Archeological Studies in Peru 1941—1942 9780231878517

Citation preview

ARCHEOLOGI GAL STUDIES IN PERU, 1941-1942

COLUMBIA STUDIES IN ARCHEOLOGY AND ETHNOLOGY • Vol. I

ARGHEOLOGIGAL STUDIES IN PERU 1 9 4 1 - 1 9 4 2

By WILLIAM DUNCAN STRONG G O R D O N R. WILLEY, and J O H N M. C O R B E T T

NEW YORK COLUMBIA U N I V E R S I T Y PRESS 1943

iiilllllllllllllllllllllMlllllllllllllMlllllllllllMllllllllimill

lllllllllillllllMlllllilllllillllll

The publication of this volume has been made possible through the generous aid of the Institute of Andean Research. A list of publications for the Institute, of which this volume is No. 3, appears on the last page, following the Index.

C O P Y R I G H T ,

1943,

BY

C O L U M B I A

U N I V E R S I T Y

PRESS,

NEW

YORK

Foreign Agents: O X F O R D U N I V E R S I T Y P R E S S , Humphrey Milford, Amen House, London, E.C. 4, England, AND B. I. Building, Nicol Road, Bombay, India M A N U F A C T U R E D

IN

THE

U N I T E D

STATES

OF

A M E R I C A

TO D R . J U L I O C. T E L L O of the Universidad Mayor de San Marcos in Lima and the Museo de Antropologia in Magdalena, this monograph is dedicated, in recognition of the fact that the founding of the Institute of Andean Research was the result of stimulation and suggestion by Dr. Tello during his visit to the United States in 1936; and in appreciation also of the importance of his decades of archeological exploration in his native country. These explorations have served as a basis for the archeological investigations which the Institute of Andean Research, as well as other North American institutions of learning, have been able to provide for a series of younger scholars in recent years, to the mutual enhancement of the intellectual interests and cultural linkage of Peru and the United States.

Preface

I

1941, the Institute of Andean Research, working under the auspices of the Office of the Coordinator of Inter-American Affairs, began a series of stratigraphic excavations in certain key areas in Latin America (see Strong, 1942, in the bibliography at the end of this volume). One of these investigations (Project 3) was particularly concerned with refuse-heap stratigraphy on the central Peruvian coast. T h e present volume presents a part of the results thus attained. Further reports on this work will be issued as opportunity permits. Our most profound thanks go to the Honorable Dr. Manuel Prado Ugarteche, President of the Republic of Peru; to the Honorable Dr. Pedro Olivera, Minister of Education; and to the members of the Patronato de Arqueología del Perú, not only for permission to carry on our survey and excavations but also for their scientific and personal assistance in ways too numerous to mention. T o Dr. Julio C. Tello, Director of the Museum of Anthropology in Magdalena Vieja, we are grateful for wise scientific guidance and for much professional and personal assistance. T o Dr. Luis E. Valcárcel, Director of the National Museum in Lima, who assisted our project in many ways, also go our warmest thanks. Dr. Manuel Beltroy and Señor José Respaldiza of the Ministry of Education likewise put us greatly in their debt in regard to both official and personal assistance. During our work at Pachacamac we were greatly aided by Dr. Tello's assistant, Señor Wapaya; and in the Museum of Anthropology in Magdalena Vieja we were helped by Señorita Rebeca Carrión Cachot and Señores Toribio Mejia Xesspe and Espejo. A t the National Museum in Lima, Señores Juan J. Delgado, Jorge Muelle, and others granted us much scientific as well as personal assistance. T h e personal associations resulting from these contacts in field and laboratory will be long remembered and will be renewed as often as opportunity offers. N JUNE,

W. R. Grace and Company extended us many courtesies both before and during our stay in Peru. We are particularly grateful for the hospitality of Señor Luis R. Fontenoy, General Manager of Hacienda Paramonga. In Lima, Señor José Quimper, Grace and Company Representative, generously gave us much time in arranging innumerable technical and other details. We are most grateful to W. R. Grace and Company and to him. T o the officers and men of the Guardia Civil of Peru in Chancay and Puerto de Supe we also owe our thanks for their cooperation in expediting our work in those regions. T h e present volume includes four parts each dealing with a different aspect of our year's work. Number 1 was written in New York in late May, 1942. A t this time a considerable part of our notes, photographs, and specimens was still in Peru and not available for study. However, since the present reports cover only a portion of our research and since it may be some time before the remainder can be presented, it seemed advisable at least to outline it at this time. This first section, therefore, consists merely of notes and makes no pretense at being either a complete survey or a detailed report. Number 2 is based on excavations at Pachacamac carried on from July 23 to October 10, 1941. T h e senior author, William Duncan Strong, was in charge of this work until September 10, and it was completed under the direction of Gordon R. Willey. T h e material was worked up after its arrival in New York by the senior author, later assisted by John M. Corbett, the junior author. For invaluable assistance in unpacking, cleaning, and supervising the collections, the authors are very grateful to Marjorie Trumbull of New York. Miss Helen L. Phipard assisted in this and in statistical and tabular work. In Number 2 the text figures of pottery types are the work of Mr. Jack Anglim and the maps and diagrams were drawn by the junior author and Mr. Bennett K. Foster. Because of University duties and work con-

viii

PREFACE

nected with the war situation, it has been impossible in the time available to make as complete analytic and comparative studies of the varied Pachacamac material as we would wish. However, the ceramic sequence in relation to the physical strata at the site of the excavation is basic to all other aspects of our problem, and it is this material that is outlined here. Much remains to be completed when the world situation again permits concentrated scientific research. Number 3 was written in Lima in the spring of 1942. T h e writer consulted or corresponded with Drs. A. L. Kroeber, W. D. Strong, W. C. Bennett, T . D. McCown, S. K. Lothrop, M. T . Newman, Bernard Mishkin, and J. S. Rowe in addition to the Peruvian scholars previously mentioned. For their advice and criticism he is most grateful. (As Numbers 2 and 3 were written in different places during the same period, it was not possible to synchronize them exactly as to format. Each is presented basically as it was written, but certain minor differences in terminology were ironed out in May, 1942, when the authors were together in New York.) In Lima, Señores Abraham Guillen and Sabino Springett of the National Museum of Peru prepared, respectively, the photographs and drawings presented in Number 3. Their timely services were indispensable. T o Katharine W. Willey, the author's wife, he is most grateful for reading and correcting the manuscript during its necessarily hasty preparation. Number 4 covers only one part of the excavations carried out at Ancon. T h e purposes of this work as a whole are mentioned in the present paper. T h e writer wishes to express his appreciation to Dr. M. T . Newman, who collaborated

in this field work, as well as at Pachacamac and Chancay. T h e statistical correlations employed were made by Dr. Roque Garcia Frias of the Peruvian Bureau of Census. T h e writer is also indebted to Dr. Alberto Area Parr6 and Senor Washington Patino Area of the same office. Representative collections from the sites worked under Project 3 were left at the disposal of the Patronato de Arquelogia del Peril under the immediate care of Dr. J. C. Tello, member of the Patronato and Inspector General of Antiquities. T h e remainder of the study collections are either awaiting transshipment from Peru or are stored for the Institute of Andean Research in the American Museum of Natural History in New York. T o the Department of Anthropology and to the Trustees, the Bursar, and other officials of this institution, we owe a great debt of gratitude for much assistance furnished the Institute of Andean Research before, in the course of, and subsequent to our field work. T h e notes, maps, negatives, and other scientific data from the Peruvian work are now stored in the Department of Anthropology at Columbia University. T o this Department, and to Columbia University, we are grateful for both direct and indirect assistance in the field researches and in preparing the manuscripts. In the latter regard, the authors wish particularly to thank Mr. Paul L. Cooper for his assistance in editing these manuscripts. W M . DUNCAN STRONG GORDON R. WILLEY JOHN M. CORBETT Columbia University August, 1942

Contents Î.

ARCHEOLOGICAL NOTES ON THE CENTRAL COAST BY

2.

W I L L I A M D U N C A N STRONG AND GORDON R .

3.

GORDON R .

GORDON R .

123

WILLEY

A SUPPLEMENT T O T H E P O T T E R Y SEQUENCE A T ANCON BY

27

CORBETT

E X C A V A T I O N S IN T H E C H A N C A Y VALLEY BY

4.

WILLEY

A CERAMIC SEQUENCE A T PACHACAMAC B Y W I L L I A M D U N C A N STRONG AND JOHN M .

1

197

WILLEY

BIBLIOGRAPHY

213

INDEX

217

ARCHEOLOGICAL NOTES ON THE CENTRAL COAST By WILLIAM DUNCAN STRONG and GORDON R. WILLEY

COLUMBIA STUDIES IN ARCHEOLOGY AND ETHNOLOGY, Vol. I, No. 1

Contents Introductory Remarks

5

T h e Northern Survey

6

Excavations in the Vicinity of Puerto de Supe

xi

Excavations in the Ancon Shell Mounds

15

T h e Southern Survey

18

PLATES 1 a.

View of Pachacamac from the Temple of the Sun, Looking toward the "House of the Cacique"

b.

Great Shell Mounds at the Mouth of the lea River

2 a.

White-on-Red Period Grave at Cerro de Trinidad, Site E, Chancay Valley

b.

Rectangular Rock Enclosure, East Side of Cerro de Trinidad, Chancay Valley

3 a. b.

4 a. b. 5 a.

Village and Cemetery Excavations at Aspero, Puerto de Supe Rock-walled Structure Enclosing "Altar," Early Ancon-Supe Period, at Aspero, Puerto de Supe "Altar" at Aspero, Looking through Doorway Connecting T w o Rooms Close-up Showing Type of Construction Used in "Altar" Cerro Hueco Ruin in the Cañete Valley

b.

T h e "House of the Women" at Inca Huasi Ruin, Cañete Valley

c.

Compound Structure at Inca Huasi Ruin, Cañete Valley

MAP of Central Coast of Peru, Showing Sites Visited and Discussed

7

Introductory Remarks

P

presenting the more specific and detailed results attained by Project 3 as the result of archeological excavations at Pachacamac, in the Chancay valley, and at Ancon during the year from June, 1941, to June, 1942, it may be worth while at least to outline the nature of the remaining survey and excavation work not reported in detail at this time. This last falls into two parts: first, partial archeological surveys to the north and to the south of Lima; and, second, limited excavations in the presumably ancient shell heaps at Supe and Ancon. T h e following account further attempts to place a number of the sites visited in the generalized central-coast culture sequence suggested in Part 3 (see chart, p. 196), which has been set up on the basis of preRIOR T O

viously published materials, as well as on our recent ceramic collections and observations at these sites. Both of these surveys were incomplete inasmuch as only a relatively small proportion of the great number of sites on the central coast were examined in any detail. Further, at the time of this writing many of the collections, notes, photographs, and other records pertaining to the excavations outlined in Part 1 were either in transshipment from Peru, in the customs, or otherwise unavailable to the writers. Nevertheless, it seemed possible on the basis of our journals and available photographs to record certain data which should have value in amplifying or extending the more detailed excavation records presented later in Parts 2 to 4.

The Northern Survey

S

brief survey trips were made along the central coast between Lima and Pativilca (see map, p. 7) both before and after our work at Pachacamac. T h e personnel on these trips included the authors (Strong and Willey) in company with Mr. Junius Bird and Mr. John Corbett. During these trips a number of sites were visited and surface collections of potsherds and other artifacts were made. Plans for future excavations at Ancon, Chancay, and Puerto de Supe were also laid, and arrangements were made with local authorities and laborers. We will consider these sites from south to north. T o the north of Lima, in the Chillon valley, is the extensive cemetery of Zapallar (map, p. 7, No. 14). This site lies just beyond the cultivated land north of the Pan-American Highway. T h e modern, walled cemetery of the little village of Puente Piedra lies over a part of the aboriginal burying ground. Considerable huaquero or treasure-hunter excavation of graves is going on at Zapallar at the present time, and surface material discarded from these graves is easily obtained. A very superficial examination of surface materials there, in July of 1941, suggested late burials marked by Chancay Black-on-white and sub-Chancay styles. It is quite probable that earlier remains are also to be found. Dr. S. K. Lothrop, who has made large surface collections at Zapallar, is at present engaged in a study which will more exactly determine the nature of this important site in the Chillon valley. EVERAL

A t Ancon, a few kilometers north and west along the highway (map, p. 7, No. 13), sherd collections were made in all parts of the Necropolis area and on the shell mounds overlooking the bay. In the Necropolis the materials picked up by us on the surface were those correlating with Strong's Middle Ancon II (Epigonal), Late Ancon I (Black, white and red geometric), and Late Ancon II (Blackon-white Chancay) (Strong, 1925). Early An-

con Incised pottery was found in a modern road cut through the shell mounds on the hill overlooking the Necropolis. North beyond Ancon, on the brow of the high sand hills facing the sea and the Pasamayo arm of the Chancay valley, is another large Late Period cemetery called Cerro Pasamayo (map, p. 7, No. 12). Directed here by Dr. J. C. Tello, we made a collection of about a hundred sherds all of which were Late Chancay Black-on-white. Dr. Lothrop (personal communication) also reports Black, white and red geometric ware from Cerro Pasamayo, although none was found by us. T h e present surface of the site is sandy and windswept, and most of the surface material, including sherds, textiles, and human bones, has been bleached and eroded from long exposure. In July of 1941 very little recent huaquero excavation was evident, although at some time in the past a great many graves had been opened. In addition to graves, evidence of adobe structures, perhaps dwellings, was noted in the road cut on the inland side of the highway. Continuing north across the Chancay valley, the highway passes through a small aggregation of sandy and rocky hills just south of the Port of Chancay. T h e roadbed has been cut down for a considerable depth into the seaward flank of one of the largest of these hills, the Cerro de Trinidad (map, p. 7, No. 9). This site is discussed in detail in Part 3 of this volume. It is sufficient to note that on our preliminary survey Interlocking, Whiteon-red, and Late Chancay ceramic materials were all found in the road cut and on the surface of the site as described by Uhle (Kroeber, 1926b). There are no other important surface indications except occasional alignments suggesting remnants of structures. There has been very little grave excavation at the site. T h e occupation area cut by the highway and covering the lower slopes of this hill

MAP OF CENTRAL COAST OF PERU, SHOWING SITES VISITED AND DISCUSSED

8

THE NORTHERN SURVEY

on the southern side was named by Uhle "Cerro de Trinidad, Site E." (Kroeber, 1926b). At Site A, Cerro de Trinidad, Uhle excavated a number of graves all of which proved to be of the Late Chancay Black-on-white period. The cemetery is a large one and typical of Late Period graveyards of the coast. Present-day looting goes on at Site A, and a large surface collection was easily obtainable. Out of approximately two hundred decorated sherds gathered at random, only one Epigonal and one Chancay White-on-red fragment were inconsistent with the Late Chancay assignment of the site originally set forth by Uhle. In addition to the cemetery, some adobe structures are exposed by the highway cut, and these are probably of the same period as the grave fields. These structures are to be seen just a few meters north of the main area of the Site A cemetery toward the Port of Chancay. It seems likely that the dwellings as well as the cemetery should be included in Site A. On the east side of Cerro de Trinidad is a large rectangular enclosure, several hundred meters on a side, made up of medium- and small-sized stones (map, p. 7, No. 10). The walls on all four sides are now jumbled stones, but one small section on the upper and more protected end still shows dry cyclopean construction suggesting that the entire quadrangle was once so composed. Few sherds or other cultural remains are to be found in the immediate vicinity of the enclosure. Below, at the eastern foot of the hill and stretching for over a kilometer to the east and south, is one of the largest dwelling sites of the central coast. This extensive series of walled compounds, buildings, and pyramids, locally called Ollada, is little known. Surface indications of structures, although clear, have been eroded and sand covered. This is perhaps one of the reasons why the site has not received more attention. Architecture here involves the use of both huge, rough, uncut stone for facings on the largest terraced pyramid and brickshaped, mold-made adobes for general pyramid and wall construction. T h e largest pyramid, which lies closest to the Cerro de Trinidad, has a series of descending terraces on one side. T w o three-by-

three-meter test pits were excavated into these terraces. T h e pit on the lower terrace penetrated through sand and rubbish to strike pure sand at about 1.5 meters below surface. Sherds, which came out by the hundreds, appear to be all of the Late Chancay style. This classification, however, is most cursory and preliminary, pending further study. T h e pit located on one of the higher terraces showed similar rubbish rich in sherds. As would be expected, its depth was greater, and sterile sand was almost three meters deep. All pottery fragments in this pit were also Late Chancay in style. With the fill of the mound identified as Late Chancay on the basis of the test pits in two separate sections of the structure, and since no earlier materials were encountered on the old ground surface beneath the mound, it seems safe to assume that the construction here took place in Late Chancay or post-Late Chancay times. Another test pit some distance away in a courtyard adjoining a second mound corroborated the results of the first two pits. Rubbish was shallower in this third pit, and the sherds were uniformly plain or Black-on-white Chancay. Burials do not seem to be common at Ollada. From all appearances it is primarily a habitation and ceremonial site of the Late period. It is quite possible that the graveyard of Site A, Cerro de Trinidad, which is not far distant, was the burial place of the inhabitants. Before leaving the southern side of the Chancay valley, we should mention the sites at and in the vicinity of Banos de Boza (map, p. 7, No. 11). Turning inland at the Guardia Civil control station shortly after entering the valley from the south, a side road winds back along the south wall of the valley some six kilometers to the Boza Springs. Along this road a number of habitation and burial sites can be observed. Skeletal material has been thrown out of looted graves, but surface pottery is very scarce. Partial alignments of stones imply dwellings or other structures. At the Boza Springs there is a small site located on sand hills rising out of the marsh area back of the modern hotel. This site was excavated in the course of our survey and is described in Part 3 of this volume. All evidence from our

