas a descriptive label, and I know of no other use towhich it has heretofore been put.

The evolved Florescent stage is best representedby collections from the sites of the Puuc area in Yuca-ta'n. The collections of the three sites of this area fromwhich we have sizeable samples-Uxmal, Kabah, andSayil, as well as the smaller collection from Labna-are almost exclusively Florescent, and preponderantlyfully developed rather than Early Florescent. The Flo-rescent slatewares of both Chichen Itza' and Yaxunaoccur in some quantity but all in mixed collections, atYaxuna mixed with ceramics of Regional stage and atChiche'n with those of Mexican stage.

The general characteristics of the ceramics foundin refuse of this period have been detailed elsewherein this report. The impression given is one of excel-lent craftsmanship, but there is little sign of experi-ment. There is also a marked uniformity in the cate-gories and forms of vessels produced, and a consider-able economy in effort documented by an almost com-plete absence of small-scale decoration. The painteddesigns which do occur are cursive; sure but not me-ticulous. The graduation in wall thickness, particularlyin basal break bowls but also in other forms, seemsto be markedly functional, as though a fast-working,soft, wet clay had been used on a kabal and so formedthat in every possible case one continuous operation,without drying periods, could be used without allowingthe pot to sag. This means that areas where stressand strain would be caused from the weight of over-lying clay must be made thick, while areas not subjectto such forces can be made thinner without allowingslump while the clay is in the soft state necessary forkabal throwing.

These characteristics are those of an industrializedcraft, performed by well-trained, full-time craftsmenapprenticed under standardized, well-established work-ing methods. We have decorated pottery from Flores-cent-stage tombs, some of which shows individualityin shape and decoration, but there is some evidencethat the best of this may range Early Florescent;other fine pieces were imported, and the small num-ber of decorated fragments found in refuse are notnotable for artistry. The artists and innovators of thisculture must have worked in other media, architecturefor example.91 On the other hand, the technical compe-tence of these workmen was very high.

With little question the pottery industry was local-ized in various areas, each with its characteristic lo-cal materials and craft traditions, although the exactspotting of these manufacturing groups will need moreexcavation and technological work to determine. Athriving trade over this country seems indicated. Fur-ther study of the modern situation may aid the inter-pretation of the mechanisms of such trade.

There are indications of social stratigraphy in themarkedly variable frequencies of the thinwares in thecollections from various locations at Uxnal, and suchsocial stratigraphy is certainly documented in Mayaart of these times, as well as suggested by post-Con-quest documents. Differential use of locations within

93


Maya centers is also indicated by the above finds, bothon grounds of wealth and on grounds of religious ob-servance. Frequency of incensario fragments in col-lections, for example, is variable. These findings are

not new; they confirm characteristics long believed tohave been true of the Maya of these periods.

Some new light has been shed by the general natureof the ceramics in these collections. Almost all of theexcavations were made at temple precincts, or atleast on the edges of plazas where people must have

gathered for worship. The ceramics are utilitarianfor the most part, and include a variable but consider-able frequency of cooking-pot fragments. The Mayamust have camped at or near their plazas and temples.The mild climate of Yucatan makes a simple life rela-

tively comfortable, and it is not at all unlikely that

these deposits were made by people who had journeyedto the sites for one or several days of religious festi-

val, and who camped and ate near or on the ceremonialplazas. Although, as Pollock has recently pointed out

(in Jones, 1951, p. 177), we cannot be positive that a

townlike group of dwellings did not surround these

sites, the ceramic finds are consistent with the theoryof an open, widespread settlement pattern, with the

Maya living near their milpas and making 'pilgrim-ages" for worship during this period. My own examina-

tions of sites of the Regional and Florescent stagesrevealed no evidence of architectural or ceramic re-

mains near them in such quantity as would be expectedif the majority of the worshippers had lived nearby.A striking feature of the Florescent stage is the

large number of major archaeological sites in the

Puuc area during this time span. This area is verysparsely settled at present, probably due to the lack

of natural water sources. The Puuc sites are notable

for their strong development of water cisterns (chul-tunes), bottle-shaped excavations in bedrock which

drain the paved plazas under which they have been dug.I have estimated that at Uxmal the runoff water from

the main plazas alone could sustain a maximum con-

tinuous population of some 5,000 people through the

six-month dry season. Certainly Maya water storagenever approached perfect efficiency, but Maya cisterns

with their accompanying collecting platforms are

widespread through the country. They could doubtless

have afforded drinking, cooking, and wash water for as

many people as could have been supported by the non-

irrigated farming which was presumably used then as

now in the Maya area. The light population in this area

at earlier and later periods can with reasonable cer-

tainty be laid to a cultural rather than a natural defi-

ciency, the lack of either the knowledge or the desire

to dig chultunes, and chultunes can reasonably be

added to the traits of my provisionally named Yucatec

cultur e.

MEXICAN STAGE

The Early Mexican substage in Yucatan is notable

for the geographically restricted and yet rich nature

of our evidence for it. With the exception of scattered

sherds in mixed collections, all our Early Mexican

ceramics come from Chiche'n Itza. Here there are

several seemingly pure deposits, many deposits show-

ing varying degrees of Middle Mexican or Florescent

mixture, but no pure Florescent deposits. This is in

spite of the fact that the Monjas with its East Wing be-

longs architecturally there, that the substructure of

the Caracol is Florescent on the basis of potterycaches found in it, and the fact that we have large col-

lections of pottery from both.

The reason for the light sampling of the "Maya" or

pre-Toltec deposits at Chiche'n Itza seems to be attrib-utable to two factors; first, the heavy Mexican occupa-tion must have disturbed many earlier deposits; second,our samples come mainly from excavations done withthe prime purpose of architectural reconstruction,which thus did not often pierce earlier deposits. Theabsence of pure Florescent ceramic assemblages fromthe area of Chichen Itza'with which the Early Mexicansample can be compared makes the determination ofchronological diagnostics difficult. The Puuc collec-tions, which are quite close in general characteristicsto the Yaxuna and Maya Chiche'n slatewares, have beenused to isolate the diagnostics for the Early Mexicansubstage.

The most important ceramic innovation of the EarlyMexican substage was unquestionably X Fine Orange.This pottery, of fine texture and color and elaboratelydecorated, was imported from Veracruz in quantityand was enthusiastically copied by the local potters.The resulting Medium Slatewares and Redwares showa revealing blend of local and foreign features. Pre-

slip incising continued in the old manner as did anothertechnical procedure, the tenoning of hollow legs tobowl bodies, but new vessel shapes of several typeswere imitated often, producing intermediate forms,and the slipped and incised decorations were copiedbut rather crudely and imperfectly. There was an in-crease in the frequency of Medium Redware over theFlorescent stage, doubtless because it more closelyresembled the desirable X Fine Orange. There are a

few unusual sherds from Chiche'n Itza which are prob-ably Mexican imports, a considerable quantity of im-ported plumbate and of local new forms which must becopies of imports. The architecture and sculpture ofChichen Itza exhibit foreign influence to an even more

striking degree.In addition to the innovations in form and decoration

of the Early Mexican stage, there is a technical changein the slateware pottery from the waxy, translucentslip of Florescent times to an opaque white slip. Thischange probably took place gradually during the sub-stage rather than at its beginning, and it is quite pos-sible that various of the other innovations describedabove did not arrive simultaneously since no certainchronological sequencing has been determined withinthe substage.

In general, it seems that the ceramic Early Mexi-can substage began contemporaneously with the erec-

tion of the "Mexican Style" architecture, although we

have no sure ceramic evidence that the East Wing ofthe Monjas, for example, was as early as Florescent.The presence of Florescent pottery in caches in thelower platform of the Caracol dates at least this muchof the structure earlier than our Early Mexican sub-

stage, and this suggests that Mexican mainland reli-gious influence arrived earlier than ceramic changes.I believe from the ceramics that the Mercado andprobably all of its adjacent colonnades are Early Mex-ican in date; they probably date late in the substage.All of the Monjas complex dates no later than EarlyMexican, save for the East Building which was surelystill in use and possibly constructed during the MiddleMexican substage.