THE NORTHERN SURVEY excavations shows it to have been a habitation as well as a burial site of the White-onred or earliest known period of the Chancay valley. On the north side of the Chancay valley are two big sites (map, p. 7, Nos. 7, 8) known from Uhle's earlier investigations, Lauri (Lauren) and Jecuan (Kroeber, 1926b). T h e former site, which is in a quebrada on the seaward slope of a range of hills separating two sections of the valley, is probably the richest of all Late Chancay sites of the coastal area. Remnants of large adobe structures are situated at the bottom of the quebrada facing the cultivated land and farther back on higher ground are numerous graves. Clandestine excavation is proceeding at a rapid rate at the present time so that a large collection of sherds, some of which constitute large portions of vessels, can be made from the surface in a few minutes. T h e site was visited three separate times by our party between July and November, 1941. Virtually all of the sherd material collected on these three trips was Late Chancay. Jecuan is reached by following the road around the landward end of the range of hills and then turning back toward the sea. T h e site is in a large quebrada or pocket in these hills facing inland toward an enclosed arm of the Chancay valley. Almost directly over the range is the Lauri site just discussed. At Jecuan Uhle found Epigonal, Black, white and red geometric, and Late Chancay vessels in the various graves, the sequence presumably being in that order from earliest to latest (Kroeber, 1926b). Fragments of vessels of all of these styles were found by us in 1941 on the surface at this site. In addition, at another, smaller quebrada near by, sherds of these three styles and a few Interlocking style sherds were collected. T h e Jecuan sites seem to be primarily burying grounds, although in the larger quebrada there are two small rectangular platforms which may have been house or ceremonial structure foundations. North of the Chancay valley at the ninetykilometer mark on the Pan-American highway is the isolated railroad station of Dona Maria (map, p. 7, No. 5). T w o important sites may be reached from this point. Turning

9

off the pavement and driving inland along a little-used road, one arrives at Teatino, some ten kilometers from Doña María (map, p. 7, No. 6). T h e Teatino cemetery is one of the most unusual in appearance along the coast, and the site, in the opinion of Dr. Tello, is of great importance. Its location is in a small ravine or quebrada of the foothills of the higher ranges to the east. T h e elevation at Teatino is enough greater than that of the immediate coast to allow for the growth of low, green vegetation or pasto during the Peruvian winter months. T h e floor of the quebrada for several hundred meters in all directions is dotted with monolithic, upright grave markers. Excavation by unknown persons is in an advanced state and continues at present. Well over a hundred graves have been opened. These are extremely deep, rectangular, and adobe lined. Near the top they are often partly stone lined, in which cases the above surface grave-marking stone has been mortised into the wall masonry. T h e bottom of many of these shafts as well as the surface of the ground was covered with large sherds discarded by the grave excavators. Dr. Tello (personal communication) mentions Late Chancay or sub-Chancay, Three-color geometric, Epigonal, and an incised pottery from Teatino. All of these styles were found at the site by us. T h e incised ware is, in the opinion of the authors, closely related to the Middle Ancón I Incised pottery from the Ancon Necropolis. Vessel form and face-spout modeling also relate a number of incised fragments from Teatino to the Tiahuanacoid influences of the coast. Besides the above-mentioned styles, the junior author found one typical Early Lima fragment at Teatino. On the opposite side of the highway, six kilometers north along the narrow-gauge railway from the Doña María station, is the Doña María site (map, p. 7, No. 5). This is a large cemetery and refuse area along the high steep hills or dunes facing the sea. A surface collection gathered here includes all of the principal pottery styles recorded at Teatino, with the Middle Ancón I type of incising also present. Refuse beds of midden and shell, apparently quite deep in some

io

THE NORTHERN SURVEY

places, make Dona Maria a promising site for stratigraphic excavation in the future. In the Huaura valley there are a great many sites about which little is known or published (map, p. 7, No. 4). T w o of these were briefly visited by the writers in July of 1941 and again in October of the same year. Entering the Huaura valley from the south, at the 130kilometer mark on the highway, one sees a great burial site on the slope of the hills facing the valley close to the right side of the road. Most of the material from this cemetery appears to be a slightly variant form of Late Chancay Black-on-white and White-slipped. Farther north, beyond the city of Huacho, there is a large dwelling site and cemetery on a promontory just above the Punta Pegota beach. A cursory inspection of pottery at this second location shows it to be much like the material at the 1 go-kilometer site. On the southern edge of the Supe valley, at

the 164-kilometer marker on the highway, there is an extensive and much dug-over cemetery extending on both sides of the highway (map, p. 7, No. 3). Late Chancay sherds are absent, and the dominant decorated style is a pressed or relief-decorated red ware. Some of this is undoubtedly "Middle Period," and this identification is further substantiated by the occurrence of Epigonal and Tiahuanaco sherds at the site. It seems reasonable to expect, however, that the pressed-ware tradition continued in this region and is also contemporaneous with the Late Chimu-Incaic occupation of the Supe valley (Kroeber, 1925b). On the other side of the valley, near Puerto de Supe, are the two "shell-mound" sites excavated first by Uhle and recently by us during the present season's work (map, p. 7, Nos. 1, 2). These sites and our excavations here are briefly described in the next section of this paper.

Excavations in the Vicinity of Puerto de Supe A LMOST forty years ago Uhle excavated in / \ two shell-mound sites in the vicinity of JL JL.Puerto de Supe. These were quite unlike the other archeological sites of the immediate region and showed affinities to the shell mounds of Ancon and Bella Vista near Callao. T h e first of these sites at Puerto de Supe lies in a small, deep quebrada just over the hill from the port in a southerly direction. It is marked by an old lighthouse which is situated on the seaward promontory of the hill or ridge which separates the site from the port. Uhle referred to the location as the "site below the lighthouse," and the term "Lighthouse site" seems convenient for the present discussion (map, p. 7, No. 1). Farther down the coast, two to three kilometers south and east from the port, is the second site, unnamed by Uhle but locally called Aspero (map, p. 7, No. 2). T h i s midden is also situated in a quebrada facing the open fields of the San Nicolas Hacienda. T h e ocean is about a kilometer distant from the site.

Aspero presents a black, ashy ground surface dotted with many small knolls of shell and other rubbish and outcroppings of rockwalled structures which are completely tumbled and covered over. From the mouth of the quebrada upward, the extent of the refuse area is about three hundred meters, and at the widest point across the quebrada the distance is about two hundred meters. In a small sandy arm of the main area of the site is a cemetery (see Plate 3a). Previous digging in this section disclosed its location, although from the scant scattered surface debris it was obvious that the graves were not rich in ceramics. In the excavations at Aspero it was decided to direct attention to three types of remains: refuse, structures, and burials. Dividing the crews, the junior author began two

test pits at random spots in the approximate center of the blackened occupation area. Mr. Corbett, with another group of workmen, started clearing around a rock-walled structure which was located a little to the north of the center of the rubbish area; and Dr. Newman, with the remainder of the crew, put down a number of exploratory trenches in the sandy burial area. (See Plate 3a.) T h e two rubbish tests can be seen in the right foreground, the clearing of the structure is in progress in the center of the picture, and the work in the cemetery is going on at the back. In the rubbish, both test pits were carried down to a depth of approximately 2.50 meters below surface. T h e soil removed was a mixture of black earth, burned rock, ash, shells, decayed wood, and organic matter. With the exception of a small whistle made of a bird bone which was found with an infant burial near the bottom of one of the pits, not a single artifact was found in either of the pits. T h e refuse clearly indicated an occupation of some duration at and adjacent to the spot, but, other than the texture and appearance of the soil, there were no evidences of human culture to distinguish the site. Furthermore, not a single surface sherd was found anywhere within the quebrada enclosing the site. Uhle is not absolutely clear on the point of whether or not he found sherds in the rubbish at his second Puerto de Supe site (Aspero). If he did, we were not so fortunate in our attempts. In clearing the rock structure, Corbett found a lance or dart point of quartzite, a cache of corncobs, a llama bone, and a large grooved stone. T h e lance point is about five centimeters in length, is rather crudely chipped, and has only one shoulder. T h e grooved stone is approximately twenty centimeters in length and fifteen centimeters in

12

EXCAVATIONS AT PUERTO DE SUPE

diameter and is completely encircled by a shallow groove. A second stone of very much the same appearance was found by us on the surface of the site at the opposite side of the quebrada. The use of these stones is unknown. No pottery fragments were encountered in the clearing of the structure. The structure itself proved to be the most interesting feature of the site; it is the first construction of this type brought to light in connection with the Early Ancon-Supe shellmound culture. A detailed report with plans and measurements will be published in the future. In describing the mud and rock walls and the clay "altar" which were uncovered at Aspero, it is necessary at present to write from memory and from the few photographs which are at hand, since our notes are in transshipment from Peru. The remnants of the walls of this structure are less than a meter in height and are composed of uncut stones, crudely coursed, and set in mud plaster, (Plates 3b, 4a). They average about forty centimeters in width. One large room with a prepared clay floor forms a central unit for the structure. This room is roughly four meters square. T w o smaller rooms adjoin this larger enclosure. The small room adjacent to the large room is connected with it by a narrow doorway formed of two stone slabs set vertically in the ground with a stone slab placed in the floor as a doorstep (Plate 4a). All three rooms show some evidences of prepared clay floors which had been built on top of a meter or more of black refuse. The walls extend downward only slightly below the floor levels and are also superimposed upon considerable rubbish (Plate 3b). Within the large room, near the doorway and not quite in the center of the room, was a pediment completely covered with hard clay. This pediment or "altar" was about 1.50 meters square and stood .50 meter above the floor level. The surface and the upper part of its four sides were blackened as though by fire and, probably as a consequence, the clay had the consistency of cement. A quadrant of the "altar" was removed to permit examination of its construction (Plate 4a, 4b). It had been made of large uncut boulders of the type used in the walls of the buildings, set on end to

form a square. Smaller boulders were placed on top and in the interstices of the larger rocks. The interior of the square was filled with dry clay and small rock rubble in addition to a few larger rocks. The whole had been covered over with a coating of clay some five to ten centimeters thick and subsequently firehardened. A second similar "altar" was found a meter or more from the first, within the walls of the same room. This one was constructed in a fashion similar to the first, differing only in being slightly smaller and in being partially destroyed. These rock-walled rooms and the two "altars" probably indicate a religious edifice. If so, it is highly important as the only evidence of this kind extant on the Early Ancon-Supe culture of the Peruvian central coast. T h e assignment of the buildings and "altars" to the Early Ancon-Supe culture is based upon the burials in the adjoining cemetery. The possibility that these burials belong to a culture different from that responsible for the structure has been considered. True enough, there were no typical Early Ancon-Supe sherds, or for that matter any pottery sherds at all, found in the rubbish associated with the structures or "altars." Three small bottle-shaped vessels were found in the twelve graves excavated. One of these was decorated in the Early Ancon-Supe style of incising, and the other two were very similar to this one in form and ware quality. In the immediate vicinity of the cemetery there are no sites other than those of the Early Ancon-Supe Period. Adjoining the quebrada which contains the Aspero site are two smaller quebradas, one on each side of Aspero and both opening onto the San Nicolas fields. They are almost identical in surface appearance. Both of these sites have fire-blackened rock and ash covering the ground; both are dotted with small mounds of rubbish and shell; and both show evidence of crude rock structures. They differ from Aspero in that surface sherds of the Early Ancon-Supe style are found at both locations. Most of these sherds are plain, but about five per cent are decorated with incised lines, punctations, rocker-stamping, or combinations of these techniques. At the site on the north of the Aspero quebrada, designated

EXCAVATIONS A T PUERTO DE SUPE only as L i - 3 1 in our survey, one three-bythree-meter test pit was put down to test the rubbish. Early Ancon-Supe style sherds were found from top to bottom for a depth of two meters. T h e identification of the Aspero structures with this style of pottery, as Uhle has identified the Aspero site with this style, seems probable even though it is unproved. Possibly Aspero represents a stage in the early development of cultures on the central coast w h e n pottery was extremely scarce, and the dwelling sites both north and south of the main quebrada are slightly later in time. T h e grave goods from the Aspero cemetery check very closely with those Uhle found at that site (Kroeber, 1925b), and it seems reasonable to believe that this was the cemetery d u g by that investigator. Furthermore, the textiles and basketry taken from the Aspero graves are virtually identical with those found by us in graves at the Lighthouse site at Puerto de Supe. Small baskets (some containing cotton, others shellfish), fish nets, and gourd containers and bottles were found besides the pottery bottles mentioned above. In almost every grave the body was covered or underlaid by a simply woven mat of rushes. T h e bodies were textile wrapped, usually one to a grave, and in a flexed or crouched position. Interment was direct in sterile sand at a depth of about one meter. T h e Lighthouse site (map, p. 7, No. 1) is in a quebrada no more than fifty meters wide which opens at one end onto a small rocksheltered beach. Evidence of occupation begins about fifty to seventy-five meters back from the beach and extends for at least a hundred meters back u p the ravine. Still farther u p the ravine, in a clean sandy area, Dr. Newman, through test excavations, located the cemetery. In the small knolls of shell and other detritus, two three-by-three-meter test pits were made. One of these ran out into sterile sand at a depth of only two meters; the other was continued down to about 4.25 meters below surface, where it was necessary to terminate the excavation before sterile soil was encountered. Judging from the nature of the soil profile in the completed pit, it seems likely

18

that the deeper excavation was approaching sterile sand when the work was stopped. Sherds were found at all depths in about the same quantity in both pits. Plain ware predominated, but incised, punctated, and rockerstamped material was plentiful enough to give one an idea of vessel decorative patterns. T h e most common form was a flattened globular bowl with a thickened but otherwise unmodified lip. Uhle has referred to this rim form, characteristic of the Early Ancon-Supe style, as the "comma-shaped" lip. Judging from the samples obtained, the pottery from the Lighthouse site and that from L i - 3 1 , near Aspero, are identical. It is the impression of the junior author, although the Supe test pit materials have not yet been examined in the laboratory, that no changes in type or style will be noted to correlate with depth. A l l pits were excavated in .25- or .50- meter levels to check such a contingency, however. Obviously, our work in the village rubbish material in the Puerto de Supe sites was of a brief and exploratory nature. Future excavations at all of these sites and perhaps at others yet unnoted in the region will be of importance to any thorough study of the Early Ancon-Supe horizon on the central coast. T h e burials at the Lighthouse site were similar to those at Aspero in basketry, textiles, nets, and matting. Pottery was a little more abundant in the Lighthouse site graves, and the flat, open bowl form, either incised or plain, was the only shape encountered. In one grave at the Lighthouse site a thin sheet of beaten gold was found with a burial, as was also a feather headdress. Uhle mentions one of the latter from his Supe excavations (Kroeber, 1925b). Perhaps the most important difference between the Lighthouse and Aspero sites is that some of the burials at the former site were cremated or partially cremated, presumably in situ, whereas all the burials at the Aspero site were simple inhumations. T h i s , as well as the scarcity of pottery at Aspero, may be indicative of a difference in time. It should be mentioned that rock alignments on the surface, indicating structures, were as common at the Lighthouse site as at Aspero. In a number of instances these seemed

i4

EXCAVATIONS AT PUERTO DE SUPE

to form conjoined rectangles, implying compound dwellings or buildings. As so little is known regarding the houses of this early

period, a portion of any future excavation at this site should be concentrated upon clearing some of these structures.

Excavations in the Ancón Shell Mounds

C

our investigations of the distribution and period of the incised Early Ancon-Supe pottery and the problem of the early shell mounds on the central Peruvian coast, the last series of excavations of our research program was made at A n c o n (map, p. 7, No. 13). A b o v e the modern town of A n c o n and the ancient Necropolis, which lies just back of the modern town, is a high hill forming the southern boundary of the small, dry valley of Ancon. T h i s hill is covered with finely broken shell, giving the appearance, from a distance, of a white capping. T h i s white capping not only crowns the top of the hill but extends inland or eastward over the southern flank of the series of hills or ridge which is a part of the southern wall of the A n c o n valley. From the crown of the highest, most westerly or seaward hill to the limit of the shell to the east is a distance of some three hundred meters. T h e shell beds slope downward on the hill and the side of the ridge for about a hundred and fifty meters. T h e i r northern, or downhill, edge is just a few meters above the railroad line which comes into A n c o n from L i m a and follows the low ground at the foot of the hill. T h e lower limit of the shell deposits and the nature of their cultural materials are attested by reports from some of the inhabitants of a few fishermen's shacks, which stand on the south side of the railroad a quarter of the way up the hill, that pottery and skeletal materials were encountered in the digging preparatory to the building of their houses. T h e i r description of these materials fits that of the Early Ancon-Supe Period. ONTINUING

Uhle's excavations in these shell mounds were admittedly limited and of short duration (Uhle, 1912). H e found only two burials, both of these in the rubbish and unaccom-

panied by grave goods. H e recovered a number of pottery sherds and some bone artifacts from the midden. According to his data, the refuse averaged two to three meters in depth. T h e excavations made by the Institute of A n d e a n Research were confined to the short period of three weeks, during which time we employed only four workmen; Dr. M. T . Newman and the junior author supervised the work. In that time Newman tested the small sandy quebradas which extend southward u p into the hills and ridges as likely locations for a cemetery. It will be remembered that the burial grounds at both of the Puerto de Supe sites were located in such side areas out of the main occupational zone. However, some ten test trenches, all carried down below a meter in depth, failed to reveal any burials or even an occasional fragment of pottery, textile, or human bone. East of the main hill on the southern flank of the ridge the shell is so deposited as to form a series of irregular terraces, each several meters in width. A test pit was sunk in each of three of these terraces. T h e first pit was put down one hundred meters or so from the crown of the main hill to the east, the second was excavated another hundred and fifty meters farther east, and the third was made in a lower terrace just a few meters downslope from the second pit. It was necessary to carry Pits 1 and 2 down to the unusual depths of 8.90 and 5.00 meters respectively. Sterile sand was encountered at these depths. Pit 3, which was of very small surface dimensions, was abandoned at a depth of about 3.50 meters below the surface. T h e two m a j o r test pits were each threeby-three meters in size at the surface and were sloped inward just enough to allow for safety from cave-ins. A great number of sherds

i6

EXCAVATIONS IN ANCON SHELL MOUNDS

were taken from each pit. All materials were saved by .25- meter levels for future stratigraphic analysis. The rubbish in the two pits was quite similar. In both, the upper meter, or less, was the same mixture of pulverized and broken shell and sand that is observable on the surface of the site. Below this was tightly packed shell and decayed organic refuse. This varied in depth and was underlaid by a brownish sandy and dusty rubbish which in some places was very loose and soft and in others quite compact. Thinner bands and lenses of shell were also encountered, occurring alternately with brownish and grayish dusty rubbish at greater depths. In Pit 2, clean yellow sand began to appear at a depth of 4.50 meters and all trace of refuse and cultural materials had disappeared at 5.00 meters. In Pit 1 this same clean sand was not reached until a depth of over 8.50 meters below surface had been attained. In both excavations the Early Ancon-Supe style of pottery was abundant in all levels. The ware is either black, brownish red, mottled with black, or reddish. It is medium to thick, grit tempered, and the surfaces are usually well smoothed and sometimes polished. T h e common forms, suggested by rim and basal sherds, are flattened globular bowls with thickened or "comma-shaped" lips, bottles with short, wide, everted-lipped spouts, and shallow bowls with flattened bottoms and direct, outslanting rims. Decorated material is a small but noticeable percentage of the total sherd count. Broad, round-bottomed incised lines are devoted to rectilinear and curvilinear designs. Large and small dot punctations are common, and plain and notched rocker-stamping is used alone or in conjunction with incised line designs. Some of the designs are simple, but others are very involved and complicated. Zoomorphic representations are present, among which the feline face is to be noted. One excellent fragment from a stirrupmouthed jar, decorated with incision and rocker-stamping, was found in one of the test pits. At Ancon, as at Puerto de Supe, the pottery recently recovered seems to check very closely with that found by Uhle at the two sites (Uhle, 1912; Strong, 1925). Sherds bear-

ing red paint were exceedingly scarce in our collections as well as in Uhle's. We recall only one or two out of several thousand from each site. Decorated sherd material from Ancon was more plentiful than from the Lighthouse site at Puerto de Supe. The Ancon site is, of course, more extensive than the Supe site, but our sample from Ancon was no greater than that from Supe. Uhle stated that the pottery from the two locations was similar, with only very slight differences. Our impression, prior to detailed analysis, is that Early Ancon ceramics seem somewhat more elaborate. In this regard Early Ancon may more closely approach the Cupisnique style of the north coast of Peru. The only burials we encountered at Ancon, like those earlier encountered by Uhle, were in the rubbish. We located an adult and a child wrapped in simple cotton textiles some 3.5 meters deep in a subsidiary excavation to facilitate the removal of dirt from Pit 1. Both had been buried in a flexed position and neither had any accompanying grave goods. In Pit 1 were the only evidences of structures or what might possibly have been structures. These were crude rock walls, of dry construction, through which the excavation cut at a depth of about 2.50 meters below surface. It is possible that these were cairns or fireplaces. A small amount of ash was found in conjunction with the rocks in one place. However, the alignment of the rocks as they were exposed in the pit gave strength to the supposition that they were parts of a wall or walls. In this connection it should be mentioned that the surface appearance of Ancon is quite unlike that of either Aspero or the Lighthouse site at Puerto Supe. The ground at Ancon is smooth and shell covered. Except for the terracelike ridges, there are virtually no noticeable aboveground features at Ancon. There are no surface rocks or rock alignments suggesting buildings as at Supe. It may be that because of its topographical position on the slope all such features have long been covered over by wind erosion. Other than pottery, only a few artifacts of interest were found. Two of these were pyrites mirrors. One surface of the artifacts still retained its beautiful luster and high polish;

EXCAVATIONS IN ANCON SHELL MOUNDS the reverse side of each was unpolished, was rounded in contour, and appeared to have been set at one time in some plastic substance. T h e mirrors were ovate-rectanguloid and about ten by five centimeters in size. A carved bone spoon was one of the most unusual and interesting of all the specimens. This artifact, perhaps fifteen centimeters in length, was covered on its top surface with deep, roundbottomed incised lines which formed a design suggesting the pottery decoration of the Early Ancon-Supe style. T w o worked stones were also recovered from the pits. These were roughly circular and had a worn pit in the center of each surface. The edges showed some marks of abrasion and pecking. In the entire Early Ancon site, only one specimen was found which suggested any other major style, culture, or period of the central coast of Peru. In the top meter of Pit 1, a sherd with black and white on red decoration

17

came out with the other material. The ware of this sherd was thin and hard. The red was light in tone and well surfaced. A purplish black line and a white line, side by side, paralleled each other. The importance of such a small and solitary fragment would not warrant mention except that the Early AnconSupe style has been so uniformly found in complete disassociation from any other known style that even the occurrence of such fragmentary evidence may be worth noting. It is difficult to place so small a fragment, but its closest affiliations would seem to be with the three-colored pottery of the Interlocking or pre-Interlocking period of the Chancay valley (see Willey, Part 3, below). Such, in briefest form, are the outlines of our recent stratigraphic testing in the middens and shell heaps of Supe and Ancon. Further results must await full laboratory analysis of the materials themselves.

The Southern Survey

D

the period of our intensive work I at Pachacamac, eight brief survey trips to the south were made. These ranged from one to six days in duration. The majority of these trips were made under the direct guidance of Dr. Julio Tello. Dr. Samuel K. Lothrop, Dr. Luis E. Valcàrcel, and Senor José Respaldiza also guided or accompanied us to a number of interesting sites. Members of these parties included at different times Dr. Marshall T . Newman, Dr. Theodore McCown, and the authors of this volume. In the following pages we will discuss briefly a number of the sites visited, proceeding in general from Lima to the south. As one travels along the highway from Lima to Pachacamac and enters the desert strip about halfway between these two places, a more or less discontinuous line of white areas can be seen extending for several kilometers along the sand-covered hills about a kilometer or less to the east of the road. Junius Bird, who examined many of these areas, reports them to be shallow shell heaps ranging from thirty to sixty centimeters in depth. The surface shell gives the white color which can be seen from a considerable distance. Here Bird found a few crudely chipped stone artifacts and a few incised potsherds of the Early Ancon-Supe type. This occurrence of early shell-mound ceramics on old camp sites so close to Pachacamac is interesting. Larger shell heaps characterized by this type of pottery are reported also from near Chorrillos, but we did not examine these. At Pachacamac, Dr. Tello showed us an interesting portion of the site where pottery does not seem to occur. This is on the northern bank of the Lurin river at the extreme eastern end of the ruins about a kilometer and a half east of the highway. The river banks are steep here and capped by about one and a half to two meters of midden deposit forming three or four shelves. The midden material URING

contains much shell, burned broken stone, ash, burned clay, and broken bone. Both animal and human bone fragments were noted amidst layers of both whole and fragmentary mussel shells. A careful examination of this entire face and some two hours' digging with two workmen failed to yield a single recognizable artifact or potsherd, although the deposit is obviously of human origin. Near by are some buried adobe structures, and about ninety meters to the northeast is a looted "Middle Period" cemetery. However, aside from a few obviously superficial sherds we encountered no artifacts in the refuse itself. This fact is difficult to explain. Either the refuse was laid down in some strange manner that excluded broken pottery and other artifacts, or it may be very ancient and culturally limited. Careful excavation, accompanied by screening, will be necessary before this problem can be answered. All other refuse deposits that we examined at Pachacamac are full of broken pottery and other cultural materials. Between Lurin and Chilca, in a picturesque cove of the Pacific, is the little fishing and, more recently, resort village of Cruz de Huesos (map, p. 7, No. 16). Here, capping the steep dunes, are low shell heaps one to two and a half meters in height. Some years ago Señor José Respaldiza trenched one of these middens, encountering Early Ancon-Supe ceramics and cultural materials. We did a little digging and surface collecting, securing a considerable number of plain sherds but no obviously diagnostic material. Near the shell heaps are a number of recently opened round cists of boulders said to be graves. No trace of their contents remained. The Cruz de Huesos site is obviously important in relation to the old shell-heap culture of the coast. Proceeding on toward Chilca, we looked for a Late Period cemetery shown to us earlier by Dr. Tello. Searching for this, just south of the 66-kilometer highway marker, we explored a

THE SOUTHERN SURVEY dry barranca which had a depth of between four and five meters. The northern bank of this barranca is here capped with refuse materials ranging from pne to two meters in depth. The party, composed of some six persons, searched the face and talus of this extensive midden for a considerable distance but found no pottery or artifacts. The composition of the midden is similar to that of the one at the eastern end of Pachacamac, previously described. However, just as we were leaving, Señor Respaldiza discovered three cut and nested calabashes at the base of this refuse at a depth of 1.8 meters. These were obviously of human manufacture, but no other artifacts were noted at this time. On a later visit, Newman and Willey did find a few pottery fragments at the upper or eastern end of this barranca. These sherds had the swollen or "comma-shaped" lip characteristic of the flattened globular bowl form occurring in the early shell heaps at Ancon and Supe. It is obvious that this Chilca site should be helpful in tracing the distribution of this early horizon and, when properly investigated, may cast light on still earlier stages. Slightly south of this site, just to the east of the highway at the 68-kilometer marker, is the Late Period cemetery referred to above. This has been badly looted, and many skulls, as well as ceramic and textile fragments, litter the site. Graves are ovoid or rectangular and are excavated to a depth of two to three meters into hardpan. T h e looted graves have pole roofs covered with reeds. Dr. Tello says that this is a Late Coastal and Inca site representing a period which has as yet received little study. Coastal brachycephals and highland dolichocephals are said to be found here. A number of skulls showing cradle-board deformation were noted. Dr. Hrdlicka made a collection here many years ago, and Dr. Newman gathered a small series during one of our visits. We were first taken to the site of Banduria by Dr. S. K. Lothrop. This large site (map, p. 7, No. 17) is on a high ridge facing the sea. There are some large adobe-block structures at the eastern foot of this ridge, and the eastern slope is marked by very extensive refuse. The main ruin is on top of this steep

19

ridge and consists of an adobe-block temple, vast partially looted graveyards, and abundant refuse which appears to be rather deep in places. The main site looks relatively late, but early deposits may be present. We turned over to Dr. Lothrop our sherd and other samples, since he hoped to make a more thorough study of the ruin. Proceeding south along the highway from Chilca, and turning off on a small side road just before descending into the Mala valley, one comes to the grim gorge of Ollería. This rock-walled little canyon is without vegetation, and its sides are lined with bleached skulls many of which have been rolled down into the road. The bleak gorge notched by the gray sky and filled with bleaching skulls would furnish morbid inspiration for a Doré. The place has been sadly looted, and no very distinctive sherds were encountered during our brief examination. The beautiful Mala valley has as yet received almost no scientific attention, though ruins seem to be both numerous and extensive. At the northern edge of the valley near the sea is the site of El Salitre (map, p. 7, No. 19) whose most striking feature is an adobe-block "sun temple" crowning a ridge facing the sea. Somewhat farther to the north is another ridge marked by a large number of looted graves. There is considerable refuse below the "sun temple" and elsewhere, but no distinctive sherds were collected. The site, on superficial examination, appears to be late. Across the central valley floor, just back from the sea, is a series of ridges which are said to contain a number of ruins having vast and apparently deep refuse deposits. We did not go to any of these ruins but did visit a picturesque site at the extreme southern edge of the valley directly above the rocky coast. This, called Culeta de Burema (map, p. 7, No. 20), is very close to a small country place recently built by the late Dr. Moisés Saenz, the eminent Mexican statesman and sociologist, who in June, 1941, kindly invited the authors, in company with Dr. Luis E. Valcárcel and Dr. Bernard Mishkin, to visit him here. This most pleasant visit also gave us an opportunity briefly to reexamine Culeta de Burema and some adjacent sites. T h e Culeta

20

THE SOUTHERN SURVEY

de Burema site consists of refuse heaps and some rough stone structures located in a barren hanging valley terminating in cliffs that descend abruptly to the sea. A slope leads down one side to a small rock-bound harbor now used as a guano port. T h e refuse consists of consolidated shell and detritus which, looked at from farther along the cliff, appear to have great depth. However, it is possible that this appearance of depth is an illusion due to the overhanging midden masking the upper extension of the rocks. Only excavation can determine the facts. Some more or less rectilinear rough stone and boulder enclosures suggest houses, and some smaller, more tightly built structures may well be looted tombs. Potsherds seem scarce on the surface here, and no diagnostic sherds were encountered. Just back of, and above, Dr. Moisés Saenz's house is a burial ground where a few textile-wrapped mummies have been excavated. Miniature tear-shaped pots and small figurines of apparently late types, as well as a few other artifacts, occur with the bodies. There are also some partially covered adobe structures the purpose of which is not clear. Obviously, there is much to be done in the Mala valley; but time was lacking for any excavation here. We visited the Cañete valley twice in company with Dr. Julio Tello, but did no actual digging there. Sites visited in the lower valley included those at Cerro Azul, the remains of La Centinela, and the adjacent great adobeblock structures which have been briefly described by Kroeber (1937). So far as is known at present, all of these represent a late period but the latter structures are enormous and merit much more extensive study than they have yet received. Recently, stone walls including both rough and finely cut stone have been reported here. We passed the Cerro del Oro site where Kroeber (1937) excavated Late Period graves overlying those of the Middle Period, and briefly examined La Muralla, the great wall or roadway which winds its sinuous way for a long distance along the valley's edge. This wall is made of large frame-made adobe blocks and averages from one to three meters in width, with occasional wider buttresses or ramps ascending to the top. Its height varies greatly, following the contour of the country.

According to Dr. Tello, it probably served as a highway to the interior. In the middle valley we visited the great fortress-temple of Húngara (which has also been called the Fortress of Chuquimancu; see Means, 1931, p. 191). This consists of a series of impressive adobe-block structures atop a high hill which dominates the central valley. A sort of observatory crowns the highest point, overlooking several lower plazas and structure-crowned ridges and spurs. Several graves had recently been opened in the largest lower plaza, and many textiles had been discarded by the huaqueros. Pottery was not abundant, but among the more plentiful plain sherds were a few decorated pieces. T h e ceramic and other sample collections from this site have not yet been analyzed. Earlier in the season Junius Bird and the authors accompanied Dr. Tello on a quick trip through the lower portion of the upper valley. We visited Cerro Hueco, an Inca site built of adobe and rough stone perched on the side of a cliff facing a beautiful stretch of the Cañete valley. T h e ruin is divided into three units, each with its plaza. T h e walls of the most westerly of these have numerous small square niches, and their tops are decorated with tilted adobe blocks forming triangles (Plate 5a) as at T a m b o Colorado in the Pisco valley. Just above Cerro Hueco is the larger ruin of Inca Huasi (Plate 5b, 5c), which is well preserved and is laid out in the best traditional Inca fashion. T h e ruin consists of four main parts: a terraced platform-capped hill on the north, which Dr. Tello designated as the Cusi Pati (Plate 5b, middle background); a large terraced structure to the west, which he designated as the women's house or Mamacona (Plate 5b, lower foreground); a great court or marketplace foundation measuring some two hundred by one hundred meters slightly up the valley to the south (Plate 5c); and, finally, a series of rooms and courts on the eastern rim of the little side valley where the ruin is located. Potsherds were rare, and those that we found seem nondistinctive. A t the time of our visit, July 16, one portion of the ruin was occupied by a picturesque band of lomeros or pastoral Indians from the highlands. At this time of year

THE SOUTHERN SURVEY they drive their herds down into the lower valleys and lomas to pasture on the pasto or fog plants which for a time make the dry hills verdant. W e saw many of these colorfully costumed people in the Cañete and adjacent valleys, and Dr. T e l l o , w h o had shared this type of life, gave us a vivid picture of their wanderings. Before leaving the Cañete valley we drove up to Machu R a n g a where some of the old andenes or terraces are again being irrigated and put to modern use. T h e mouth of the lea river marked the most southerly point we reached during this year's activities. W e were anxious to visit this rather inaccessible location to inspect the great shell heaps described by Uhle (Kroeber and Strong, 1924b, p. 123). Under Dr. Tello's guidance, two trips, the first from July 15 to July 17 and the second from July 29 to August 3, 1941, were made to this place. W e will first mention, in order from north to south, certain sites visited en route. Each time we visited the lea valley we went by way of the very important Paracas site just south of Pisco (map, p. 7). T h i s great Necropolis is famous as a center for the early and rich Necropolis and Cavernas cultures (Tello, 1929). T h e former culture, represented by crowded burials in old subsurface structures, is characterized by rich textiles with designs similar to those from Nazca and a simple polished but unpainted ceramic type. T h e Cavernas culture from closely adjacent deep tombs lacks these elaborate textiles but has a highly distinctive incised and painted ceramic style. T h i s last style has close affiliations with that of Chavín de Huantar. T h e Paracas site is located on the desolate sandy Paracas peninsula. T h i s region possesses a strange, lonely charm due in part to the ever-changing light effects over land and sea. Rosy flamingos and clamoring wild fowl along the wind-swept beaches add to this effect. T h e bleak terrain of sand dunes and rocky hills, however, would seem to offer little to a settled population. Surface materials are not abundant at the site. O n an old beach line near the Cerro Colorado, Bird found a crudely chipped and highly patinated stone artifact, suggesting that adequate examination might reveal traces of