It seems likely that the changes in social organiza-tion and settlement pattern caused by the Toltec con-

quest were as striking as those in the arts and crafts.There is little doubt that the Puuc religious centerswent out of use due to this change. Thompson haspointed out that the Maya used words of Nahua originfor a defended town and for the administrative buildingof a town (Thompson, 1943, p. 23). The recent survey

94


of Mayapan has shown that it can definitely qualify asa town, a concentrated walled settlement 4.2 squarekilometers in extent, with an estimated 3,500 struc-tures (Jones, 1951, p. 180). I have stated above that,although proof is lacking, it is very unlikely that thePuuc religious centers were towns in the sense thatMayapan can be considered one. The likelihood thusstands at present that the Florescent Maya, living ina widespread fashion in the Puuc, were concentrated,after the Toltec incursion, into towns on the north Yu-catan plain. The two well-known sites of this period,Mayapan and Chiche'n Itz ',92 as well as Dzibilchaltunwhere a Middle Mexican settlement existed, are placedbeside cenotes, natural water sources. At Acanceh,where a sample from this period was obtained, no ce-notes were found in use at present, but the water tableis high and several of the wells now used may havebeen dug in pre-Conquest times. It is quite possiblethat the militaristic Toltecs concentrated the Maya inlarger settlements for easier control, choosing cenotesites on the northern plain for the easy availability ofa water supply sufficient for the large human groupsthey had assembled.

The Middle Mexican substage is a distinctive ceram-ic period firmly placed in time despite our relativelysmall ceramic samples. The diagnostic slipped ware,Coarse Slateware, is characterized by a coarse-tex-tured, rough-surfaced paste, with a slip which has thesame opaque white color as that of the Early Mexicansubstage. At Chichen Itza the latest architectural re-mains by stratigraphic placement at four locationswere at least occupied and possibly built during thisperiod. These four locations are at the Southeast Col-onnade, the Mercado, Temple of the Wall Panels, andthe East Building of the Monjas group. There is alsoconsiderable pottery of this period in collections fromfallen masonry at various buildings at Chiche6n Itza',but none in pure collections.93 At Dzibilchaltun, CoarseSlateware is distinctive of this period, but there are al-so a few Early Mexican Medium Slateware sherds in thecollections. At Mayapan, Coarse Slateware increasestoward the lower strata of trenches, but was found inminor frequencies, always mixed with Late MexicanCoarse Redware. The occurrence of a small amountof X Fine Orange at Mayapan shows that this ware con-tinued to be imported to Yucatin during Middle Mexi-can times, and thus argues against the isolation ofYucatan from the Mexican Mainland.

Architecture of this period is unknown, an unfortu-nate situation because it is known that there was a de-terioration in masonry between the Early and LateMexican substages. There is a very uncertain hint thatthe deterioration of masonry techniques had begun be-fore or during the Middle Mexican substage. The CasaRedonda Chichen Itza collection was tabuiated by Rob-erts. There was no Coarse Redware in the collectionfound within the structure, but it is uncertain whetherCoarse Slateware was present; the collection has sincebeen lost. The mud-mortared masonry found in theCasa Redonda may have been contemporaneous withsuperior Early Mexican masonry, and Tulum masonrywith good lime mortar is contemporaneous with ourLate Mexican substage which shows considerable useof mud mortar at Mayapan.

The estimation of the length of the Middle Mexicansubstage is based on two lines of evidence. The firstis the fact that we do not have extensive evidence ofan occupation of this period at any site yet sampled;the occupations of Chichdn Itzi and Mayapan are pre-dominantly earlier and later respectively. Anotherline of evidence is based on the amount of building atChich6n Itza and Mayapan, reinforced by documentary

accounts of the lengths of their occupations. If the11.16.0.0.0 correlation is correct, if the documentarydate of 1441-1460 A.D. for the abandonment of Mayapanbe allowed. if approximately one katun cycle is allowedfor each site (Mayapan and Mexican Chiche'n) and theEarly Mexican occupation is begun in accord with thedocuments at 10.8.0.0.0, only five katuns-less than100 years-remain for this substage. If the Middle Mex-ican substage lasted longer than our estimate, it seemslikely that the entire Mexican stage must be lengthenedto accommodate it. This could be done by beginning thestage earlier, either by changing to the 12.9.0.0.0 cor-relation or by beginning the Early Mexican substageat some date such as 10.3.0.0.0. With present evidenceno such course seems advisable; the Middle Mexicanstage can be assumed at present to have lasted for ap-proximately a century. Although the Early and LateMexican substages have been fitted to documentary re-constructions and to architectural styles, this horizonis known only by its ceramics.

The Late Mexican substage is of particular interestbecause concentrated work on this horizon is underway by the Carnegie Institution of Washington. Thediagnostic slipped ware of the substage is Coarse Red-ware, which bears a red slip, sometimes polished, ona paste very close to that of Coarse Slateware. The pe-riod is also distinguished by a clear-cut change in thesurfacing and form of unslipped jars, and by the intro-duction of figurine incensarios during the period. Thepresence of imported Mayapan Fine Orangeware con-firms the suggestion of western influence given by thefigurine incensarios. The ceramic evidence thus wouldnot seem to be for a period of "Mexican Absorption"or "Maya Resurgence" after the Toltec incursion, butfor continued influence, part military, part by trade,stemming eventually from the Aztecs. The figurineincensarios, found in surprising quantity, suggest awestern religious influence as well. These archaeolog-ical lines of evidence are well supported both by Mayaand Spanish documentary evidence.94 There is a likely,though slim, dating crosstie with the Valley of Mexicoin the presence of terraced slab legs, an Aztec III de-terminant, at Mayapan. Of interest is the clean-cut,though sporadically documented, distribution of potteryrelated to the Mayapan-Chichen Itza assemblage atTulum on the east coast of the Peninsula and farthersouth at Santa Rita on Chetumal Bay, the presence offigurine incensarios over a much larger area, and thepresence of orangewares very similar to if not identicalwith Mayapan Fine Orange at Cintla, Tabasco, and atXicalango and Champoton, Campeche. These distribu-tions conform neatly to the extensive coastal tradedocumented historically for the Spanish Conquest pe-riod and archaeologically for the Toltec period, andsuggest that cultural isolation did not accompany thedecadence of sculpture, architecture, and calendricswhich have perhaps unduly influenced previous charac-terizations of this period. The ceramics of Late Mexi-can times are technically good. The degree of theirornamentation is approximately equal to that of mostperiods in Yucatan, save for the Early Mexican sub-stage when design was copied from the ornate Vera-cruz ceramics then in vogue.

The excavations now under way in Mayapan willsoon furnish details on Late Mexican substage archi-tecture, since all the standing architecture of the maingroup seems to be of that date. Tulum adds furtherarchitectural detail; the few ceramics described fromthere, which are said to come mainly from wall mor-tar, show several Late Mexican characteristics (AngelFernandez, 1942). The Tulum and Santa Rita frescoesadd more substance to the documentation of this hori-

95


zon on the eastern part of the Yucata'n Peninsula.

POST-CONQUEST STAGE

The conservatism in the Yucatecan ceramic indus-try since the Spanish conquest is notable, despite thelong period of European contact and replacement ofpottery by metal vessels. The industry is organizedon a family basis, and there is some evidence that themarketing and distribution methods likewise beartraces of pre-Conquest custom. Most craft techniquesshow every evidence of being of pre-Conquest origin.Materials and tools are largely of local materials.The use of imported ochre as slip has a prehistoricparallel in the occasional use of imported specularhematite.