21

still earlier man in this vicinity. T h e Paracas site is guarded by the Peruvian government, for it is obvious that in addition to the important results already attained here by Dr. T e l l o much work yet remains to be accomplished in the vicinity. T h r o u g h the kind invitation of Señor Pablo T r u e l l and his charming wife it was possible for the senior author, in company with Dr. T e l l o and his party, to spend several days at the well-known Hacienda Ocucaje, located in a green valley south of lea (map, p. 7, No. 21). Of late years huaqueros in the vicinity of Ocucaje have been turning u p with pottery, textiles, and other materials in both the Necropolis and the Cavernas style. Señor T r u e l l has preserved many of these, and his collection is tremendously important since this is the first time the Cavernas and Necropolis styles have been reported outside of the Paracas site. O n the invitation of Señor T r u e l l , Dr. T e l l o has recently made extensive studies of these rich materials which will shed new light on various phases of the cultures. Guided by Señor T r u e l l , we visited a number of sites in the Ocucaje valley. Judging from surface collections, the cultures represented at these sites include Inca, Late and Middle lea, Nazca, Necropolis, and Cavernas. One of these sites, L a Capilla, seemed particularly important. O n this hill huaqueros had opened many graves ranging from small, rather nondescript pits to large rock tombs, the latter being made of rough stones cemented with adobe and surmounted by layers of dried grass and refuse. Sherds of the Necropolis and Cavernas cultures are not abundant, since the value of these newly discovered pottery types has made treasure hunters keep all broken vessels. However, at L a Capilla we picked u p small Cavernas incised and painted sherds (the paint having the same lacquerlike appearance as at Paracas); dull, crackly white Necropolis-type sherds; plain brown ware with cuneiform incisions on the inside (a common Paracas type); plain sherds with external incised designs; and a considerable number of negative-painted sherds with dark background and orange-red designs. Over the main hill and adjacent ridges at L a Capilla are scattered between

22

THE SOUTHERN SURVEY

thirty and fifty opened graves all of which are said to have contained Cavernas-Necropolis materials. On the top of the main hill a solid adobe wall about one meter high has been partially revealed. Both deformed and undeformed skulls occur. On the flats below La Capilla hill, graves containing materials of either Nazca or lea type are reported, and some three fourths of a kilometer south, at a place called Cordova, is a small ruin of large adobe blocks which is apparently late, although Nazca-type graves are reported in the vicinity. Other small, rather badly looted cemeteries on hill slopes about two kilometers east of the Hacienda Ocucaje yielded similar materials. Opened graves here were rather shallow, ranging from one to two meters in depth. Some were rectangular and clean-cut, others irregular. All pottery types mentioned above were found here in small quantities, and no Nazca or later types were noted at the site. On the basis of this superficial examination one might tentatively conclude that the Cavernas and Necropolis ceramic styles were contemporary and were probably two types within one major culture, since they occur together at a number of sites at Ocucaje. Further, since no Nazca or later styles occur in this association, it seems probable that the Cavernas-Necropolis styles represent a distinct and early period in the valley. The association of negative painting with this complex has widespread implications. As is pointed out below, a somewhat similar negative painting style occurs early at Pachacamac and in the Chancay valley, while a connection with Recuay to the north and east may also be implied. All these hypotheses await stratigraphic demonstration, but the stylistic implications open wide vistas. Superficial examination of complete pieces in the Truell collection from Ocucaje indicates that many of the incised and painted Cavernas-style vessels have hump-necked bird, human face with projecting tongue, and other designs strongly suggestive of Early Nazca and that the form and surface finish of the Necropolis vessels is likewise similar to the Early Nazca ware. In regard to the Cavernas vessels from Ocucaje, these Nazcoid resemblances oc-

cur in addition to even more striking and basic Chavin designs such as the modeled, incised, and painted feline deity. In other words, these new materials from Ocucaje, which are predominatingly in the Chavin decorative style, also show Nazcoid trends in the increased use of painting over incision and in the occurrence of other motifs formerly regarded as being typical of the Early Nazca style alone. Since this manifestation occurs near the center of Nazca cultural development, Ocucaje being in a sense geographically intermediate between the upper lea and Nazca valleys (see map, p. 7), a reasonable interpretation might be that the Ocucaje Cavernas and Necropolis styles were ancestral to the later Nazca polychrome developments in that region. Wider considerations appear to support this still-unproved hypothesis. On the central coast of Peru, as we have already stressed, occurs the incised (and rarely one-color painted) Early Ancon-Supe style of ceramics characteristic of the shell mounds. This ceramic style also occurs at Chavin de Huantar in the north-central highlands; hence the name "Coastal Chavin" has sometimes been applied to it. The richer Cavernas style is obviously related to this simple Early Ancon-Supe style of the central coast and the north-central highlands, although the exact linkage between the two manifestations remains to be established. On the north Peruvian coast another even more closely related style called the Cupisnique has also been distinguished. Recently Larco Hoyle (1941) has presented quite convincing grave stratigraphic evidence that this Cupisnique incised style precedes the elaborately painted and modeled Mochica or Early Chimu style in the northern region. It is the belief of the authors that similar evidence of the Cavernas incised style preceding the colorful Nazca styles will soon be forthcoming in the south. Uhle's important discoveries demonstrating the relative antiquity of the Early Nazca and Mochica styles in north and south respectively would seem to have blinded him, as well as many others, to the even greater importance of another one of his own discoveries: the occurrence of the incised (and, rarely, painted and modeled) Early Ancon-

THE SOUTHERN SURVEY Supe style in shell mounds on the central coast. T e l l o has long argued the case for the greater antiquity of this relatively primitive ceramic style, and one of the present authors (Strong, 1925, pp. 183, 186), on the basis of Uhle's own materials, many years ago suggested the probability that this style marked the earliest pan-Peruvian horizon yet known. In its relative simplicity it held latent the possibility of a transition from incising to modeling which was apparently emphasized in the later northern development, and from incising to many-colored painting which was apparently stressed in the south. In the north this hypothesis has already been partially demonstrated by objective methods. In the south the discovery of the Cavernas-Necropolis styles isolated in sites in the heart of the Nazca culture area lends support to the hypothesis for that region. Here, however, stratigraphic work will be necessary before satisfactory demonstration is possible. En route to the mouth of the lea river we stopped also at the Hacienda of Callango, formerly owned by Dr. Barranca and now the center for extensive charcoal-burning operations in the almost unpopulated lower lea valley. Callango is downstream from Ocucaje about an hour's ride over very bad roads or sand tracks (map, p. 7, No. 22). In this vicinity there have recently been revealed a considerable number of sites characterized by Middle and Late lea ceramic types (Kroeber and Strong, 1924b) of which we briefly visited about seven. T h e first three of these sites, visited on our first trip to the mouth of the lea river, had only recently been cleared of the low, dense algarroba bush as a result of the extensive charcoal-burning which is further stripping an already eroded and desolate dry country. T h e sites consist of mounds, probably formed on natural hillsides, which are literally covered with broken pottery. On two of these sites near Callango this pottery was so thick as to color the hillock red when seen from a distance. T h e depth of these deposits was not determined. There were a few low adobe structures visible, and the massed sherds seemed to be almost entirely Middle and Late lea in style. On our second trip to Callango, while en-

25

joying the hospitality of the Hacienda Ocucaje, the senior author in company with Dr. T e l l o and several of his assistants visited five other sites characterized by the same general type of ceramics. Proceeding from Ocucaje to Callango, we first visited what may be termed Site A, or the Santos site, if we accept the name of a houseowner who lives near here. Distances and directions were not observed on this hurried inspection trip, and it is essential to secure local guides in any case due to the roughness and inaccessibility of the brushcovered terrain. Site A is a looted graveyard covering several acres. T h e graves, so close together in places as to suggest a gridiron pattern, are both rectangular and square, ranging from one and a half to six meters on a side and one to three meters in depth. T h e majority are square and clean-cut. A few algarroba roof beams were noted, and a few graves contained wall construction of small adobe blocks. Broken pottery and other material were not abundant at this site, but the ceramic sample gathered seemed intermediate in style between Middle lea and Late Nazca. Dr. T e l l o designated this style as Chanka, and his assistants reported finding a few incised and painted sherds of Cavernas type as well. It would be extremely interesting to work such a site as this if it is actually transitional between Late Nazca and Middle lea. From here our guide took us to another looted burial ground about two kilometers away. This may be termed Site B and consisted of about one acre of closely crowded looted graves. T h e graves were mainly cleancut and cubical and of the same general size and depth as those at Site A. Poles and reeds had been used for grave roofing. From the abundance of cultural materials at this site, as well as from the general appearance of the digging, the extensive looting seems to have been quite recent. Undeformed skulls, mummy bundles, textiles, cordage, wooden implements, and broken and some unbroken pottery were abundant. There were numerous large brick-red urns ranging up to a meter or more in diameter. T h e pottery was mainly of Middle lea style with a few fragments suggesting Chanka or Late Nazca types. Woodwork was abundant, including long-handled spades

24

THE SOUTHERN SURVEY

with carved figures on crutch handles, short spades, weaving implements, and numerous large, knob-headed clod-smashers. A few crude, apparently percussion-flaked, stone implements were also noted. Good sample collections were secured by Dr. Tello and by the senior author, but these have not yet been studied in detail. We drove from Site B along deep, sandy trails another kilometer or so to the west (that is, down the dry river course) to a small adobe ruin which topped a knoll or dune close to the steep river bank. This ruin consisted of from six to eight rooms composed of both adobe blocks and other smaller blocks that appeared to have been cut from earth, all being plastered over with mud. Natural boulders which had been used as grinding basins, small stone mortars and pestles, and several large polished and perforated stone maceheads or clod-smashers were noted. The abundant sherds seemed to be mainly Middle and Late lea in style. Not far down the dry lea river course, on knolls just back from the steep river bank, are two more sites of very similar type. Small adobe and mud structures crown the two hillocks, which are only a few hundred meters apart. Natural flat boulders, cupped for grinding purposes, occur at both sites, and between the two sites the steep river bank shows a deposit of densely packed sherds that is at least a meter thick. These, like the abundant sherds on the sites themselves, are mainly of Middle and Late lea types. None of these collections has as yet been carefully analyzed, so the exact range of types is yet to be determined. In a modern shack close at hand Dr. Tello secured a very large collection of whole pottery vessels of Middle and Late lea type which had apparently been looted from these and adjacent burial sites. This excellent collection from the local sites should go far to illustrate the varied vessel forms, sizes, and designs which are characteristic of these distinctive but as yet littleknown Middle and Late lea periods.

such it might be called, took less than a day. As the charcoal burning continues, more and more sites are being revealed, and it is already apparent that in earlier times this part of the lea valley was thickly populated. Today, owing to the almost complete diversion of the water of the lea river in the upper part of its drainage area, the middle and lower portions of the valley, below Ocucaje, are desolate and largely without inhabitants. Even the most cursory examination indicates that this was not always the case. The above notes are the fortunate byproduct of our original desire to inspect the shell heaps at the mouth of the lea river. We did reach this remote place twice (map, p. 7, No. 23; Plate ib). We hoped actually to accomplish some work, but our time in Peru proved too short to permit excavations here, although we regard this as one of the most promising places on the south-central coast for stratigraphic excavation. Because of the remoteness of the site, its inaccessibility by any fairly good automobile road, and its vast size, this task should not be too lightly undertaken. The road from Ocucaje to Callango is rather poor and very hard to follow, but the sandy trails from the latter place to the mouth of the lea river are incredibly bad and necessitate a good local guide. The country of the lower lea valley is very rough, with alternate desert, steep ridges, mountains, and dense algarroba thickets. The trails or roads, if one can so dignify them, have recurrent areas of deep sand, and there is no water between Callango and the river mouth and no certainty of getting good water even there, although brackish wells exist. With proper guidance, enough water for both the engine and the passengers, and a skillful driver to manipulate the steep grades and sloping dune faces which must be crossed, the site can be reached by car in about three hours. A fisherman's shack at the river's mouth was the only habitation we found between Callango and the sea.

It is obvious that even small excavations at sites such as those briefly mentioned above could easily yield important structural, burial and other data on these periods, while the river-bank and mound sites are crying for careful stratigraphic testing. Our entire survey, if

Because of the delays caused by being stuck in the sand and running out of water, our two visits at the site were very brief. The two main shell mounds to the south of the river mouth are very impressive. They can be seen from a considerable distance because of their size and

THE SOUTHERN SURVEY their shell covering, which gleams a pure white (Plate lb). T h e height of the highest mound, according to an aneroid reading made by Junius Bird, is ninety feet. It is impossible to tell how much of this is artificial deposit and how much the nuclear sand dune on which the refuse was deposited. Dr. T e l l o was inclined to agree with Uhle's estimate of great depth for the refuse in these two mounds. Bird and the authors originally guessed at there being only four or five meters of artificial deposit. O n the basis of our later experiences at Pachacamac, at Ancon, and in the Chancay valley, however, we would now be inclined to raise our estimates to depths possibly as great as those suggested by the more experienced Uhle and Tello. However, this is a very important question that only excavation can answer. O n the top of the highest mound are the remains of some seven late aboriginal houses

25

as described by Uhle (Kroeber and Strong, 1924b, p. 123). T h e lower portions of the posts and the vertical tortora or reed walls are still visible for a few inches above the refuse. Scattered around these houses are numerous llama bones, whale vertebrae, and what appear to be the bones of sheep as well. H u m a n bones are also present. T h e surface pottery we picked u p consisted mainly of Late and Middle lea types as well as possible earlier styles, but this material has not yet been carefully studied. T h e Uhle pottery collections from this site at the University of California contain a few Nazca-influenced and incised sherds and there seems a high probability that deep stratigraphic excavations here might well yield a long and important series of superimposed styles. Such excavation is vitally needed in the lea and Nazca regions, and the great shell heaps at the mouth of the lea river should be borne in mind in this regard.

PLATE I a. V I E W O F P A C H A C A M A C LOOKING T O W A R D

THE

FROM "HOUSE

THE

TEMPLE

OF T H E

OF T H E

SUN.

CACIQUE"

PLATE I b. GREAT SHELL MOUNDS A T THE MOUTH OF THE ICA RIVER

PLATE 2 a. W H I T E - O N - R E D P E R I O D G R A V E A T C E R R O D E T R I N I D A D , S I T E E , CHANCAY

PLATE a b.

VALLEY

R E C T A N G U L A R ROCK ENCLOSURE, EAST SIDE OF CERRO TRINIDAD, CHANCAY VALLEY

DE

PLATE 3A.

VILLAGE AND

CEMETERY P U E R T O DE

EXCAVATIONS AT

ASPERO,

SUPE

PLATE 3 b. R O C K - W A L L E D S T R U C T U R E E N C L O S I N G " A L T A R , " E A R L Y SUPE PERIOD, A T

ASPERO, P U E R T O

DE

SUPE

ANCON-

PLATE 4A.

"ALTAR"

A T ASPERO,

LOOKING

CONNECTING TWO

PLATE4b.

THROUGH

DOORWAY

ROOMS

CLOSE-UP S H O W I N G TYPE OF C O N S T R U C T I O N USED IN

"ALTAR'

PLATE5A.

CERRO

HUECO

RUIN IN T H E

CAÑETE

VALLEY

PLATE 5 b. T H E "HOUSE OF T H E WOMEN" A T INCA HU ASI RUIN, CAÑETE VALLEY

PLATE 5 c .

COMPOUND

STRUCTURE CAÑETE

AT

VALLEY

INCA

HUASI

RUIN,

A CERAMIC SEQUENCE AT PACHACAMAC By WILLIAM DUNCAN STRONG and JOHN M. CORBETT

COLUMBIA STUDIES IN ARCHEOLOGY AND ETHNOLOGY, Vol. I, No. 2

Contents Introduction

31

T h e Main R u i n

32

Uhle's Excavations below the T e m p l e of Pachacamac

34

T h e T e m p l e of the Sun

37

O u r Excavations below the T e m p l e of the Sun

40

Ceramic Styles and T y p e s

49

Stratigraphie Distribution of Ceramic Styles (Cut 2)

79

Stratigraphie Distribution of Ceramic Types (Cut 2)

83

Discussion of Ceramic Styles and Periods at Pachacamac

86

Conclusion

91 PLATES

1.

T h e T e m p l e of the Sun at Pachacamac

2.

Excavations below the T e m p l e of the Sun

3.

Cross Sections of Cut 2 Showing Inca Chala and Lowest Adobita

4.

Pachacamac Interlocking Style Vessels, Pit 1; Hand-molded Adobe, C u t 1

5.

Pachacamac Interlocking Style Sherds, Modeled T y p e , C u t 2

6.

Pachacamac Negative Style Bowl Fragment, Inner T r i a n g l e and D o t T y p e , Pit 1

Wall

FIGURES 1.

Simplified Plan of Main Ruins at Pachacamac (after Uhle, 1903)

32

2.

Simplified Cross Section of Uhle's Excavations below the T e m p l e of Pachacamac (after Uhle, 1903)

35

3.

Simplified G r o u n d Plan, T e m p l e of the Sun (after Uhle, 1903)

38

4.

G r o u n d Plan of Present Excavations below the T e m p l e of the Sun, Cuts 1 and 2

41

go

CONTENTS 5.

North Face of Cut 2 below the Temple of the Sun

6.

Cross Section of Soil and Cultural Strata, West End of Cuts 1 and 2 46

7.

Inca Polychrome Style Sherds: Fern, Serrated, Hatched Zone, Serried Diamond, Maltese Cross, and Checker Types

51

Inca Polychrome Style Sherds: Naturalistic, Seminegative, Bold Geometric Types

53

Inca Polychrome and Polished Black Style Sherds: Aryballoid Neck and Nub, Effigy Handles, Hatched Handles, Hatched Lip, Banded Border

55

Inca Polychrome and Inca-associated Style Sherds: Effigy Handle; Aryballoid Nub; Modeled Brown Style, Relief and Pressed Relief Types; White-on-Red Style, Unpolished Type

57

Inca-associated Style Sherds: Polished Black Style, Plain, Modeled, and Incised Types; Plain Red Painting Style

59

Inca-associated Style Sherds: White-on-Red Style, Polished Type; Black and White on Red Style

61

8. 9. 10.

11. 12.

43

13.

Pachacamac Interlocking Style Sherds: Dot, Circle and Dot, Crosshatched, Geometric, and Inner Geometric Types 64

14.

Pachacamac Interlocking Style Sherds: Fret, Inner Fret, Linear, and Inner Linear Types

66

15.

Pachacamac Interlocking Style Sherds: Triangle and Dot, Inner Triangle and Dot Types 68

16.

Pachacamac Interlocking Style Sherds: Black-on-Red, White-on-Red, and Figured Types

69

Pachacamac Interlocking Style Sherds: Flaring Necks, Handles and Spouts; Dot, Triangle and Dot, Geometric, and Linear Types

72

Pachacamac Negative Style Sherds: Inner Line and Dot, Inner Dot, Inner Triangle and Dot, Inner Scroll Types

74

19.

Punctate (and Incised) Type Sherds

77

20.

Diagram of Percentage Distribution of Ceramic Styles, Cut 2

81

17. 18.

TABLES x.

Suggested Sequence of Ceramic Styles and Periods at Pachacamac

90

2.

Occurrence of Ceramic Styles and Types with Style Percentages in Blocks and Strata, Cut 2

92

3.

Provenience of Pottery Illustrated

120

PLATE I .

T H E TEMPLE OF T F

OF THE SUN AT PACHACAMAC

Introduction

T

ruined city of Pachacamac, located some thirty kilometers south of Lima, is one of the most famous archeological sites in Peru. Here, in 1896, Dr. Max Uhle reported the first demonstrable case of cultural stratification in Peru (Uhle, 1903). In 1941 this great site had for several years been the scene of extensive excavations carried out by Dr. Julio Tello, of the University of San Marcos and the Museum of Anthropology, Magdalena Vieja. Thanks to his invitation, it was possible for us to make a stratigraphic cut in rubbish heaps at this important coastal site. In our limited work we were greatly aided by Dr. T e l l o and his assistants, who were carrying on their vast and important excavations radiating out from the so-called Convent of Mamacona (see map, Fig. 1). This cooperation made available to us the services of some four or five of Dr. Tello's trained workmen, a factor which speeded our work considerably. Due to the excessive dust in our portion of the site it was necessary to change our crew rather often so that the men would not be too long exposed. T h e workers generally proved to be skillful, cheerful craftsmen who soon became adept at our necessarily meticulous and painstaking techniques. HE GREAT

T h e primary purpose of our limited exca-

vation was the determination of a part of the ceramic sequence at Pachacamac through the methods of detailed stratigraphy. In the present report we have confined ourselves entirely to the ceramic and sequential aspects of the problem. Studies of the plentiful textiles and many other cultural organic remains have been begun but cannot be included here. In regard to the abundant human skeletal material recovered in one test pit adjoining our main cut, we were assisted by Dr. Marshall T . Newman, Supervisor of Project 8 of the Institute of Andean Research. This aspect of the problem will be treated by him in a forthcoming report. In all the present study we have leaned heavily on the earlier work of Dr. Max Uhle. Our simplified maps (Figs. 1-3) of the great ruined city and the Sun Temple are based on his painstaking surveys. T h e reader is referred to his monumental report (1903) for the descriptive and historical aspects of Pachacamac. Here, as elsewhere in Peru, Dr. Uhle blazed a broad trail for those scientists who follow. So far as the work at Pachacamac is concerned, our indebtedness to Dr. Uhle for his past researches and to Dr. Julio Tello for his present guidance is far greater than any words of acknowledgment can indicate.

The Main Ruin

P

ACHACAMAC, while smaller in extent than Chanchan on the north coast, is one of the larger Peruvian ruins. According to Uhle (1903, p. 10), the city ruins extend over an area some two and one-half miles in length by one and one-third miles in width. T h i s estimate apparently includes outlying structures not indicated on the map (Fig. 1). T h e general description of Pachacamac and its history, from both the documentary and the archeological standpoint, has been presented in considerable detail in Uhle's monograph (1903), to which the reader has already been referred. T h e present brief comments are intended merely to orient our minor excavations in relation to the ruin as a whole. T h e ancient city is located on the western desert edge of the Lurin river valley, on a hilly promontory dominating that charming green valley. At present Pachacamac is separated from the city of Lima and the Rimac river valley to the north by several miles of barren sandy hills but, if the account of the chronicler Estete is to be believed, the two valleys were connected by green fields and groves at the time of the conquest (Uhle, 1903, P- 3)Approaching Pachacamac from Lima, one arrives first at the structure which Uhle designated as the Convent Mamacona (see map, Fig. 1). Here, during recent years, Dr. T e l l o has conducted extensive excavations which reveal amazing new structures and architectural features, including beautifully cut stonework hitherto unknown at the site. These excavations of Dr. Tello's have already added new and important chapters to the cultural history of this astonishing city. From the Mamacona one looks south to the highest hill ( W on map, Fig. 1), which is crowned, some seventyfive meters above the level of the valley, by the great Inca Sun Temple. Despite the ravages of time, earthquake, and treasure hunter, the profiles of this great structure are still im-

pressive and dominating. Its outlines were restored in 1939 by Dr. Albert Giesecke. While the severe earthquake of 1940 destroyed much of the detailed restoration, the battleshipprowed silhouette of the main structure remains intact (Plate 1). Below the T e m p l e of the Sun, and some one hundred and twentyfive meters to the northeast, is the less impressive but extremely important terraced temple structure which Uhle designated as the T e m p l e of Pachacamac. According to Uhle, this was the famed temple of the chief creator god Pachakamaj. I n the words of Estete, "Pilgrims coming from all parts of Peru, three hundred leagues or more, flocked to this shrine, as the Turks and Moors flock to Mecca" (Uhle, 1903, p. xi). Below this structure Uhle excavated in the crowded graves of a cemetery containing Inca and Late Coastal and earlier Tiahuanaco-Epigonal ceramic and other cultural materials. While the exact identification of these various structures with those mentioned in the chronicles is a complex problem on which there is still considerable disagreement, the present authors have followed the designations employed by Uhle, both for the sake of uniformity and convenience and because their own very limited researches do not justify their drawing independent conclusions. T o the southeast the structure designated as the Temple of Pachacamac follows the slope of a hill much of which has been built up of adobe blocks to form a vast structure which is at present masked by earth and rather shapeless. All three of these mounds or pyramids, including the Temple of the Sun, the Temple of Pachacamac, and the great adobe structure on which the latter impinges, are surrounded by a low temple enclosure wall (map, Fig. 1) which, though badly eroded, is still visible. It is interesting to note that the Temple of the Sun, which is presumably (in its outer structure at least) Inca and late in origin, lies in consider-

T H E MAIN RUIN able part outside this enclosure wall, which slightly overlaps the eastern corners of the structure (see maps, Figs. 1, 3). T h e major portion of the city, a vast complex of adobe-block city walls, compounds, and high-walled streets, lies on a lower level to the northeast (Fig. 1 and present volume, Part 1, Plate 1a). According to the Uhle (1903, p. 11), the oldest part of the city lies between the hills X , Y, and Z (map, Fig. 1). T h e even more ruinous nature of this section of the site lends weight to this possibility but, on the basis of published excavation data, it has not yet been clearly demonstrated. Since our own tests were made outside of these sections, we have nothing to add concerning this vast major portion of the city. Undoubtedly, when the present extensive excavations of Dr. T e l l o are completed the relative ages of the different

33

sections at Pachacamac will be much clearer than they are at present. Despite four centuries of almost uninterrupted digging, this great site has as yet hardly been touched scientifically. Its deposits extend down to great depths, and many great structures are still masked by sand dunes or by superficial debris from treasure hunting. Uhle's magnificent pioneering work is being followed up by other excavators, but it will be many, many years before the complete story of this great and ancient city will be available to science. Though our trenches seem almost infinitesimal in comparison with the enormous amount of excavation done by others and still to be done, in their sequential strata they add objective facts to the important story of the rise of the coastal civilizations of prehistoric Peru,

Uhle's Excavations below the Temple of Pachacamac

I

1896 UHLE discovered a case of burial stratification in the crowded graveyard at the n o r t h e r n base of the T e m p l e of Pachacamac (Uhle, 1902 and 1903). Since this graveyard is located only about 250 meters below and to the north of our own excavations (U o n Fig. 1) it may be well to outline briefly his m a j o r results. T h e cemetery is about 750 feet long, with a general width of 200 feet (Uhle, 1903, p. 12). Uhle estimated that there were some 30,000 burials in this one cemetery alone. H e cites Cieza de Leon a n d Zarate to the effect that a final resting-place before the famous shrine was highly desired a n d that burial here was a privilege reserved to princes, priests, and those pilgrims who could bring rich offerings to the sanctuary (Uhle, 1903, p. 16). However, he points out that the great majority of skulls from this graveyard are of the common coastal or "Chincha" deformed type (occipitally flattened), a n d that other types are few and generally limited to skulls f r o m the graves closely adjacent to the f r o n t terrace walls of the temple. These latter he considers exceptional. H e notes few archeological evidences to prove a high reverence for the sanctuary b u t points out that there are marked period differences in the cemetery contents. H e records a few definite Inca burials b u t states that the majority of burials were made in times preceding that period. N

T h e exact nature of Uhle's stratigraphic findings here seem somewhat complex, due to occasional conflicts of textual and figure designations (Uhle, 1903, pp. 19-45 and Figs. 2-5) but may be analyzed as follows. For the sake of convenience, a simplified version of one of Uhle's cross sections (Uhle, 1903, Fig. 3) is given here (Fig. 2). Near the surface in the new soil just before the lowest terrace of the u p p e r T e m p l e of Pachacamac he encountered

individual m u m m y bundles at a depth of five feet which appear from their ceramic associations to be pure Inca and Late Coastal (Fig. 2, No. 2). U n d e r n e a t h the terraces, b u t above the older cemetery (Fig. 2, No. 3), he encountered graves containing several mummies ranging from one and one-half to five feet in depth. These burials contained red and black on white pottery with little polish a n d rather crude shapes (Uhle, 1903, Plate 7, Nos. 1-6). H e states that the surface looks like whitewash with red a n d black designs. T h r e e vessels, ranging from open bowls to narrow-mouthed jars, have conventionalized faces on the neck. One shows a conventionalized cat of Coastal type and another a conventionalized doublesnouted head of Epigonal style. O n e of the modeled heads suggests Late Chimu influence (Uhle, 1903, Plate 7, No. 4). T h e vessels illustrated by Uhle find parallels in Kroeber's Epigonal a n d Three-color geometric styles at Chancay (Kroeber, 1926b, Plates, 83, 84) and in Late Ancon I (Strong, 1925, Plate 43i-p). Graves containing this Coastal type of pottery and, in one case, a T i a h u a n a c o tapestry poncho also occur in the uncovered portion of the graveyard beyond the T e m p l e of Pachacamac. These finds, as illustrated (Uhle, 1903, Plate 8, Nos. 1, 5, 7-10), seem also to include Late Chimu or Chimu-influenced types as well as the red and black on white ware (Uhle, 1903, Plate 8, Nos. 2, 3, 4, 6). According to his caption (Uhle, 1903, Plate 8), he classes all these as belonging to the late pre-Inca period at Pachacamac b u t regards them as imported pieces probably from the north. H e states that the Inca period was not represented in his finds in the open cemetery (Uhle, 1903, p. 41). From what he considers the original burial ground (Fig. 2, No. 4) u n d e r the front terraces of the later T e m p l e of Pachacamac and in

UHLE'S EXCAVATIONS

35

(2) FIG. 2. S I M P L I F I E D C R O S S S E C T I O N O F U H L E ' S E X C A V A T I O N S B E L O W T E M P L E O F P A C H A C A M A C (after Uhle, front of, b u t deeper than, the earlier stone w a l l (Fig. 2, o l d t e m p l e ) h e e x c a v a t e d a n u m b e r of g r a v e s c o n t a i n i n g a f e w C o a s t a l T i a h u a n a c o a n d m a n y E p i g o n a l style c e r a m i c s a n d textiles. T h e s e g r a v e s w e r e of v a r i e d d e p t h s b e l o w the o r i g i n a l surface, r a n g i n g f r o m one a n d one-half to ten feet underg r o u n d . T h e m a j o r i t y of t h e g r a v e s c o n t a i n e d c e r a m i c s in t h e m o r e c o m m o n E p i g o n a l style ( U h l e , 1903, P l a t e 5, N o s . 1 - 7 , 13) b u t o n e int e r m e d i a t e g r a v e (Fig. 2, N o . 4') c o n t a i n e d t e x t i l e s a n d p o t t e r y of the m o r e e l a b o r a t e a n d r i c h e r C o a s t a l T i a h u a n a c o style ( U h l e , 1903, P l a t e 4). T h e s e l a t t e r o b j e c t s w e r e i n t e r m i n g l e d w i t h those of the E p i g o n a l style, w h i c h U h l e c o n s i d e r e d as later, a n d t h e g r a v e cont a i n i n g t h e m (Fig. 2, N o . 4') is s u r r o u n d e d b y d e e p e r g r a v e s (Fig. 2, l o w e s t N o . 4) c o n t a i n i n g o n l y E p i g o n a l types. U h l e t h e r e f o r e c o n c l u d e d t h a t these o b j e c t s i n the r i c h e r C o a s t a l T i a h u a n a c o style w e r e a l r e a d y a n c i e n t a t the t i m e of the i n t e r m e n t s , since a g r a v e m u s t b e d a t e d b y its y o u n g e s t o b j e c t s ( U h l e , 1903, p . 22). U h l e figures a n d describes b o t h t h e C o a s t a l T i a h u a n a c o style, as f o u n d b y h i m at P a c h a c a m a c , a n d its c r u d e r c o u n t e r p a r t o r d e r i v a tive, t h e E p i g o n a l style, at c o n s i d e r a b l e l e n g t h ( U h l e , 1903, p p . 22-34). W e w i l l discuss this c o m p l e x p r o b l e m later.

THE

I9OJ)

I n a d d i t i o n t o the a b o v e c e r a m i c styles, U h l e also illustrates, b u t does n o t d i s t i n g u i s h i n this 1903 r e p o r t , a n o t h e r a n d a p p a r e n t l y e a r l i e r c e r a m i c style. T h i s m a y b e d e s i g n a t e d as t h e P a c h a c a m a c I n t e r l o c k i n g style, f r o m its obvious relationship to similar ware f r o m C h a n c a y ( K r o e b e r , 1926b, p. 279). I n his m a j o r m o n o g r a p h U h l e states t h a t a l a r g e n u m b e r of sherds d e c o r a t e d w i t h " t e x t i l e p a t t e r n s " ( U h l e , 1903, p. 28 a n d Figs. 26-29) w e r e f r o m vessels d e s t r o y e d b y l a t e r b u r i a l s , p r e s u m a b l y of the E p i g o n a l p e r i o d , since it w a s i n t h e E p i g o n a l p o r t i o n of the g r a v e y a r d t h a t h e enc o u n t e r e d t h e m (Fig. 2, N o . 5). T h e f o u r sherds i l l u s t r a t e d ( U h l e , 1903, Figs. 26-29) a r e o b v i o u s l y i n the P a c h a c a m a c I n t e r l o c k i n g style, as are c e r t a i n c o m p l e t e vessels ( U h l e , 1903, P l a t e 5, N o s . 10, 12, a n d , possibly, 6, 7, a n d 8, t h o u g h the last t h r e e h a v e o t h e r aff i l i a t i o n s as w e l l ) . I n his n o t e s o n e x c a v a t i o n s at C h a n c a y i n 1904 ( K r o e b e r , 1926b, p. 296) U h l e states i n r e g a r d to this I n t e r l o c k i n g style: I was attracted by the distinctiveness of style of the fragments in the [railroad] cuts, though at first sight I attributed a post-Tiahuanaco origin to them. A very small number of similar sherds had been recovered at Pachacamac in 1896 at considerable depths between the adobes of the platform of the

UHLE'S EXCAVATIONS

36

a n c i e n t t e m p l e of P a c h a c a m a c [present p a p e r Fig. 2, N o . 5], a n d were r e p r o d u c e d Pachacamac

[Uhle,

in m y

work

1903, Figs. 36-39], since

on they

were different f r o m all o t h e r finds in the t o w n . I h a d then

also assumed

a

post-Tiahuanaco

origin

for

them, suggesting that they were r e m n a n t s of vessels b r o k e n b y the masons d u r i n g

their w o r k o n

the

t e m p l e . N o w I see t h a t this p o t t e r y d a t e d f u r t h e r back, a n d that its f r a g m e n t s were l a i d b e t w e e n the adobes f o r some other reason.

He continues here to give reasons why he believes this Interlocking style to be anterior to Coastal Tiahuanaco and related to ProtoNazca in the south. In a later article (Uhle, 1908, p. 356) he states that: I n central Peru I c o u l d f i n d the oldest vestiges of the l e a a n d N a z c a cultures o n l y i n the most i m m e diate p r o x i m i t y of the coast, in C h a n c a y a n d Pachacamac. I n P a c h a c a m a c o n l y recently, o n a potsherd I f o u n d by chance u n d e r the floor of the lowest eastern terrace of the Sun T e m p l e . Perhaps a more imp o r t a n t cemetery of the earliest P e r u v i a n

culture

h a d b e e n covered b y the I n c a b u i l d i n g .