Vessel form and design show only minor Europeaninfluences. The displacement of the striated cookingjar by the smooth-surfaced cauldron is the most re-cent major vessel form change, but this change oc-

curred in the Mayapan period, probably influencedfrom the Mexican mainland. Small, deep, cup-shaped

vessels on flaring pedestals first appear in the ManiColonial deposits and are still made as a nonutilitarianform, but again this shape probably stems from theMexican mainland. The influence of the Toltecs was,without question, far more pervading on Yucata'n ce-

ramics than has been European influence since theConquest. This is not surprising in view of the natureof the two incursions. The Toltec way of life was muchmore similar to that of the Yucatecans than was theSpanish, and thus minor innovations could more easilybe fitted to local customs. It also seems possible thatthe local ceramic techniques and materials, like thoseof Maya lowland corn farming, were peculiarly welladapted to local needs and limitations, and thus couldnot advantageously be displaced. Conservatism alongmany lines, such as language and population, are strik-ing in Yucatan as compared to other areas in Mexico.Some of this conservatism is certainly due to the lackin Yucata'n of natural resources coveted by peoples ofhigher technological advancement. Yucatan has beenleft in a backwater which provides the archaeologistwith tantalizing glimpses of the past.

96

NOTESIJohn L. Stephens, an American explorer, published

the first popular book on the Maya ruins in 1841. Thisbook, entertainingly written and amazingly sound ar-chaeologically for its times, went through numerousreprintings in the United States and abroad. Encour-aged by its reception, Stephens chose Yucatan for asecond voyage, publishing his "Incidents of Travel inYucataTn" in 1843.

2For this reconstruction, compare Spinden, 1923,pp. 145-151, with Morley, 1938, Morley, 1946, pp. 72-97 and table V, and Barrera Va'squez and Morley, 1949.Morley was the last active field worker to adhere tothe above general outline.

3Vaillant's listings (1927 and 1935) of Yucatecan ar-chaeological sites as representatives of various pe-riods was based on the reported provenience of wholespecimens, and has proved of little value since sitesof several periods are found at or near almost everylocality listed by him.

4I have not been able to check the early stratigraph-ic placement of Peten polychromes at Kabah, perhapsdue to confusion in these collections since Roberts'determination, but have checked this sequence at othersite s.

5For a more detailed account of these data, seeBrainerd, 1941.

6See revisions in the dating of fine orangewares,and of the Holactun collection in this volume. I cannotrecognize the bases for Roberts' chronological subdi-visions within the Puuc and Mexican period ceramicsin his notes or collections, and they can most logicallybe considered as casualties of the time between ourstudie s.

7Another later version of this reconstruction is con-tained in Barrera Vasquez and Morley, 1949.

8I have elsewhere given an account of these discrep-ancies (Brainerd, 1948).

9Brainerd, 1942; also, see Andrews, 1942 and 1943,for descriptions of architecture dated by these excava-tions, Brainerd, 1948, for detail on the late periods,and Brainerd, 1951, for detail on early periods.

10The above reasoning, aimed at explaining the ab-sence of pottery from house sites in our samples, mayin some details be disputable. However, the fact re-mains that in spite of a systematic search of the envi-rons of each site, no domestic dumps were found.

"1This statement would be unnecessary but for Tay-lor's inaccurate allegation (1948, pp. 46-68) that underthe direction of A. V. Kidder the archaeological pro-gram of the Carnegie Institution in the Maya area hasbeen one-sided. Taylor comments as follows: "Car-negie has sought and found the hierarchal, the grandi-ose. It has neglected the common, the everyday. Andeven within its chosen segment of Maya culture, it hasproduced data for the most part in a narrowly descrip-tive range."

[97]

12Brew, 1946, p. 46, concepts 6 and 7, seems to bedefining the use of classifications in a later stage ofstudy than that discussed here, cf. Rouse, Modes (1939,pp. 26-27), or perhaps his house types do not allow ad-vantageous use of the hierarchic type of classificationso well suited to both Southwest and Maya pottery.Krieger (1944, pp. 276-277) also discusses these prob-lems of varied material. See further discussion of au-thor's viewpoint in Brainerd, 1951 and 1951a.

13See Beals, Brainerd, and Smith, 1945, pp. 87-89for an earlier statement of this plan of procedure.

14Descriptions defining Uaxactun Tepeu 1 potteryhave not been published at this writing. Material illus-trating Tepeu, but not subdivided, may be found in R.E. Smith, 1936 and 1936a, and in Ricketson and Ricket-son, 1937, pp. 270-283.

15Although Shook (1940) believes that the lintel is inits original position, it should be recorded that thecarved inscription on this lintel is slightly longer thanthe width of the doorway it spans, suggesting that itmay have been refitted to the building into which it isset.

16This section was written before the intensive ex-cavations of the Carnegie Institution at Mayapan, andhas not been revised. It is thus subject to correctionby later, fuller information.

17The shape of this specimen also equally resemblesthat of a Spanish olive oil jar.

18Unfortunately for the argument presented here,the bowl illustrated by Merwin and Vaillant as plate28, h, and dated Holmul IV, is also similar in shapeand design to those of the Thompson collection. Thisbowl seems atypical for its period. Can it have beenreused in a tomb?

19By Tzakol-style polychrome I mean Tzakol poly-chrome of the central Maya area, and pottery fromother areas, such as Yucata'n, which is very similarto it in form, color, and design.

20Prospectuses and progress reports of this surveywere published as follows: Pollock, 1931; Roberts,1931; Kidder, 1932; Roberts, 1933; Morley, Ruppert,and Bolles, 1934; Roberts, 1935; Pollock, 1935; Pol-lock, 1936.

21Of these, the largest assemblage from a singlesite is the Late Formative ware from Kabah. See fig.60 with caption.

221 hope that it will be possible to include close com-parative figures on pottery types between Puuc andChenes sites in a later report.

23A summary of the evidence from ceramics foundin the caves of the Puuc area by Mercer and Hatt isgiven in an article by Brainerd (1953). These caves,some of which are water-bearing, are not characteris-tically found near the large ruins. As might be expected,the cave time ranges exceed those of the sites, but ce-


ramic collections are small and mainly of Florescentstage. The cave of Chac, midway between Labna andSayil, presents a somewhat different picture. Thiscave, although difficult of access (see Stephens, 1843,v. 2, pp. 31-35), is still a water source of considerableimportance during the dry season. A ceramic collecotion made here by Edwin M. Shook in 1935 shows a

high proportion of Regional Polychrome, suggestingPeten Tzakol wares in color and finish (see fig. 63, Oj.This collection documents an early Regional occupa-

tion, suggesting that slatewares did not become domi-nant in the Puuc area until a time considerably laterthan the end of the Formative stage.

24See Kidder and Shepard, 1936, pp. 411-422, for a

study of the reactions operative in the use of organicpaint s.

25Beals, Brainerd, and Smith, 1945, contrast plate23, c and d.

26See Proskouriakoff, 1950, p. 4, figs. 1, 2, for thedistribution of Maya-dated monuments of this horizon.

27The potter's status in the Puuc must have beensimilar to that implied for Highland Guatemala in thispassage from the Popul Vuh: "Not for you shall be theball game. You shall spend your time making earthenpots and tubs and stones to grind corn" (Recinos,Goetz, and Morley, 1950, p. 161).

28The following climatic and agricultural data isgeneral and poorly documented. I have been able togather little concise information for this area.

29Roberts (diary, January 9, 1935) describes a man

living at Sayil who was carrying water from Chac Cave(1 1/2 leagues plus 425 meters of tortuous undergroundpassages each way) as the sole supply for his family.The man was carrying 2 gallons per day, probably inone trip. Stephens (1843, pp. 31-35) describes the carry-ing of water in gourds held in slings, so made as to al-low the men to crawl through the passages of this cave.

30Landa (see Tozzer, 1941, p. 289, under storage)seems to describe the use of chultunes for maize stor-age as well as for cisterns, but see Ricketson, 1937,p. 172, for the disintegrating effects of chultun storageon horn, wood, and steel. Although the characteristicchultunes at Uaxactun seem not to have been used as

cisterns (ibid., p. 123), their use for water storage inthe Puuc seems indisputable from evidence of theirplacement and structure.