T h e large rim and body sherd illustrated (Uhle, 1908, Fig. 5) is a fine example of Pachacamac Interlocking. In his 1908 article Uhle continues to point out stylistic and architectural links between the Interlocking and the Proto-Nazca cultures and the fact that the former is older than the Tiahuanaco style on the coast. He particularly emphasizes the common inclusion, by both the Interlocking and the Proto-Nazca style, of the interlocking fish motif (compare Kroeber, 1926b, p. 287 and footnote 42). Since Uhle's discovery of the Interlocking style below the Temple of the Sun anticipates our own findings, it will be mentioned again in connection with that structure. It is rather difficult to sum up the ceramic sequence reported by Uhle from his excavations in front of (and under) the Temple of

Pachacamac because of his inadequate description, particularly of the later styles, and also because of some confusion between ceramic styles and time periods. This matter is considered below; but it may be well to list the major styles reported in this portion of the site and to indicate their general provenience: 1. Inca, Cuzco ceramic types, in superficial graves in new soil in front of the Temple of Pachacamac (Fig. 2, No. 1). 2. A Late Coastal style associated with Inca, apparently containing Late Chimu and nonEpigonal red, white, and black ware, in superficial graves as above (Fig. 2, No. 2). 3. A n earlier Coastal style, including ceramics very similar to Chancay Three-color geometric and Chancay Epigonal with some Late Chimu influences, in graves in the open cemetery and in those deeper than the above style but above the graves containing the styles listed in 4 and 5 below (Fig. 2, No. 3). 4. T h e Coastal Tiahuanaco-Epigonal style, in deeper graves under the front terraces of the Temple of Pachacamac and also, rarely, in the open cemetery (Fig. 2, Epigonal No. 4, Coastal Tiahuanaco No. 4'). 5. Interlocking style, sherds amidst Tiahuanaco-Epigonal graves and between adobes of the ancient Temple of Pachacamac (Fig. 2, No. 5). Obviously, five or six styles are represented here. However, considering the fact that all occur crowded in burials within a limited vertical distribution it may be questioned whether five or six distinct time periods are represented as well. This complex matter of styles and periods is discussed below in connection with Uhle's ceramic materials from other sections of the town as well as with those from our own stratigraphic cut.

The Temple of the Sun JTS C A N BE SEEN from the sketch map (Fig. 1) / \ the southwest corner of the T e m p l e of ^ P a c h a c a m a c is almost exactly one hundred meters from the northeast corner of the T e m p l e of the Sun. Placed on a natural hill, some 250 meters in height, the Sun T e m p l e , like other coastal Inca structures of similar nature, dominates both the ancient city and the adjacent L u r i n valley. As Uhle says, this great adobe-block and unpolished-stone structure has been more frequently described than any other at Pachacamac, and since it was carefully studied and mapped by him (1903, pp. 73-99, and maps, Plates 15, 16, 17), all that is attempted here is a brief recapitulation of certain salient details that have a direct bearing on our own excavation.

T h e T e m p l e of the Sun at present consists of a series of from four to five terraces and ruined structures covering the top of the highest hill, which is at the extreme southern corner of the site. It covers an area of some twelve acres and is built in the form of an irregular trapezoid having, according to Uhle, a southwestern length of 715 feet, a northwestern length of 515 feet, a northeastern length of 565 feet and a southeastern length of 347 feet. T h e temple seems to have been constructed so as to face the sea. T h e southwestern or sea front is the most elaborate structurally, with rows of rooms and niches and, originally, covered galleries. A l t h o u g h the summit is actually about a half kilometer from the beach, the T e m p l e gives the impression, as Squier says, "that it reaches close to the shore, so that the ocean may be said to break at its feet." O n the side facing the ocean the most delightful sea-breeze is felt all day long, m a k i n g the crest of this hill a pleasant place, even on the hottest day. [Uhle, 1903, p. 73.]

T h e height of the artificial structure is variable and difficult to determine save on the southwestern front. Uhle gives the height here as 75 feet, with a variation on the other

three sides of from 49 to 60 feet. O n the northeastern front, just above our Cuts 1 and 2 (Fig. 3), the structure, according to Uhle's profile (Uhle, 1903, Fig. 17), has a height of 60 feet. O n the basis of passageways running up the northeastern front, Uhle postulated that the main entrance was on this side. Since his work at the site, an entrance with a stairway connecting with these passages has been uncovered and restored near the northern end of this front (Plate 1 and Fig. 4). Uhle also postulated another entrance near the southern end of the northeastern front. It is notable that the two main Inca refuse heaps (R, R on Fig. 3) are located just outside and below these entryways (see Plate 1 for north entry). It was in the more northerly of these middens that we made our stratigraphic cuts (Figs. 1, 4, Cuts 1 and 2). In regard to the occurrence of these refuse heaps in front of the temple entrance (or entrances?), it is interesting to note that several rooms had been added to this front subsequent to the completion of the temple (see Fig. 3). T h e fact that the red paint which once covered the major portion of the temple is continuous behind the rooms in question suggests later construction. In regard to the use of the lower terraces as habitations, Uhle quotes the following: R o m a n furnishes an exceedingly clear description of the purpose served by the lowest terraces of the T e m p l e , which may have a deeper foundation. " W i t h i n the two lower terraces," he says, "were the apartments of all the servants, priests and priestesses connected with the T e m p l e . T h e r e were rooms for the manufacture of the T e m p l e decorations, cellars and storerooms for the liquors (vinos), and living and slaughtered animals and birds were kept here for the sacrifices. T h e r e were sacristies where might be seen hangings of wool and cotton of the finest colors and texture, and so the entire edifice was excellently arranged (estava muy claro todo) and everything easily to be f o u n d . " T h i s account, if founded upon truth, w o u l d largely explain the dis-

4 ..

VVK.

CMI- *

,„

I

»•.«»

H

„_

I

ENCLOSURE ftlWALL

I S * '

1

I

ENCLOSURE WALL

'»• 'II W

C U T S 1 A N D 2 - / ^

"

V X ' I ;

.5

I I F J SF: ¿ 5 0 /-£ZT rOO M'S

FIG. 3. S I M P L I F I E D G R O U N D P L A N , T E M P L E O F T H E S U N (A/TER Uhle,

ipoj)

THE TEMPLE OF THE SUN position of the lower terraces of the Temple. [Uhle, 1903, p. 82.] As will be seen in sequel, the content of the northern refuse heap reveals varied materials strikingly reminiscent of the activities postulated by this sixteenth-century churchman and chronicler. T h e r e seems little doubt that the T e m p l e of the Sun, at least as regards the surface structure, is Inca in origin and pertains to Inca times. Uhle points out various Inca features in construction, such as certain types of niches, but he also emphasizes that the structure as a whole actually represents a blending of Inca (i. e., Cuzco) and Coastal architectural styles. T h a t the great superficial structure may mask other older temples is highly probable. A great cut, running from the highest terrace down the northeastern side, made by a treasure hunter some fifteen years ago, revealed structures of hand-molded adobes and refuse containing Interlocking potsherds just below the present floor of the top terrace. Likewise, our own cut just below the base of the temple revealed similar conditions (see Fig. 5). T h i s is discussed below, but the fact that the superficial structure of the Sun T e m p l e was Inca, or at least Incainspired, is borne out by the contents of the upper portions of our refuse heap, which are predominantly Inca, and also by the almost pure Inca burial ground on the lowest terrace of the southeastern front, excavated and described by Uhle (Uhle, 1903, pp. 84-96). T h i s cemetery, which contained only women and children (the former, according to Uhle, probably including sacrificial victims from the highlands), showed an almost pure Cuzco type of culture. T h e pottery (compare Uhle, 1903, Plate 18, Nos. 1-10) compares with the best and most typical specimens from Cuzco. T h e textiles, not often preserved in the highlands, are said to be of highland types, the skulls show a variety of noncoastal deformations, and the vegetal foods deposited with the dead are likewise predominantly highland

39

rather than coastal in origin. Uhle estimates that only two percent of the objects found in the graves are of local—that is, coastal—types. H e implies that the majority of the objects were imported from the highlands rather than locally manufactured; but this important point will remain open to doubt until detailed comparisons have been made. Without entering further into the various theoretical issues raised and discussed by Uhle in connection with this most important burial ground, it will suffice to point out the predominatingly Inca character of this cemetery lying entirely within the precincts of the Sun T e m p l e . Even here other associations do occur, as indicated by illustration of two Late C h i m u type pots from this burial ground (Uhle, 1903, Plate 18, Nos. 11, 12). In graves of the outer city Uhle also found Inca type ceramics associated with various local and Late Coastal type vessels (Uhle, 1903, Plate 13). These associations and types are discussed below in connection with Uhle's ceramic discoveries below the T e m p l e of Pachacamac and with our own investigation in the refuse heap below the T e m p l e of the Sun. Here it is important to point out that the ceramics encountered by Uhle at the T e m p l e of the Sun were predominantly Inca or Late Coastal, with one important exception. T h i s was his later discovery, below the surface of the lowest eastern terrace of the T e m p l e of the Sun, of one very characteristic fragment of Interlocking type pottery (Uhle, 1908, p. 356, Fig. 5). T h i s sherd he considered as representing the earliest Peruvian culture known to him at Pachacamac, and he stated that the Inca Sun T e m p l e might cover a cemetery of this early culture. T h i s prediction was borne out by our discovery of a crowded tomb of the Interlocking period in T e s t Pit 1 (Fig. 4) below and in front of the Inca refuse heap on the northeastern front of the T e m p l e of the Sun. T h i s possibility had also been pointed out by Dr. T e l l o when he advised us as to where to make our cut.

Our Excavations below the Temple of the Sun

T

location of our main cut at Pachacamac was determined by the occurrence of a large, dark midden below and to the south of the main entrance to the Temple of the Sun (Plates 1, 2 and Figs. 1, 3). This is the most northerly of the two middens clearly indicated on Uhle's map (see R on Fig. 3). Abundant surface material here was all of Inca type, but, for reasons already mentioned, we hoped to encounter other cultural strata below. Our cuts were oriented in relation to the outer wall of Terrace 1 of the Sun Temple. Since the nearest corner of the Sun Temple points almost exactly north, the terrace wall (Terrace 1) on the northeast side of the structure actually runs from northwest to southeast (Figs. 3, 4). T o facilitate description, however, we have here used only cardinal directions in referring to our cuts and faces, as though this northeastern wall of Terrace 1 actually ran from north to south and our cuts from east to west. Their exact alignment in each case may be determined by reference to the diagram (Fig. 4). HE EXACT

Our procedure in excavation included four stages: first, the orientation of our excavations so they would cut through a high undisturbed portion of the midden and also miss several huaquero or treasure-hunter holes to the south; second, the digging of a small series

of test pits along this orientation line beyond the midden to determine its extent and the nature of the underlying terrain; third, the excavation of Cut 1 to the north of this line (Cut i would cross-section the midden and all underlying artificial deposits as rapidly as was possible, considering the obvious mass of material to be removed; all cultural features and all materials from this first cut were to be noted and then removed, but stratigraphic levels were not to be maintained); fourth, the making of Cut 2 to the south of the central line (which would mark the vertical cross section revealed by Cut 1, see Fig. 5), carefully removing all materials by stratigraphic blocks (see Fig. 5). Since Cut 2 would be merely a meter-wide extension of Cut 1, the pottery collections from each cut should be strictly comparable. It would thus be possible to utilize a ceramic style and type classification derived from the abundant sherds from the larger cut (Cut 1) in classifying the sherds in each stratigraphic block of the smaller cut (Cut 2). Due to the great depth of the deposit revealed in the middle stage of completing Cut 1, it proved most practicable to remove about half of this, then remove half of Cut 2, then complete Cut 1, and finally complete Cut 2. In the following description of the four main stages of excavation we will describe each of these operations as a unit.

ORIENTATION T o line up the cuts and the preliminary test pits, a reference point was established on the outer wall of Terrace 1, forty meters south of the southern corner of the main entrance of the Sun Temple (R x on Fig. 4). From here a line was marked out at right angles to the

temple wall, marking the division point between Cuts 1 and 2 (R 3-4 on Fig. 4). Because of the slope and owing to the need of narrowing the trench toward its eastern extremity this line later deviated slightly to the north.

EXCAVATIONS A T TEMPLE OF THE SUN

SUN

41

TEMPLE ENTRANCE

CUT 2

scale 2

;m

m FIG. 4. G R O U N D P L A N OF P R E S E N T E X C A V A T I O N S B E L O W T H E T E M P L E OF T H E SUN, C U T S 1 A N D 2 TEST These were originally laid out on both sides of the line R 3 - 4 and its projection (Pits 1 - 5 , Fig. 4). T h e i r original size was two by two meters in surface extension, and they were excavated by half-meter levels. One of the pits (1) was subsequently enlarged (Fig. 4). While the materials from these pits are not considered here in detail, their general characteristics may be mentioned. Pit 1: T h i s turned out to be a crowded tomb bounded on the east by a rough stone wall (see Fig. 5, "rock wall," eastern end of Cut 2). Despite the small size of this excavation even when enlarged (Fig. 4; the area B was not completely excavated), it yielded over one

PITS hundred skeletons, most of them in excellent condition. T h e majority of these skeletons were removed by Dr. Marshall T . Newman and will be reported on in another place. Similarly, the associated pottery (Plates 4, 6) and other artifacts, a considerable number of which are not at the moment available for study, will also be reported on in full at a later time. T h i s burial area impinged only slightly on the southern edge of Cut 2, and its main extension was to the south and east. It was not exhausted when our work at the site terminated. T h e burials here were associated with pottery of the Pachacamac Interlocking style.

42

EXCAVATIONS A T TEMPLE OF THE SUN

Pit 2: T h e upper layers of this pit were composed of dust, consolidated soil, some cultural detritus, sherds, and undressed structural rocks. A t a depth of 2.50 meters sterile sand was encountered. This sand had the same southeastward contour of slope as has the present surface. Pit 3: T h e upper 25 centimeters of this pit were composed of dust and small rocks. An irregular layer of gray-black ash and consolidated refuse extended below this to a depth of 1.30 meters. Beneath this was a clayey layer of light-brown earth which continued to the bottom of the cut at 2.70 meters. A t a depth of 1.80 meters in this layer a child's burial and fragments of a large red olla were encountered. T h e bottom 30 centimeters of this layer were culturally sterile. Pit 4: T h e top 50 centimeters of this pit consisted of dust and scattered rock underlain by a loose rock and dirt formation which became consolidated at 1.70 meters. Clean, sterile, brown sand was encountered at 2.25 meters. Immediately above this was a streak of gray-black ash. Cultural detritus but no diagnostic sherds came from this pit.

Pit 5: T h e upper levels of this pit were similar to the above. A t a depth of 85 centimeters remnants of an adobe floor and a large stone mortar were encountered. Sterile brown sand occurred at a depth of 1.50 meters and extended to the bottom of the excavation at 2.50 meters. As a group, the test pits indicated that east of the main midden Inca deposits were superficial and underlain by another cultural horizon. In ceramics, specifically, the pits yielded a few Inca and Inca-associated sherds from near the surface, with Pachacamac Interlocking and Negative style sherds below. This gave promise of stratification beneath the midden itself. With the exception of Pit 1, where the main burials extended to the south of our proposed cut, the pits indicated that in all probability no important structures or grave concentrations would be covered by our dump heaps in this tested area or encountered by extending our cuts. T h e ceramic materials from certain of these pits, with the complete vessels and other artifacts from Pit 1, are less fragmentary than the materials from Cuts 1 and 2 and merit a special study.

CUT 1 Cut 1 was originally laid out to extend 15 meters east and 7 meters north of R 3 (Fig. 4), thus including a rectangular area of 7 by 15 meters. T h e eastern boundary reached the visible edge of the midden, which at this point slopes to the northeast at an angle of approximately 45 degrees. It was originally planned to have Cut 1 consist of a vertical face on the south (for Cut 2) with a series of terraces stepping down from the north. This stepped arrangement would economize on earth removal. For this reason one terrace (see A on Fig. 4) was left on the north side at a depth of 2.20 meters below the surface at the northeast corner of the cut. However, the original plan for Cut 1 had to be modified when the great depth and extent of the deposits became obvious. T o facilitate the removal of earth and expose the underlying deposits which extended eastward beyond the limits of the Inca midden, the length of the cut was extended 12 meters, giving a total

east-to-west length of 27 meters. T h e final 3 meters were solely to facilitate passage of wheelbarrows and are not included in the profile here given (Fig. 5). During the process of removal, the north and south walls were allowed to converge gradually, thus lessening the amount of digging necessary prior to making stratigraphic Cut 2 (Fig. 4). When completed, Cut 1 had a width of 7 meters on the west and 1.60 meters on the east. Its depth at R 3 (Fig. 4) was 10 meters, and at R 4, 2.60 meters (Plate 2b and Figs. 4, 5, 6). Excavation was commenced in the southwest corner of the cut (R 3 on Fig. 4) and this point (R 3) was used as datum for establishing vertical and horizontal positions of burials, special finds, and other features. All potsherds were saved, but their exact levels were not recorded. Starting in the reference corner, the midden material was peeled off and thrown down the slope, by necessity being rehandled many times. Thus a sufficient

EXCAVATIONS A T TEMPLE OF THE SUN

44

collection was secured and few artifacts of importance were missed. Due to the nature of the material and the time factor involved, screening was not attempted. Seven burials were encountered in Cut 1. T h e majority of these were near the east end of the cut. T w o were textile-wrapped mummies pertaining to the upper or Inca layer. T h e others were without textiles and were in the lower earth strata. None was accompanied by complete

pottery vessels. These and all other features encountered, such as walls or floors, were cleaned, measured, and photographed. They were then removed in order to clear entirely the face of our next stratigraphic cut, Cut 2 (R 3 - R 4 on Fig. 4). When completed, Cut 1 had been carried down to sterile sand throughout its entire length (Fig. 5). We will discuss the nature of the soil deposits and cultural features of this cut in relation to Cut 2.

CUT 2 T h e profile at the juncture of Cuts 1 and 2 (Fig. 5) shows the entire north face of Cut 2 and the manner in which it was divided into stratigraphic blocks. As previously indicated, Cut 2 had a width of one meter. Ideally, each stratigraphic block was one meter by one meter in length and breadth and half a meter in depth. Factors of slope necessitated the use of irregular blocks, especially near the surface. All cultural, as well as significant animal and vegetal, material from each block was saved. T h e blocks were marked off on the north face of Cut 2 (Fig. 5) by vertical and horizontal strings. In removing these blocks, burlap cloths were spread on the floor of Cut 1 below the block being removed. By selecting those far enough apart, it was possible to remove two blocks at a time. One workman removed each block while another helped him watch for and sack all specimens. T h e dirt from each square was thrown below onto the sacking, and it was looked over again when it was shoveled from the sacking into the wheelbarrow. T h e only obvious chance of mixture of materials occurred when falling earth from upper squares brushed the walls below and dislodged sherds which would fall upon the sacking. T h e entire process of removing, sacking, and labeling the material from the stratigraphic blocks was carefully supervised STRATA

AND

STRUCTURAL

by us at all times, and very few occurrences of this sort were noted. Like the occasional intermixture of sherds from the borders of immediately adjoining blocks, such slight intrusions seem statistically negligible. At the eastern end of the excavation, Cut 2 just impinges on the tomb encountered in Test Pit 1 (see Fig. 5, rock wall). A wall made of rough undressed rocks occurs just above the top adobe layer at this point. This marks the eastern border of the concentrated burial area which encroaches slightly on the southern half of Cut 2 in stratigraphic Blocks 150, 165, and 166 and the uppermost portions of Blocks 181 and 182 (Fig. 5). In these blocks, the badly mixed bones of some six individuals were encountered. T w o other burials were encountered in the lower eastern section of Cut 2, an adolescent in Blocks 205 and 206 and a child in Block 203. In all of Cut 2 only three other burials occurred, a child burial in Block 76, another in Block 85, and an adolescent in Block 103. These were below the chala or Inca level. As indicated in the diagram (Fig. 5), none of these burials showed on the face of the cut. With the possible exception of the blocks impinging on the burial area of Test Pit 1, these few and scattered burials seem to have had little disturbing effect on ceramic or other stratification in Cut 2.

SUPERIMPOSITION,

T h e superimposition of soil layers, detritus, and artificial structures in Cuts 1 and 2 is most clearly perceived on the north profile of Cut 2 and the west profiles of both cuts

CUTS

1 AND

2

(Figs. 5, 6 and Plate 3a, 3b). If these strata are considered from the oldest (bottom) level to the top, an idea can be gained as to the manner in which the occupied levels were

EXCAVATIONS A T TEMPLE OF THE SUN built up. T h e earliest occupation in this portion of Pachacamac occurred on the surface of a bed of coarse brown sand which underlies all of Cuts l and 2 and the test pits outside. Below the lowest ash layers, this sand is clean and absolutely without cultural admixture. None of our pits penetrated through it, and the thickness of the bed is unknown. As can be seen from the profile (Fig. 5), it has a gradual slope from west to east, in contrast to the 45-degree slope of the present surface. T h e upper portion of this sand is marked by superimposed layers of ash which slant all the way down the gradual slope. These vary in thickness from place to place but average around five or six centimeters. T h e y apparently represent the earliest occupation on the slope of the natural hill now crowned by the Sun T e m p l e . A b o v e these ash layers, or lenses, which characterize the lower one or two meters of the deposit is a series of packed clay or adobe layers, each slightly below the preceding one as they extend eastward down the slope. T h e y suggest puddled clay or adobe and probably represent habitation levels. In the southwest corner of the cut (Fig. 5, lower right) there was an uneven but definite adobe layer or irregular floor which was continuous around all three sides of the cut and underlay a notable adobe-block structure or old retaining wall at that point. Depressions with concentrated ash and charcoal suggested fire pits in this floor. A t one place (Fig. 5, diagonal wall), a crudely built wall, of undressed blocks of stone, cuts through one of these adobe layers and three of the uppermost ash lenses. T h i s low wall runs in a true north-tosouth direction and crosses our cuts at an oblique angle. Its purpose was obscure. M i x e d with and above these adobe layers are deposits of mixed ash, dirt, rubble, and small rocks (Fig. 5). Sherds and limited cultural detritus occur in all layers above the sterile sand. T h e adobita or hand-molded adobe-block structure which rests on one of the largest floor or habitation levels is particularly interesting. It is apparently both a platform and a retaining wall supporting an artificial terrace on the western or uphill side (Fig. 5 and Plate 3b). It is composed of layers of small

45

adobes or adobitas set in a clay plaster which is harder than the adobes. T h e plaster layers average about 4 centimeters in thickness. T h e adobita illustrated (Plate 4i) is typical and, like most of the others, retains the finger impressions of its molder. It measures 30 by 20 by 15 centimeters. T h e entire adobita structure has the same profile on both sides of our cut and, though we did not follow it beyond these limits, was apparently a rather extensive structure. Like the rock wall previously mentioned, the direction of this adobita retaining wall is from north to south and it crosses our cut at a slightly oblique angle. T h e space inside, to the west of the retaining wall, is filled with dust, broken rock, gravel, and unstratified detritus (Fig. 5). It strongly suggests an artificial fill probably made at one time. T h i s is overlain by a layer of dark, humus soil (see Fig. 5, vegetal layer) and is capped at the top by a low rock wall and a somewhat scattered flooring of adobes (Figs. 5, 6). Thus, the retaining wall itself forms one adobe platform below and supports a fill capped by another adobita floor or platform above. T h e occurrence of vegetal mold on this ancient buried slope indicates a considerable period of exposure. H o w the early terrace-builders watered this slope remains a puzzle, for no vegetation can grow there now (Plates 1, 2). It is tempting to see in this adobita structure the terraced approach to the earliest temple on the hill now crowned by the Sun T e m p l e . However, the true extent and function of this interesting early structure is only hinted at by our deep but limited excavations. A b o v e this adobita structure and the adobe layers and ash lenses which slope down from it to the east, we find a second or intermediate series of strata underlying the chala or Inca midden material (Fig. 5). Whereas the strata below the adobita structure slope gently from west to east (Fig. 5) and, in Cuts 1 and 2, not at all from north to south (Fig. 6), the intermediate series above all slope steeply from the west (Fig. 5) and, in different strata, from both north and south as well (Fig. 6). T h e nature as well as the degree of slope of the deposits also changes above this line. Instead of gradually accumulated occupation layers,

CUT 1

debris

GR A VEL

L A Y ER

LOOSE DIRT - LARGE ROCKS \ D f R T AND GRAVEL

DUST - DETRITUS 4

and

GRAVEL ROCK

HABITATION FINE

R0CK

LEVEL

' .DETR1XÜS

SAND

SCÀL e —

I

FIG. 6. CROSS SECTION OF SOIL AND CULTURAL STRATA, WEST END OF CUTS 1 AND 2

EXCAVATIONS A T TEMPLE OF THE SUN the intermediate strata are composed of loose dirt, gravel, rubble, larger rocks, and detritus. Unlike the gradually stratified habitation levels below, or the purposely concentrated "'haystack" (Plate 3a) Inca midden above, these intermediate layers appear to be the result of dumping or sliding of materials dislodged by construction or other human activities from the western hill slope above. Judging from the limited cross section revealed on the west end of Cuts 1 and 2 (Fig. 6), the intermediate levels slope down from the north (Fig. 6) as well as from the west. T h i s suggests that dumping or sliding came from the northwest at this time. A b o v e the thick "gravel-dirtdebris" layer (Fig. 6) there is a capping of yellow dirt, in its upper layers mixed with the Inca chala. O n these mixed deposits the Inca midden was built u p until its highest point extended slightly to the south of the area chosen for our cuts (Plates 2b, 3a). Considering the immediate proximity of the great Inca Sun T e m p l e on the higher ground to the west, it is quite possible that the intermediate strata accumulated on the area of Cuts 1 and 2 as a result of building activities connected with this great structure or one of its forerunners. Whereas the lower and intermediate strata represent only occupational or casual accumulations of human detritus, the upper stratum, the Inca midden, is a highly concentrated refuse deposit (Plates 2b, 3a). W i t h the exception of several dumps or lenses of intrusive yellow dirt (Figs. 5, 6) this uppermost stratum is a haystacklike mass of vegetal remains, worn-out or broken artifacts, and bones mixed with rocks, dirt, and fine dust. T h e vegetal remains, particularly reeds and cornstalks, were so thick that our profiles had literally to be trimmed as well as cut. W e have referred to this vegetal mass as chala or cornstalks (Fig. 5) although tortora or reeds, apparently old roofing material, are even more abundant. Due to the extreme dryness of this deposit, the preservation of usually perishable materials is amazing—as is the dust when one is working. T h e mass of textiles from which we selected representative samples was enormous. Both cotton and wool were represented. W h i l e some well-decorated textile

47

pieces occurred, the majority had simple stripes or were plain. Some were very heavily patched and repatched, suggesting that poverty was not lacking under the Inca regime, even in the great ceremonial centers. A n enormous amount of cordage, sandals (leather and fiber), basketry, matting, raw wool and cotton, spindles and whorls, needles (some threaded), combs, and woodwork was intermingled with masses of both well-decorated and plain broken pottery. A wide range and large numbers of dried plants and fruits occurred, as well as many broken and sometimes burned animal, bird, fish, and molluscan remains. Stonework was crude and uncommon, the main types being flat metates, manos, hammerstones, and large crudely chipped scrapers. Only one fragment of metal was found, presumably a copper pin or awl. Burials were not encountered in this stratum in C u t 2. As previously indicated, two mummies were found in this layer in C u t 1 near the eastern edge of the midden. Detailed studies of the above cultural and organic remains have been inaugurated but are not included in the present account. T h e above materials suggest a concentrated rubbish dump accumulated by a considerable population living in the immediate vicinity. Whereas much of this detritus might pertain to any predominantly Inca population, certain ceramic styles represented by fairly abundant sherds indicate that local peoples, or at least coastal wares, contributed to the chala accumulation. In regard to the Inca style pottery, there seems to be a high proportion of finely modeled and painted polychrome ware, particularly aryballoid fragments (chicha jars?) and handled plates. In addition, the refuse contains large fragments of red-painted plaster and fiber from structure or room walls; much tortora or reed roofing and floor material; numerous large worn-out fiber or bird-wing brushes or brooms; abundant llama skulls (often broken open or burned) and bones; many large fragments of llama, alpaca, and other skin; desiccated birds and much bird plumage; and a wide variety of strange molluscan and organic remains. T h i s combination could be interpreted as indicating that an aristocratic Inca group,

48

EXCAVATIONS A T TEMPLE OF THE SUN

probably the priesthood, lived on the near-by terraces of the Sun Temple with their servants, some of whom were local or coastal people. T h e structural materials, painted plaster, and brooms suggest the above location, for this refuse was dumped just outside the main northeastern entrance. Both subsistence and sacrificial patterns may be represented by the strangely variegated bird, animal, and other organic remains. T h e raw cotton and wool, with the spinning equipment, reveal some manufacture, and evidence indicates that much culinary activity was carried on. As we have already suggested (p. 37), Roman in the sixteenth century, presumably on the basis of traditional accounts, postu-

lated exactly this state of affairs. Our own excavations, like Uhle's observations regarding the lower terraces of the Sun Temple, seem to give considerable support to this early account. It is also interesting that in our entire excavation in this midden we found absolutely no evidence of Spanish contact or of the colonial period. Apparently this refuse heap was not in use at the time of the Spanish occupation. In the following section we will discuss the various ceramic styles and types encountered in Cuts 1 and 2, and in a later section we will consider the distribution of these styles and types in relation to the various strata here described.

Ceramic Styles and Types

T

is concerned mainly with the classification of the sherd material from C u t 1 into styles and types. Sherds were abundant in this cut, but because of our method of excavation only a large representative sample were preserved. O u r counts indicate that a minimum of 22,352 sherds, including 4,483 (20 percent) decorated and 17,869 (80 percent) plain, were utilized from C u t 1. O n the basis of these, a tentative classification was established in the field. Later this classification was added to and revised (Table 2). C u t 2, from which every sherd was preserved, yielded 24,038 sherds, of which 4,119 (17 percent) were decorated and 19,919 (83 percent) plain. In all, 46,390 sherds from both Cuts 1 and 2 were available. Of these, 37,788 (82 percent) were plain and 8,602 (18 percent) were decorated. Where types were inadequately represented in our collection from C u t 1, examples from C u t 2 are also included in the description and text figures (Table 3). HIS SECTION

T h e present classification is based on external and easily observable features, mainly surface treatment and decoration. Plain ware classifications were made, but time has been lacking to employ these here. As a result, this abundant type of ware (some four fifths of the collection) has been thrown into categories so large as to be almost meaningless except to show its great preponderance. Such studies can be completed later when finer distinctions in this great mass of undecorated ware become essential. Similarly, more detailed analysis of all styles and types by microscopic and other means should follow. A t present, however, it is necessary to confine ourselves to determining objectively the relationships and sequential trends of the more obvious decorative styles in Pachacamac pottery. In this section, classification is stressed. In the next section the matter of sequence of

styles and, to a lesser extent, of types will be considered. A s employed here, the classificatory term style represents a ceramic complex based upon an aggregate of similar decorative and form features which occur in association so consistently as to imply a cultural and temporal unity. Decorative and surface treatment are matters of primary consideration. T h e criterion of form is considered wherever it can be assumed with reasonable validity, primarily from the pottery fragments themselves, or, secondarily, from similar vessels illustrated elsewhere. Due to the almost entirely fragmentary nature of the ceramic materials considered in the present study, the criterion of form has much less value here than in cases where complete vessels from burials are considered. Paste, composition, and technique are also mentioned but are not emphasized as much as surface treatment because of the apparently limited possibilities of detecting more subtle changes of this sort through macroscopic examination. Where marked temper or paste changes occur, these are noted. T h e term type is used to indicate a smaller category, usually decorative, which forms a unit within the larger style. It is obvious that the number of types could be greatly increased by making finer distinctions, but an attempt has been made here to compromise on a limited number of motifs, pattern arrangements, or other distinguishing features which give some promise of indicating either vertical or horizontal (time or space) developmental trends. Since a number of type criteria may occur on the same fragment, that motif which appears to be central to the design as a whole, so far as may be determined from the fragment, is given precedence. Certain of these criteria tend to appear as markedly dominant or central in a greater majority of cases than others, hence these motifs

5o

CERAMIC STYLES AND TYPES

or arrangements have been given precedence over others which appear to have been more subsidiary in nature. This situation is discussed in relation to the various types described below. T h e following descriptions of the styles and types distinguished in the fragmentary ceramic materials recovered from Cuts 1 and 2 below the Sun Temple at Pachacamac are unavoidably incomplete. However, references are here given to other sources where many of these

stylistic distinctions aie illustrated on whole vessels. Such references, combined with the illustrative material in the present report, should make clear the major styles and types that are later employed in tracing stratigraphie changes in the materials from Cut 2. Since certain styles are better known and are better illustrated in previous reports, their description here may be more cursory than that of others which are not so well known or so well illustrated elsewhere.