31Average diameter of the Labna chultunes is 14feet, average floor area 150 square feet.

32For evidence of groups of Maya families, includ-ing married sons, habitually living in a single groupat the time of the Spanish conquest, see Roys, Scholes,and Adams, 1940, p. 14.

33See Ricketson and Ricketson, 1937, pp. 8-31, fordescription and discussion of this situation.

340ur early collections from Oxkintok, here pub-lished, contain little painted pottery, and their fewPeten affiliations are all to the Tzakol phase. I amthus at a loss to explain Ruz' statement. I have seen

many specimens from the region of the city of Cam-peche which resemble Jaina sherds and so can secondRuz' statement of similarity.

35Roberts' statement (in Pollock, 1937, p. 152) seemspuzzlingly at variance with his summary of wares givenon p. 151.

36These dates correspond quite closely to Thomp-son's reconstruction, scheme B (Thompson, 1941).

37The most recent publication in this series (Rup-pert, 1952), which has appeared since the writing ofthis section, gives much information of direct applica-bility to the ceramic studies and should be consulted inreference to this part of the report.

38Two fragments of plumbate pottery come from thetop levels of trenches at Uxmal, and another from thesurface at Oxkintok. A few grater bowl rims also oc-cur at Uxmal, and a single sherd of X Fine Orange.

39This conclusion is the author 's, supported by thefact that Vaillant's description of the one Coarse Slate-ware fragment which we can identify from his data(fig. 92, i) shows that he recognized the distinction be.tween it and Medium Slateware. See Vaillant, 1927, p.90, figs. 302, 371.

4OSee Tozzer, 1941, index and syllabus; Tozzer,1907, pp. 148-150; Thompson, Pollock, and Charlot,1932, pp. 3-5.

41Color terms, when capitalized, are from Ridgway,1912. These can be converted for use with the moregenerally available Munsell charts by consulting theconversion list published by Hamly (1949).

42A concurvate surface is here defined as a surfaceof a vessel on which a vertical plane taken radiallythrough the vessel, as well as a horizontal plane, bothcut curves which are concave on their inner surface;on a discurvate surface, in contrast, the vertical radialsection shows a convex profile on its interior surface.

43For San Jose I, see Thompson, 1939, pp. 72-74;and for Uaxactun, Ricketson and Ricketson, 1937, pp.231, 243, which shows, interestingly, a drop-off in fre-quency of this form in the two Formative phases, fromperiod IA to IB.

44See, for example, Scholes and Roys, 1948, pp. 34-35, for a discussion of the Conquest period. See discus-sion on captions for figs. 28 and 29 for evidences ofceramic influence through Mayapan Fine Orange andFigurine Incensarios, and Tozzer, 1941, p. 32, for his-torical evidence.

45Ruz (1945, pp. 68-70) describes what must be thisware from Xicalango, Tixchel, and Champoton in Cam-peche. See fig. 103 for a related type collected byBerendt at Cintla, Tabasco.

46For historical information on the relations betweenYucatan and the Campeche-Tabasco coast at and justprior to the Conquest, see Scholes and Roys, 1948.

47See Brew, 1946, pp. 67-73, for evidence againstthis hypothesis.

48See Brainerd, 1951 and 1951a, for more extendeddiscussions of various problems and viewpoints men-tioned above.

49For my views on method and validity of seriation-al techniques, see Brainerd, 1951 and 1951a.

98


50See Beals, Brainerd, and Smith, 1945, p. 115, fora design analysis of this type using variation in multi-ple factors. Direction of time changes was determinedby other means.

5IThis scheme was brought forth by Thompson(1941) as the one fitting most closely the Mexicanchronologies, which lean toward the placement of theplumbate horizon, linked to the Tula-Mazapan horizonin Mexico and to Toltec Chiche'n Itza in Yucatan, asending at least 150 years before the Spanish conquest.This scheme has been given strong support by the au-thor's excavations at Mayapan in 1942, which sampledlarge deposits all postdating the constructional periodsat Chiche'n Itza, and thus established beyond doubt thesequent positions of time spans A-B and B-C.

52See Tozzer, 1940, pp. 30-31, for discussion of thisevidence.

53The placement of the Uaxactun phases, as yet un-published, is given through the kindness of R. E. Smith.Placements are therefore tentative, and the author isresponsible for any errors which may have crept intothe chart due to his lack of critical judgment as to theUaxactun dating evidence.

54Cf. Wauchope, 1948, fig. 57, k-s, with our fig. 95,b, 1-6, 28, 37, and a comparative specimen from UaxacCanal, fig. 89, t.

55Maya names given are k'at or k'ut, with varietiesgiven by E. H. Thompson as zahk'at (white clay) andzhank'at (yellow clay).

56This suggests that the clay seen by Thompson mayhave been bentonitic. Bentonitic clays are well suitedto the retention of organic pigments; hence such a claywould be valuable for slip. See Shepard in Kidder andShepard, 1936, pp. 417-418, for descriptions of thisquality.

57See Rendon, 1947, pp. 119-120, for a fuller descrip-tion of tools and their Maya names.

58Mercer describes a kiln of brick with dry-laidbrick arches separating a fire chamber from the pot-tery area above it. Rendon describes stone with lime-mortar construction, with a k'ankob lining.

59There is a possible variation in certain clays,shown by the presence of small, dark-colored lumps,which is discussed below under tempering materials.No doubt there are also local variations in iron oxidecontent, since iron is often noted to be concentratedin veins through the clay, and these are sometimesremoved by hand. Miss Shepard believes some claysbore calcite other than that added in temper, and thereis adequate evidence of variety in the clays used forslipping.

60Although Shepard (see Thompson, 1939, pp. 264-265, and Shepard, 1951, pp. 243-244) suggests that anorganic coating may have been applied to slatewareslips, to my mind her experiments have not ruled outthe possibility that the organic substances she foundoriginated after the pottery had become buried in thehumus of the archaeological deposit. It is neverthelessquite possible that an organic coating was used.

61See Shepard in Thompson, 1939, p. 26, and 1940,p. 11, for mention of such an underslip, which was

found to be markedly calcareous in some instances, andthus may have been subject to solution by soil acids.This underslip served to heighten the brilliance of colorof the final slip.

62Cf. Beals, Brainerd, and Smith, 1945, plate 23, c,with plate 23, d, for example.

63This view is at variance with that of several writ-ers, who have assumed that in all cases a 'resist" ofwax or some similar pore-filling, water-resisting sub-stance was first painted on to form a design, followedby a water-mixed paint which did not adhere to the re-sist-decorated areas.

64See Shepard in Kidder, Jennings, and Shook, 1946,pp. 270-271, for a description and discussion of thispaint.

65fn 1942, a man near M6rida was successfully mak-ing untempered pottery from local clay. The man hadbeen educated abroad. Geographic determinism is notabsolutely certain!

66The ash temper determined in 26 sherds fromHolactun classed as Late Formative is questionablesince it may be due to sorting rrors. It is very pos-sible that volcanic ash te- .:per was used in the Forma-tive Orangeware found at Yaxuna. .No temper determi-nations have been made on this3 wve.

67Sherd temper 61 per cent, volcaaic ash 14 percent, calcite 24 per cent, clay lump 1 per cent; 186sherds in sample which may have included a few Flo-rescent-stage sherds through sorting errors, since allLate Formative sherds were found in mixed deposits.This is similar to the, situation in the Chicanel phaseat Uaxactun where 66 per cent of the temper is sherd,33 per cent calcite. No volcanic ash is listed for Chi-canel, but 10 per cent for the preceding Mamom phase(Smith, 1940). San Jos6 shows only 10-20 per centsherd; the remainder is mostly crystalline calcite,and volcanic ash is rare (Thompson, 1939, p. 75).

65Shepard, 1951, p. 243, notes that the ash in thispottery is not of Yucatan type.

69The classification of slips of Oxkintok monochromesinto the 'opaque" class is not completely clear-cut.Some are smooth-surfaced, probably burnished, andsome seem to have been unburnished and applied overa lumpy surface which makes their attribution betweenopaque and waxy difficult. Analytic work may providea more discerning classification of these slips.

70The main lack in the data is the uncertainty as towhether collections belonging to the same substage,from different areas and showing differing traits oftemper, slip, and paint, really represent regional dif-ferences or whether a time difference is also involved.

71A suggestion of rapid development may be drawnfrom the fact that during the short time since Mira-flores has been recognized, four subphases have al-ready been defined (Shook, 1951, pp. 97-99).

72See Shepard, 1951, p. 243, who points out that theYucata'n sherds are calcite tempered, but the Petensherds are ash tempered.

73The fact that the grading of vessel form and ofdecoration are consistent between the two wares sug-

99


gests a linear causative factor (see Brainerd, 1951,pp. 306-307) which is probably time, rather than a re-gional intergrading which would probably not be soconsistent between two unrelated traits.

74Wauchope, 1950, would seem to place this generalhorizon there. Also see Brainerd, 1951. Robert Smithbegins the Tzakol phase at Uaxactun at 8.12.0.0.0; thusIncised Dichrome, which I suspect precedes Tzakol 1on stylistic grounds, would distinctly precede the ear-liest Initial Series date of 8. 14.0.0.0.

75Wauchope's recent (1950) survey of pre-ClassicMesoamerican ceramics, published after this sectionwas written, seems to support my choice of stageboundarie s.

76I know of no New World archaeological evidenceother than in Mesoamerica for the manufacture andwide trade of luxury ceramics. If the absence of thisculture characteristic from other areas is real andnot due to incomplete study, it documents an importantorganizational advance in the Mesoamerican area.However, it is quite possible that such specializationexisted in Peru and perhaps even in the southeasternUnited States. In many areas where wares are simplein style, a technical study of pottery compositions isnecessary to establish a common origin for widespreadpottery types, and such studies have seldom been made.

77Which took it at least as far as the coastal areaof British Honduras; see Gann, 1900, pl. 32.

78Drucker may have been misled by postoccupation-al mixing of deposits. Drucker's San Marcos type,found mixed with Lirios type in Upper Tres Zapotesdeposits, shows definite stylistic similarity with theYucata'n Puuc figurines (see figs. 54-56 and captions).Since the Yucatan Florescent stage and Upper TresZapotes are dated as approximately contemporaneous,Drucker's placement of his San Marcos type is con-firmed by the Yucata'n evidence.

79Shook's report of sherds with similar decorationfrom Las Charcas and Sacatepequez phases in Guate-mala (see Brainerd, 1951, pp. 77-78) does not includeinformation on vessel shape.

85See Tozzer, 1941, pp. 89-90, and Villa, 1945, pp.54-55, for accounts of the preparation of maize amongthe Maya.

81Villa, 1945, p. 50. But pigs and chickens areraised by these people, and fed largely on corn.

82This term is an adaptation of Vaillant's compositesilhouette, altered because the term seems clearer.The contrasting term used here is simple profile.Composite profile is used here as a catchall for bowlswhich have in common a single angular change of di-rection in the wall profile of the bowl. In the detaileddescriptions of forms, more specific terms such asshouldered bowls. basal break, basal angle, etc., havebeen used.

83Such geographic separation is also suggested byRuppert's and Denison's survey in southeast Campeche(1943), where the Rlo Bec style sites (which bear slate-ware pottery) are separated geographically from thosein Peten architectural style.

84See Brainerd, 1949 and 1951, for preliminary re-ports.

85See Wauchope, 1950, for a survey of MesoamnericanFormative pottery.

86A further suggestion of early dating is given byVaillant's Valley of Mexico sequence, in which bowlsof somewhat similar shape and incised dichrome dec-oration occur in Early Ticoman, being replaced bypolychrome in Intermediate Ticoman, which is contem-poraneous with or slightly precedes Teotihuacan 1.Since Tzakol 2-3 and Teotihuacan 3 (Xololpan) are con-temporaneous, early Ticoman must be placed earlyenough to allow three well-differentiated ceramicphases before late Tzakol. A radiocarbon date fromCuicuilco, which dates somewhere in the Ticoman se-quence, reads 471 B.C. + 250, and other radiocarbondates support this approximate placement.

87Shook does not suggest this possibility.

88Used in Proskouriakoff's sense (1950, p. 1).

89See Thompson, 1945, for summary descriptionsand a bibliography on the first four types, Pollock,1940, for a geographic determination of the fifth.

90I refrain from listing the features of difference.Thompson's descriptions of northern styles must serveas a basis for the present. See Brainerd, 1954, for ageneral statement.

91A possibly fruitful analogy may be drawn with theOld World "early urban" cultures where Childe positsa difference in social status between the potters of theMesopotamian cities and those under Minoan kinglypatronage from marked differences in the characteris-tics of their wares: "In Crete the specialist potter wasnot one of the craftsmen already differentiated beforethe local urban revolution and degraded socially there-by. So, while in the Orient the aesthetic quality of pot-tery almost everywhere declined after the [urban] rev-olution, in Crete the new specialists in the palace work-shops turned vessels, delicate and beautiful and worthyto adorn the tables of princes' (Childe, 1942, pp. 155-166). The Florescent Yucatecan potters may be com-pared here with the Mesopotamian early urban potterswho made mainly undecorated, utilitarian wares ascompared to the exquisitely formed, delicately deco-rated wares of their village-living predecessors andof their contemporaries in peripheral areas.

92The lack of evidence at Chich6n Itza for an urbanconcentration of the type known at Mayapan seemslikely to be significant (see Wauchope, 1938, pp. 163-170, and Ruppert, 1952), thus suggesting that concen-trated settlements were not used until Middle Mexicantim e s.

93A news item reported after this manuscript wascompleted further confirms this early cessation ofarchitectural activity at Chiche'n Itza (Thompson, 1954).A series of excavations made in 1954 by Edwin Shookand R. E. Smith at the supposedly latest buildings atChiche6n Itza produced almost no Coarse Redware.

94For a discussion of some of the evidence for west-ern contacts with Yucatan at and before the Spanishconquest, see Scholes and Roys, 1948, pp. 317-322.

100

BIBLIOGRAPHY

La tercera temporada de exploracionesarqueologicas en Tula, Hgo., 1942.Rev. Mex. Estudios Antropol., vol.6, pp. 125-157.

1945. La cuarta y quinta temporada de exca-

vaciones en Tula, Hgo., 1943-44.Rev. Mex. Estudios Antropol., vol.7, pp. 23-64.

Andrews, E.1942.

W.Yucatan: architecture. Carnegie Inst.

Wash., Year Book 41, pp. 257-263.

1943. The archaeology of southwestern Cam-peche. Carnegie Inst. Wash., Pub.546, Contrib. 40.

Angel Fernandez, M.1941. El templo num. 5 de Tulum, Quintana

Roo. In Los Mayas Antiguos, pp.

159- 180.

Armillas, P.1948. A sequence of cultural development in

Meso-America. In W. C. Bennett,Reappraisal of Peruvian archaeolo-gy, pp. 105-111.

figy prongs of the pre-classic periodat Kaminaljuyu, Guatemala. CarnegieInst. Wash., Div. Hist. Research,Notes on Middle Amer. Archaeol. andEthnol., no. 97.

1951. A study of three-pronged incense burn-ers from Guatemala and adjacentareas. Carnegie Inst. Wash., Div.Hist. Research, Notes on MiddleAmer. Archaeol. and Ethnol., no. 101.

1951a. Further notes on three-pronged incenseburners and rim-head vessels inGuatemala. Carnegie Inst. Wash., Div.Hist. Research, Notes on Middle Amer.Archaeol. and Ethnol., no. 105.

Brainerd, G.1940.

W.Study of Yucatecan pottery. Carnegie

Inst. Wash., Year Book 39, pp. 271-274.

1941. Fine Orange pottery in Yucatan. Rev.Mex. Estudios Antropol., vol. 5, pp.163- 183.

1942. Yucatan: pottery. Carnegie Inst. Wash.,Year Book 41, pp. 253-257.

Barrera Visquez, A., and S. G. Morley1949. The Maya chronicles. Carnegie Inst.

Wash., Pub. 585, Contrib. 48.

Beals, R. L., G. W. Brainerd, and W. Smith1945. Archaeological studies in northeast Ari-

zona. Univ. Calif. Publ. Am. Arch.and Ethn., vol. 44, no. 1.

Bennett, W. C.1944. The north highlands of Peru. Amer.

Mus. Nat. Hist., Anthropol. Papers,vol. 39, pt. 1.

Bernal, I.1949. La ceramica de Monte Alban IIIA. The-

sis, Univer sidad Nacional Aut6nomade Mexico. Mimeographed.

Beyer, H.1937. Studies on the inscriptions of Chichen

Itza. Carnegie Inst. Wash., Pub.483, Contrib. 21.

Blom, F., and 0. La Farge1926- Tribes and temples. 2 vols. Tulane1927. Univ., Middle Amer. Research Ser.,

Pub. 1.

Boas, F.1911-1912.

Album de colecciones arqueol6gicos.Escuela Internac. Arqueol. y Etnol.Amer.

Borhegyi, S. F.1950. Rim-head vessels and cone-shaped ef-

1942a. Symmetry in primitive conventional de-sign. Amer. Antiquity, vol. 8, pp.

164-167.

1946. Wheel-made pottery in America? South-west Mus., Masterkey, vol. 20, no.

6, pp. 191-192.

1948. Review of S. G. Morley, The ancientMaya. Amer. Antiquity, vol. 14, pp.133-136.

1948a. Decorative techniques of Maya pottery.Southwest Mus., Masterkey, vol. 22,no. 4, pp. 131-135.

1949. Campeche. Carnegie Inst. Wash., YearBook 48, pp. 232-234.

1951. The place of chronological ordering inarchaeological analysis. Amer. An-tiquity, vol. 1o, pp. 301-313.

1951a. The use of mathematical formulationsin archaeological analysis. In J. B.Griffin, ed., Essays on archaeologi-cal methods, pp. 117-127. Univ.Michigan, Anthropol. Papers, no. 8.

195 lb. Early ceramic horizons in Yucatan. InS. Tax, ed., The civilizations of an-

cient America. Selected papers ofthe 19th Int. Cong. Amer., vol. 1,pp. 72-78.

1951 c. A Peruvian whistling jar. Southwest Mus.,Masterkey, vol. 25, no. 1, pp. 18-22.

[101]

Acosta, J. R.1944.


Brainerd, G. W. (continued)1953. On the design of the Fine Orange pottery

found at Chichen Itza, Yucatan. InHuastecos, Totonacos y sus vecinos,pp. 463-473. Rev. Mex. Estudios An-tropol., vol. 13, nos. 2 and 3.

1953A. Archeological findings [in Yucatancaves]. In R. T. Hatt, Faunal and ar-cheological researches in Yucatancaves, pp. 109-118. Cranbrook Inst.Science, Bull. 33.

1954. The Maya civilization. Publ. by theSouthwest Mus., Los Angeles.

Brenner, A.1931.

Brew, J. 0.1946.

Caso, A.1935.

The influence of technique on the deco-rative style in the domestic potteryof Culhuacan. Columbia Univ., Con-trib. to Anthropol., vol. 13.

Archaeology of Alkali Ridge, southeast.ern Utah. Papers Peabody Mus.,Harvard Univ., vol. 21.

Las exploraciones en Monte Alban, tem-porada 1934-1935. Inst. Panamer.Geog. e Hist., Pub. 18.

Cervantes, E. A.1939. Loza blanca y azulejo de Puebla. 2 vols.

Charnay, D.1887. The ancient cities of the New World.

Cresson, F. M.1937. Foot forms of pottery vessels at Pie-

dras Negras. Pub. Philadelphia An-thropol. Soc., vol. 1, pp. 37-46.

Digby, A.1951. The technical development of whistling

vases in Peru. In S. Tax, ed., Thecivilization of ancient America. Se-lected papers of the 29th Int. Cong.Amer., vol. 1, pp. 252-258.

Drucker, P.1943. Ceramic sequences at Tres Zapotes,

Veracruz, Mexico. Smithsonian Inst.,Bur. Amer. Ethnol., Bull. 140.

1952. La Venta, Tabasco. Smithsonian Inst.,Bur. Amer. Ethnol., Bull. 153.

Du Solier, W.1943.

Dutton, B. P.,1943.

A reconnaissance on Isla de Sacrificios,Veracruz, Mexico. Carnegie Inst.Wash., Div. Hist. Research, Noteson Middle Amer. Archaeol. and Eth-nol., no. 14.

and H. R. HobbsExcavations at Tajumulco, Guatemala.

School Amer. Research, Mus. NewMexico, Monogr. 9.

Ekholm, G. F.1942. Excavations at Guasave, Sinaloa, Mexi-

co. Amer. Mus. Nat. Hist., Anthropol.Papers, vol. 38, pt. 2, pp. 23-139.

1944. Excavations at Tampico and Panuco inthe Huasteca. Amer. Mus. Nat. Hist.,Anthropol. Papers, vol. 38, pt. 5, pp.320-512.

Fewkes, J. W.1907. The aborigines of Porto Rico and neigh-

boring islands. Smithsonian Inst.,Bur. Amer. Ethnol., 25th ann. rept.,pp. 3-220.

Franco C., J. L.1945. Comentarios sobre tipologla y filogenia

de la decoraci6n negra sobre colornatural del barro en la cera'mica"Azteca II." Rev. Mex. Estudios An-tropol., vol. 7, pp. 163-186.

Gann, T. W. F.1900. Mounds in northern Honduras. Smithson-

ian Inst., Bur. Amer. Ethnol., 19thann. rept., pt. 2, pp. 655-692.

Garcia Payon, J.1949. Zempoala-compendio de su estudio

arqueol6gica. Uni-Ver, vol. 1, no. 8,pp. 449-470.

Griffin, J. B., and A. Espejo1947. La alfarerla correspondiente al iultimo

periodo de ocupacidn Nahua del Vallede Mexico. I: Tlaltelolco a trave's delos tiempos, vol. 9, pp. 11-26. Mem.Acad. Hist., vol. 6, no. 2. Mexico.

1950. Alfarerla hago de Texcoco, tipo Cullina-can negro-sobre-anaranjado II. Tlal-telolco a trave's de los tiempos, vol.11, pp. 15-66. Mem. Acad. Hist., vol.9, no. 1. Mexico.

Guthe, C.1925.

Huntington, E.1914.

Pueblo pottery making. Papers PhillipsAcademy Southwe stern Expedition,no. 2.

The climatic factor as illustrated inarid America. Carnegie Inst. Wash.,Pub. 192.

Joyce, T. A.1914. Mexican archaeology.

Kidder, A. V., J. D. Jennings, and E. M. Shook1946. Excavations at Kaminaljuyu, Guatemala.

Carnegie Inst. Wash., Pub. 51.

Kidder, A. V., and A. 0. Shepard1936. The pottery of Pecos. Vol. II: The

glaze-paint, culinary and other wares.Papers Phillips Academy Southwe st-ern Expedition, no. 7.

Kidder, A. V., and E. M. Shook1946. Rim-head vessels from Kaminaljuyu,

Guatemala. Carnegie Inst. Wash.,Notes on Middle Amer. Archaeol.and Ethnol., no. 69.

102


Krieger, A. D.1944. The typologic concept. Amer. Antiquity,

vol. 9, pp. 271-288.

Linne, S.1942. Mexican highland cultures: archaeologi-

cal researches at Teotihuacan, Cal-pulalpan and Chalchicomula in 1934-35. Stockholm.

Lothrop, S. K.1924. Tulum: an archaeological study of the

east coast of Yucatan. Carnegie Inst.Wash., Pub. 335.

1933. Atitlan: an archaeological study of an-cient remains on the borders of LakeAtitlan, Guatemala. Carnegie Inst.Wash., Pub. 444.

1936. Zacualpa: a study of ancient Quiche ar-tifacts. Carnegie Inst. Wash., Pub.472.

Mera, H. P.1945. Negative painting on southwest pottery.

Southwe stern Jour. Anthr opol., vol.1, pp. 161-166.

Mercer, H. C.1895. Potter's wheel in Yucatan. Amer. Natu-

ralist, vol. 29, p. 511.

1896. The hill caves of Yucatan.

1897. The kabal, or potter's wheel of Yucatan.Univ. Pa., Free Mus. Science and Art,Bull. 2, pp. 63-70.

Merwin, R. E., and G. C. Vaillant1932. The ruins of Holmul, Guatemala. Mem.

Peabody Mus., Harvard Univ., vol. 3,no. 2.

Montgomery, R., W. Smith, and J. 0. Brew1949. Franciscan Awatovi. Papers Peabody

Mus., Harvard Univ., vol. 36.

1952. Metals from the Cenote of Sacrifice,Chichen Itza, Yucatan. Mem. PeabodyMus., Harvard Univ., vol. 10, no. 2.

Makem son, M. W.1950. The katun calendar of the Book of Tizi-

min. Amer. Antiquity, vol. 16, pp.166- 168.

Morley, S. G.1931. Report of the Chichen Itza Project. Car-

negie Inst. Wash., Year Book 30, pp.104-108.

1938. The Maya new empire. In Cooperationin Research, pp. 533-565. CarnegieInst. Wash., Pub. 501.

Maler, T.1895-1902.

Mar quina, I.1951.

Martin, P. S.1927.

Mason, J. A.1939.

Yukatekische Forschungen. Globus, vol.68, pp. 247-259; vol. 68, pp. 277-292;vol. 82, pp. 197-230.

Arquitectura prehispanica. Mem. Inst.Nac. Antrpol. e Hist., no. 1.

Report of Paul S. Martin on the Templeof the Interior Atlantean Columnsand the House of the Grinding Stones.Carnegie Inst. Wash., supplementalreport bound with reprints of reportof Division of Historical Researchfrom Year Book 26, following p. 267.This report does not appear in thecomplete Year Book.

Archaeology of Santa Marta, Colombia:the Tairona culture. Part II, section2. Field Mus., Anthropol. Ser., vol.20, no. 3.

1946. The ancient Maya.

Morris, E. H., J. Charlot, and A. A. Morris1931. The Temnple of the Warriors at Chichen

Itza, Yucatan. 2 vols. Carnegie Inst.Wash., Pub. 406.

Noguera, E.1934. Estudio de la cera'mica encontrada en

el sitio donde estaba el templo mayorde Mexico. Anal. Mus. Nac. Aqueol.,Hist., y Etnog., vol. 26, pp. 267-281.

1940. Ceramica de Quintana Roo. El M6exicoAntiguo, vol. 5, pp. 9-40.

1940a. Excavations at Tehuacan. In The Mayaand their neighbors, pp. 306-319.

1947. Ceramica de Xochicalco. El Mexico An-tiguo, vol. 5, pp. 273-300.

Nuttall, Z.1910. The Island of Sacrificos. Amer. Anthro-

pol., n.s., vol. 12, pp. 257-295.

Mayer, B.1852. Mexico: Aztec, Spanish and Republi-

can.... 2 vols. Hartford.

Meggers, B. J.1948. The archaeology of the Amazon basin.

Smithsonian Inst., Bur. Amer. Eth-nol., Bull. 143. Handbook of SouthAmerican Indians, vol. 3, pp. 149-166.

1951. A pre-Columbian colonization of theAmazon. Archaeology, vol. 4, pp.

110-114.

Palacios, E. J.1937. Arqueolo'gia de Me'xico: culturas arca'ica

y tolteca. Encic. Ilus. Mex., no. 4.

1942. Cultura totonaca. El Nacional. Me'xico.

1943. Los "yugos" y su simbolismo. Proc. 27thInt. Cong. Amer., 1939, pp. 509-545.

1948. Review of R. Girard, El calendarioMaya-Me'xica. Anal. Soc. Geog. eHist. de Guatemala, vol. 23, pp. 17-28.

103


Pearse, A. S., E. P. Creaser, F. G. Hall, et al.1936. The cenotes of Yucatan: a zoological

and hydrographic survey. CarnegieInst. Wash., Pub. 457.

Peflafiel, A.1890.

Piffa Chan, R.1948.

Monumentos del arte Mexicano antiguo.Berlin.

Breve estu'dio sobre la funeraria deJaina, Campeche. Mus. Arqueol.,Etnog. y Hist. de Campeche, Cua-derno 7.

1934. Ceramic problems. Carnegie Inst. Wash.,Year Book 33, pp. 92-93.

1935. Ceramics. Carnegie Inst. Wash., YearBook 34, pp. 126-127.

Robinson, W.1951.

Rouse, I.1939.

S.A method for chronologically ordering

archaeological deposits. Amer. An-tiquity, vol. 16, pp. 293-301.

Prehistory in Haiti: a study in method.Yale Univ. Pub. in Anthropol., no. 21.

Pollock, H. E. D.1931. Architectural problems in the Maya

field. Carnegie Inst. Wash., YearBook 30, pp. 117-119.

1935. The architectural survey. Carnegie Inst.Wash., Year Book 34, pp. 124-126.

1936. The architectural survey. Carnegie Inst.Wash., Year Book 35, pp. 122-125.

1936a. The Casa Redonda at Chichen Itza, Yu-catan. Carnegie Inst. Wash., Pub.456, Contrib. 17.

1940. The architectural survey of Yucatan.Carnegie Inst. Wash., Year Book39, pp. 265-267.

Popenoe, D. H.1934. Some excavations at Playa de los Muer-

tos, Ulua River, Honduras. MayaResearch, vol. 1, pp. 61-85.

Porter, M. N.1953. Tlatilco and the pre-Classic cultures of

the new world. Wenner-Gren Found.for Anthropol. Research. Viking FundPub. in Anthropol., no. 19.

Proskouriakoff, T.1950. A study of Classic Maya sculpture. Car-

negie Inst. Wash., Pub. 593.

Rendon, S.1947. Notas sobre la alfarerla indlgena de la

peninsula de Yucatan. Rev. Mex. Es-tudios Antropol., vol. 9, pp. 107-123.

1948. La cera'mica juguetera de Maxcana, Yu-catan. El Nacional. M6xico.

Ricketson, 0. G., and E. B. Ricketson1937. Uaxactun, Guatemala: Group E-1926-

1931. Carnegie Inst. Wash., Pub. 477.

Roberts, H. B.1931. Problems in the study of Maya ceram-

ics. Carnegie Inst. Wash., YearBook 30, pp. 114-116.

Roys, L.1941.

Roys, R. L.1933.

Masonry traits found at Mayapan. In LosMayas Antiguos, pp. 145-156.

The Book of Chilam Balam of Chumayel.Carnegie Inst. Wash., Pub. 438.

1943. The Indian background of Colonial Yuca-tan. Carnegie Inst. Wash., Pub. 548.

Roys, R. L.,1940.

Ruppert, K.1931.

F. V. Scholes, and E. B. AdamsReport and census of the Indians of Co-

zumel, 1570. Carnegie Inst. Wash.,Pub. 523, Contrib. 30.

Temple of the Wall Panels, Chichen Itza.Carnegie Inst. Wash., Pub. 403, Con-trib. 3.

1943. The Mercado, Chichen Itza, Yucatan.Carnegie Inst. Wash., Pub. 546, Con-trib. 43.

Ruppert, K.,1943.

and J. H. Denison, Jr.Archaeological reconnaissance in Cam-

peche, Quintana Roo, and Peten. Car-negie Inst. Wash., Pub. 543.

1952. Chichen Itza: architectural notes andplans. Carnegie Inst. Wash., Pub. 595.

Ruz L., A.1945. Campeche en la arqueol6gia Maya. Acta

Antropol., vol. 1, nos. 2, 3.

n. d. La costa de Campeche en los tiemposprehispanicos: prospeccion ceramicay bosquejo hist6rico. Thesis, Univer-sidad Nacional Aut6noma de Me'xico.

Scholes, F. V., and R. L. Roys1948. The Maya Chontal Indians of Acalan-Tix-

chel: a contribution to the history andethnography of the Yucatan peninsula.Carnegie Inst. Wash., Pub. 560.

Seler, E.1901.

1932. The architectural-ceramic survey. Car-negie Inst. Wash., Year Book 31, pp.96-97.

1933. Ceramic research. Carnegie Inst. Wash.,Year Book 32, pp. 86-88.

Die alten Ansiedlungen von Chacula imDistrikte Nenton des Departem ent sHuehuetenango der Republik Guate-mala.

1915. Die Teotihuacan-Kultur des Hochlandsvon Mexico. Gesammelte Abhandlung-en zur amerikanischen Sprach- undAltertumskunde, vol. 5, pp. 557-585.

104


Shepard, A. 0.1948. Plumbate, a Mesoamerican trade ware.


1951. Ceramic technology. Carnegie Inst.Wash., Year Book 50, pp. 241-244.

Shook, E. M.1940. Exploration in the ruins of Oxkintok,

Yucatan. Rev. Mex. Estudios Antro-pol., vol. 4, pp. 165-171.

Taylor, W.1948. A study of archaeology. Mem. Amer.

Anthropol. Assoc., no. 09.

Thomp son, E. H.n.d. The Maya potter of Yucatan. Typewritten.

Peabody Mus., Harvard Univ.

1897. The chultunes of Labna, Yucatan. Mem.Peabody Mus., Harvard Univ., vol. 1,no. 3.

1951. The present status of research on thepre-Classic horizons in Guatemala.In S. Tax, ed., The civilization ofancient America. Selected papers ofthe 29th Int. Cong. Amer., vol. 1, pp.93-100.

Shook, E. M., and A. V. Kidder1952. Mound E-III-3, Kaminaljuyu, Guatemala.

Carnegie Inst. Wash., Pub. 596, Con-trib. 53.

Smith, A. L.1950.

Smith, A. L.,1943.

Smith, R. E.1936.

Uaxactun, Guatemala: excavations of1931-1937. Carnegie Inst. Wash.,Pub. 588.

and A. V. KidderExplorations in the Motagua Valley,

Guatemala. Carnegie Inst. Wash.,Pub. 546, Contrib. 41.

Preliminary shape analysis of the Uax-actun pottery. Carnegie Inst. Wash.Mimeographed for distribution tostudent s.

1936a. Ceramics of Uaxactun: a preliminaryanalysis of decorative technics anddesign. Carnegie Inst. Wash. Mimeo-graphed for distribution to students.

1940. Ceramics of the Peten. In The Mayaand their neighbors, pp. 242-249.

Spinden, E. S.1933.

Spinden, H. J.1913.

The place of Tajin in Totonac archaeol-ogy. Amer. Anthropol., vol. 35, pp.225 -270.

A study of Maya art. Mem. PeabodyMus., Harvard Univ., vol. 6.

Thomp son, J. E. S.1931. Archaeological investigations in the South-

ern Cayo District, British Honduras.Field Mus., Anthropol. Ser., vol. 17,pp. 215-362.

1937. A new method of deciphering Yucatecandates with special reference to Chi-chen Itza. Carnegie Inst. Wash., Pub.483, Contrib. 22.

1939. Excavations at San Jose, British Hondur-as. Carnegie Inst. Wash., Pub. 505.

1940. Late ceramic horizons at Benque Viejo,British Honduras. Carnegie Inst.Wash., Pub. 528, Contrib. 35.

1941. A co-ordination of the history of ChichenItza with ceramic sequences in centralMexico. Rev. Mex. Estudios Antropol.,vol. 5, pp. 97-111.

1943. Pitfalls and stimuli in the interpretationof history through loan words. TulaneUniv., Middle Amer. Research Inst.,Philological and Documentary Studies,vol. 1, no. 2.

1945. A survey of the northern Maya area.Amer. Antiquity, vol. 11, pp. 2-24.

1954. Maya area. Amer. Antiquity, vol. 20, p.208.

Thomp son, J.1932.

Tozzer, A. M

1923. Fresco technique on pottery. In Ancientcivilizations of Mexico and CentralAmerica. Amer. Mus. Nat. Hist.,Handbook Ser., no. 3.

Stephens, J.1843.

Stewar d, J.,1946.

Strebel, H.1899.

L.Incidents of travel in Yucatan. 2 vols.

ed.Handbook of South American Indians,

vol. 2. Smithsonian Inst., Bur. Amer.Ethnol., Bull. 143.

Uber Tierornamente auf Trongefassenaus Alt-Mexico. Berlin.

E. S., H. E. D. Pollock, and J. CharlotA preliminary study of the ruins of Coba,

Quintana Roo, Mexico. Carnegie Inst.Wash., Pub. 424.

1907. A comparative study of the Mayas andthe Lacandones.

1930. Maya and Toltec figures at Chichen Itza.Proc. 23d Int. Cong. Amer., 1930, pp.155-164.

1941. Landa's Relacio'n de las cosas de Yuca-tan. A translation edited with notes.Papers Peabody Mus., Harvard Univ.,vol. 18.

Vaillant, G.1926.

C.Excavations at Station No. 13. In K. Rup..

pert, Chichen Itza: architectural notesand plans, pp. 157-162 (1952). Car-negie Inst. Wash., Pub. 595.

1927. The chronological significance of Mayaceramics. Doctoral dissertation, Har-vard Univ.

105


Vaillant, G.1930.

C. (continued)Excavations at Zacatenco. Amer. Mus.

Nat. Hist., Anthropol. Papers, vol.32, pt. 1, pp. 1-197.

1933. Hidden history. Nat. Hist., vol. 33, pp.618-628.

1935. Chronology and stratigraphy in the Ma-ya area. Maya Research, vol. 2, pp.

119- 143.

Vega Bolaf¶os,1945.

L. et al.Catalogo de construcciones religiosas

del estado de Yucatan. 2 vols. Secre-taria de Hacienda y Cre'dito Pu'blico,Direcci6n General de Bienes Nacion-ale s.

Villa Rojas, A.1945. The Maya of east central Quintana Roo.


Wauchope, R.1941. Effigy-head supports from Zacualpa,

Guatemala. In Los Mayas Antiguos,pp. 211-232.

1947. An approach to the Maya correlationproblem through Guatemala highlandarchaeology and native annals. Amer.Antiquity, vol. 13, pp. 59-66.

1948. Excavations at Zacualpa, Guatemala.Tulane Univ., Middle Amer. ResearchInst., Pub. 14.

1950. A tentative sequence of pre-Classic ce-ramics in Middle America. TulaneUniv., Middle Amer. Research Rec-ords, vol. 1, no. 14, pp. 211-250.

Weitlaner, R. J.1948. Exploracio'n arqueolo'gica en Guerrero.

In El Occidente de Mexico, pp. 77-85.Soc. Mex. Antropol.

Willey, G. R.1945. Horizon styles and pottery traditions in

Peruvian archaeology. Amer. Antiqui-ty, vol. 11, pp. 49-56.

Woodbury, R. B.1948. Progress at Zaculeu, Guatemala. Amer.

Antiquity, vol. 14, pp. 121-122.

106

Date post:	09-Feb-2017
Category:	Documents
Upload:	hacong
View:	230 times
Download:	2 times

anthropological records the archaeological ceramics of yucatan

Documents