INCA S T Y L E INCA P O L Y C H R O M E

STYLE

A well-known style generally based on consideration of ceramic materials from Cuzco or other highland sites presumably pure Inca in origin (compare Kroeber and Strong, 1924a, pp. 9 - 1 6 and Figs. 1-4). T h e style is perhaps best illustrated at present by the Machu Picchu materials, since this appears to be a pure Inca site (Bingham, 1930). In the following discussion we will give references to this well-illustrated work and, for materials of like style from Pachacamac, to the work of Uhle (1903), particularly to the ceramics from the Inca graveyard on the southeastern terrace of the Sun Temple. Inca style vessels are also illustrated by Schmidt (1929, pp. 349-5 2 )> Muelle and Bias (1938, Plates 42-44), and Pardo (1939, Plates 1 - 1 3 ) , among many others. T h e forms of Inca pottery, including both painted and plain wares, are distinctive and well known and can be briefly recapitulated. Bingham gives outlines of most of these forms but employs a number of descriptive terms more familiar to classical than to American archeologists. Here we will refer to Bingham's types (1930, Plates 70-72), employing somewhat different terms. T h e aryballus, or aryballoid, is probably the most distinctive Inca form (Bingham, 1930, Plate 70, Fig. la-d; Uhle, 1903, Plate 18, No. 1; Kroeber and Strong, 1924a, Fig. 1). Other forms include kero-shaped cylindrical vessels (Bingham, 1930, Plate 72, Fig. 15a, 15b; Kroeber and Strong, 1924a, Fig. 2d); footed goElet or beaker forms with horizontal handles, often with lids (Bingham, 1930, Plate

70, Figs. 2a-c, 3a-c; Uhle, 1903, Plate 18, Nos. 4, 5); low, two-handled dishes and bowls, most commonly with horizontal handles (Bingham, 1930, Plate 70, Figs. 4a-e, 5a; Uhle, 1903, Plate 18, No. 3); small pitchers and bottles with vertical or (more rarely) horizontal handles and slightly flaring necks (Bingham, 1930, Plate 72, Fig. i3a-h; Uhle, 1903, Figs. 67-71); and plates with effigy handles (Bingham, 1930, Plate 7 1 , Fig. na-g; Uhle, 1903, Plate 18, No. 6; Kroeber and Strong, 1924a, Fig. 2h) or with small projections on two sides (Bingham, 1930, Plate 72, Fig. i2a-d; Kroeber and Strong, 1924a, Fig. 2i). In addition, flat-bottomed face jugs (Bingham, 1930, Plate 72, Fig. 13b, 13d), openmouthed braziers with solid tripod legs (Bingham, 1930, Plate 72, Fig. 16a) and a number of more widely distributed pot forms (Bingham, 1930, Plate 71, Figs. 7 - 1 0 ) are characteristic Inca forms. T h e very long-necked Inca bottle (Pardo, 1939, Plate 7, Figs, a, b) seems rare or absent on the coast, whereas the barrelor canteen-shaped vessel (Kroeber and Strong, 1924a, Fig. 4f; Muelle and Bias, 1938, IcaChincha style, Plate 45a, 45b) often associated with late or Inca horizons on the coast does not appear to be common in the highlands. With the possible exception of the longnecked bottle, all of the above forms seem to be represented in the Inca style materials from Cuts 1 and 2. In the present analysis the distributions of only a few of these forms are considered, due to certain practical considerations previously mentioned. T h e restraint and precision of the Inca style, as well as its general uniformity in both

e BLACK

LTRED

RED

WHITE

FIG. 7. INCA POLYCHROME STYLE SHERDS: FERN, SERRATED, H A T C H E D ZONE, SERRIED DIAMOND, MALTESE CROSS AND CHECKER TYPES

52

CERAMIC STYLES AND TYPES

form and decoration over the far reaches of the later Inca Empire, have often been stressed. T h e ware is generally hard, well modeled, highly polished, and carefully shaped; and, in the Inca Polychrome style, designs are nearly always applied with great care. Designs tend toward the geometric, but, as analysis of types below indicates, naturalism is also employed. Colors in the Inca Polychrome style at Pachacamac range from two to four, with three-color ware the most common. T h e colors most commonly employed are black, white, and dark red designs on a white, yellow, orange, or light red background. Modeling in the Inca Polychrome style is in most cases limited to aryballoid nubs and effigy-head handles to dishes. It is usually conventional (Fig. 7) but at times is pleasingly naturalistic (Figs. 9b, 9c, gd). In regard to internal structure, the paste in Inca Polychrome ware is very fine, as is the tempering, which is hardly apparent to the naked eye. Further details, such as exact hardness and the ingredients of paste, temper, and slip, must await later microscopic and technological analysis. Such studies should also throw light on the coastal or highland origin of types and individual pieces. T h e Inca Polychrome style includes the following types. j. Fern pattern (Fig. 7 a ) . — T h e Fern type is characterized by a highly stylized fern with central stem and branches. T h e branches consist of two or three parallel lines terminating in round solid dots. T h e design is usually black against an orange or red background, but color variations occur. T h e design is seen commonly on the exterior of aryballoids, in the interior of plates, and elsewhere. Bingham calls this the "necklace pattern" (1930, p. 179). T h e Fern pattern, on account of its great vogue in Inca decoration, has been illustrated many times. (Examples: Bingham, 1930, Plates 75, 76, 80a, 81b, 83, 84, 85, 112, 113, 116, 119c, 127c; Uhle, 1903, Plate 18, No. 7-) 2. Serrated, pattern (Fig. 7b; border: Fig. 7c, 7d).—In the Serrated type the colors, distribution, and type of vessel on which the design occurs are similar to those of the Fern type. T h e characteristic design is a field of parallel

lines with serrate or saw-tooth attachments to the lines. It occurs both as a border and as a central design. (Examples: Bingham, 1930, Plates 74, 80c, 82e, 95; Uhle, 1903, Plate 18, No. 1.) 3. Serried diamond (Fig. 7d).—In the Serried diamond type the characteristic decoration is a series of diamond figures with corners touching. It appears on both interior and exterior surfaces. O n the exterior of larger vessels one series of diamonds is often placed in a white-bordered zone not far below the neck of the vessel. O f t e n the diamonds are hatched or linear rather than solid. W h e n used on plates, diamonds are usually of solid color, red, black, and orange being the most common colors. (Examples: Bingham, 1930, 78a, 8of, 81a, 86g, 95, 108a, 108b, io8f, logd, 117b, 121b). 4. Naturalistic (Fig. 8a, 8 b ) . — T h e Naturalistic type is characterized by semiconventionalized or, occasionally, very realistic life figures including birds of various species, frogs, fish, fruits, seed pods, and worms or other insects. As interior decoration it is very common, particularly inside plates, but it also occurs in exterior use. T w o or three colors, with a lighter or, more rarely, a darker background color in addition, are common. (Examples: Bingham, 1930, Plates l i g f , 122, 123; Pardo, 1939, Plates 2f, 2g, 7a, 7b, IOV, IOX, loy, IOZ, 10a', 10b', 10c'.) 5. Checker pattern (Fig. 7 f ) . — T h e Checker type is a rare pattern, occurring usually on exteriors but occasionally on interiors, characterized by a field of alternate checks of black or red contrasted with checks of white, yellow, or orange. T h e Checker pattern occurs in conjunction with Maltese cross, Hatched zone, Bold geometric, and some other types. In this classification it takes precedence over all of these. (Examples: Bingham, 1930, Plates 82d, io8e; Pardo, 1939, Plates gq, i2d.) 6. Maltese cross (Fig. 7e).—In the Maltese type, a solid or half-solid Maltese cross design of varying size is used against a plain background or against the background of another decoration. It is applied both internally and externally, with considerable variation in form and color. T h i s type may occur in association with Checker, Hatched zone, or B o l d

-Sto^ BLACK

PURP.LE LT HEP

g

KEP

WHITE'

m m is M i r

c». . .

5

FIG. 8. INCA POLYCHROME STYLE SHERDS: NATURALISTIC, SEMINEGATIVE, BOLD GEOMETRIC TYPES

54

CERAMIC STYLES AND TYPES

geometric. In our classification it takes preced- and white on red style (Fig. i2c-f) is predomience over all of these but the first. (Examples: nantly an unpolished ware with paint rather crudely applied. However, there seems to be Bingham, 1930, Plates 73a, 73b, 82a, 100a.) 7. Seminegative (Fig. 8c, 8d).—The Semi- some blending locally of the Bold geometric negative type has a complex decorative area type (Inca Polychrome style) with the Black characterized by a background darker than and white on red style. (Examples: Bingham, the design elements. Commonly three-colored, 1930, Plates 86d, iogf; Uhle, 1903, Plate 18, its background is often painted over an ele- No. 8.) mentary base or slip. This slip sometimes T h e following six are distinguishing feashows through as part of the design, or addi- tures unclassifiable as to major decorative tional colors outlining the design may be . typeused. Angular and rectangular scroll and step Aryballoid nubs (Figs. 8f, 9a, 10b).—The elements are common. T h e type is well illus- single lug or knob characteristic of aryballoid trated by Uhle (1903, Plate 18, No. 6). Here jars is very distinctive. This is modeled, inthe major Seminegative design takes preced- cised, or, more rarely, painted. T h e most ence over the isolated Maltese cross. (Other characteristic form suggests a bat, and the examples: Bingham, 1930, Plates 78, 79, 109b, majority of more conventionalized forms seem 127b.) to be derived from the same model. (Ex8. Hatched zone (Figs. 7c, ga-shoulder).— amples: Bingham, 1930, Plates 78-88; Uhle, T h e Hatched zone pattern is interior and ex- 1903, Plate 18, No. 1.) terior decoration of considerable variation. Aryballoid ring necks (Fig. 9a).—This charZones of black and red lines forming wide- acteristic form with flaring lips and two small spaced hachures and cross hachures on a white vertical lugs may occur with any of the above background are common. Large zones of black decorations. It is usually dark red with thin, on orange with hachures and cross hachures widely spaced white lines, although there is used as decorative motifs occur commonly on some color and size variation. Where not the upper portion of vessel exteriors. Fine, enough of the body sherd shows for typing, close-spaced cross hachures in rectangles are the sherd is listed as a ring neck. (Examples: used as an interior decoration on plates. Bingham, 1930, Plates 76, 84; Uhle, 1903, Hatched zones occur with Fern, Serrated, Plate 18, No. 1. Note necks.) Serried diamond, and Checker types. In our Effigy plate handles (Figs. 9b, 9c, 10a).— classification, these other types take preced- This is a characteristic form, often naturalistic, ence. (Examples: Bingham, 1930, Plates ggj, depicting ducks, other birds, and possibly alli99k, 100c, io8d, loge, 119a, 124a; as a sub- gators. These become highly conventionalized, sidiary design, Uhle, 1903, Plate 18, No. 1.) and this classification includes blunt-ended p. Bold geometric (Fig. 8e).—The Bold geo- handles bearing no likeness to any animal. metric type shows wide zones of solid color (Examples: Bingham, 1930, Plates 89-101; bordered with another color against a back- Uhle, 1903, Plate 18, No. 6.) ground of a third color. Concentric rectangles Hatched handles (Fig. ge).—These are flat are commonly used. Red, black, orange, white, strap handles with hachure or cross hachure. and purple are common colors. Bold geo- (Examples: Bingham, 1930, Plate 108c, io8e, metric is found in association with Fern, io8f. Note handles.) Serrated, Serried diamond, Checker, Maltese Hatched lips (Fig. gf).—These lips, usually cross, and Hatched zone. In our classification from smaller vessels, have hatched markings these others take precedence unless used for in black or red paint usually on a white backpurely subsidiary design. While Bold geomet- ground. ric, as indicated by its other type associations, Banded borders (Fig. gg).—These are white is predominantly Inca, our classification prob- bands about one centimeter wide which occur ably includes some polished Black and white immediately beneath the lip of ollas or bowls. on red, which is coastal in range. T h e Black T h e bands are usually bordered with black

BUCK

•m

WHITE

RED

Crtii

5

Fie. 9. INCA POLYCHROME AND POLISHED BLACK STYLE SHERDS: ARYBALLOID NECK AND NUB, EFFIGY HANDLES, HATCHED HANDLES, HATCHED LIP, BANDED BORDER

CERAMIC STYLES AND TYPES

56

lines. They are always exterior and commonly occur on vessels of the Serried diamond or Hatched zone types. T h e Banded border type

is counted only where other main designs are not present or not observable due to fragmentation.

INCA-ASSOCIATED S T Y L E S T h e Inca-associated styles include styles occurring with but not all culturally cognate with Inca. Some of these styles are subdivided into types. PLAIN

RED P A I N T I N G

STYLE

(FIG.

lib)

Dark red or purplish red painting over all or part of the area of dull fired surface occurs on many vessels of Inca form. Where the red painting is partial, as inside the lips of aryballoids (Fig. 9a) or on the inner rims of plates, it is highly distinctive. Where complete, it tends to blend with any polished dark red slip and becomes almost valueless as a diagnostic. Plain red painting occurs alone, hence its classification as a style, but it also occurs with all the foregoing Inca Polychrome types. Where observable, the latter take precedence to it in the classification. This is obviously an Inca style. POLISHED B L A C K S T Y L E (FIGS, g d ,

11 a,

HC-h)

This style, while superficially uniform inasmuch as it was reduced in firing whereas almost all other pottery from this site was oxidized, is obviously a combination of at least two black-ware ceramic traditions. One of these is apparently Inca from the highlands, another apparently Late Chimu from the north coast. It is interesting to note that Late Chimu forms in black ware are also associated with Inca wares in the highlands as well as on the coast (at Machu Picchu; see Bingham, 1930, Plates 111, 115a, n8f). T h e majority of Polished black ware from Cuts 1 and 2 at Pachacamac consists of sherds from vessels of Inca form. These are very highly polished and lustrous and run the gamut of the forms previously described as typically Inca. In addition, there is a much smaller number of sherds from stirrup-handled and other vessels of Late Chimu type (compare Uhle, 1903, Fig. 78 and Plate 18, No. 11). In the present sherd material it is often impossible to distinguish between these styles, and the attempt has not

been made here. Even where complete Polished black vessels are present, the styles often appear too closely blended to make very consistent distinctions (Uhle, 1903, Plate 13, Nos. 2, 3, 8, 12, 14). T h e Polished black style includes the following types. 1. Plain (Fig. 11a).—Pottery of this type is plain black, lustrous and thin, with quite hard polished black exterior, and dark gray in cross section. T h e paste is fine, as is the tempering material. T h e majority of fragments are from small jars and other bowls, low jars, and plates. Strap handles are fairly common. Inca forms seem to predominate. 2. Incised (Fig. 11c, n d , l i e ) . — T h i s type is the same as the above in relation to quality of ware. T h i n line incising, forming geometric and a few curvilinear designs, is common. Small incised motifs include birds, animals, and fish. Occasionally decoration is achieved by fine punctation or stamping (Fig. n e ) between incised designs. (Examples: perhaps best illustrated by Inca and Late Chincha II vessels; see Kroeber and Strong, 1924a, Plate i3a-c and Fig. 5.) 3. Modeled (Figs, gd, 1 if, n g , n h ) . — T h e Modeled type is the same in ware quality. Characteristic high relief decorations include naturalistic, conventionalized, and geometric elements. Stirrup handles (usually with a small figure where the handle joins the spout), pot necks with faces in relief (with modeled earplugs), raised and incised collars, small fillets, and bird-head plate handles all occur. T h e last two are obviously Inca, the stirrup handles seem Late Chimu, and many of the human effigy pots appear to be Late Coastal (Chancay Black-on-white) in general form. Coastal forms are perhaps most common in this type. (Examples: Uhle, 1903, Figs. 67, 69-72, 74, 76-79; Plate 13, Nos. 1-3, 8, 9, 11-14.) MODELED

BROWN S T Y L E

(FIG.

lOC-e)

This style is similar to the above Polished black Modeled pottery of Inca form, but in

a

BUCK

WHITE

b

RED

FIG. 10. INCA POLYCHROME AND INCA-ASSOCIATED STYLE SHERDS: EFFIGY HANDLE; ARYBALLOID NUB; MODELED BROWN STYLE, RELIEF AND PRESSED RELIEF TYPES; WHITE-ON-RED STYLE, UNPOLISHED TYPE

58

CERAMIC STYLES AND TYPES

this case the modeling techniques are applied to a coarse, dull brown ware. T h e paste is crumbly and has grit temper, and the ware is apparently poorly fired. A red slip is occasionally applied, but the majority of the pieces are apparently unslipped and are rough to the touch. Decoration is primarily by modeling in high relief, various serpentlike designs being most common. Sometimes punctations along these designs are filled with a chalky white paint. White paint occasionally occurs as a band where the neck of the vessel joins the body. T h e majority of these vessels appear to have been small, and they are usually externally smoke-blackened from use in cooking. Small bowls with vertical and horizontal strap handles and direct or low flaring rims are suggested by the fragments. This ceramic style apparently occurs with Inca associations in the highlands (Machu Picchu; see Bingham, 1930, pp. 148-55 and Plates 102-7) anc ^ o n the coast (Pachacamac; see Uhle, 1903, Figs. 84-85 and Plate 13, No. 10). T h e Modeled brown style includes the following types. 1. Relief (Fig. 10c, lod).—The majority of sherds of this type are of the coarse brown ware described above. Designs on these sherds are most often wavy serpentlike ridges undulating horizontally around the body of the vessel. These were sometimes formed by pressure from the inside while the vessel was soft. About half of these have small circular depressions filled with coarse white paint. From Cut 1 were six sherds with a pouting "baby face" in relief (Fig. lod), but, strange to say, none of this type came from Cut 2. Other sherds suggest low-relief figures of obscure outline, and one has a modeled squirrel or monkey peering over a swollen and incised rim (similar to Uhle, 1903, Plate 13, No. 14; Figs. 72, 76). A few rims are collared and fingernailimpressed around the lower edge of the collar. One of these has cross-hatched incisions around the upper body portion of the sherd. A very few sherds with hand-modeled decorations of the above types are plain polished brown ware of a familiar Inca type. We believe that the highland examples illustrated by Bingham (Bingham, 1930, Plates 103-7) a r e mainly of this latter type. They have relief decorations equally characteristic of the more abundant

coarse brown ware from Pachacamac (Fig. 10c; Uhle, 1903, Figs. 84, 85.) 2. Pressed relief (Fig. loe).—This is a highly distinctive type rather sparsely represented in our collection. T h e ware is of the coarse brown type, and several sherds have a thin white line around the neck where the rim joins the body. T h e designs appear to have been impressed on the outer surface below the rim. Within angular lines are attractive little bird designs, dots and circles, and dots (Fig. loe). Uhle (Uhle, 1903, Plate 13, No. 10) illustrates a pot which appears to be of this type. This method of decoration is particularly characteristic of Late Chimu (re Supe, see Kroeber, 1925b, p. 239 and Plate 7og, 70k, 70I, 70m). WHITE-ON-RED

STYLE

(FIGS,

lof,

12a,

12b)

Pottery of this style appears to have been largely utilitarian in purpose. T h e vessels range from abundant large flaring-rim ollas to medium-sized swollen-rim pots. T h e sherds are usually very large in size, and large strap handles are common. Fire blackening is rare, and the large bowls or ollas were probably used for storage of liquids. T h e common features of the types discussed below are size, to a certain extent shape, and the use of white paint as a complete or partial covering over a red body slip. Painted decorations, save for very simple linear designs, are largely lacking. Like so many of the plainer utilitarian wares, these large types are not well illustrated in reports. This style should not be confused with the distinct and much earlier Chancay White-on-red style (Kroeber, 1926b). 1. Unpolished (Fig. lof).—Pottery of this type is thick and coarse. From the fragments available, flaring rims and rounded bodies with heavy strap handles seem to be most common. A few of the rims apparently once had simple face and earplug representations, and the forms as well seem to have been similar to the Black and white on red style described next (see Uhle, 1903, Plate 13, Nos. 5-7). According to the size of the vessel, sherds of this type vary considerably in thickness (five to ten centimeters), but the surface is always rough, the paste crumbling, and the tempering usually coarse grit or gravel. Surface slip

6o

CERAMIC STYLES AND TYPES

is d a r k brick-red in color, a n d d e c o r a t i o n is a c h i e v e d by the a p p l i c a t i o n of very c r u d e l y d r a w n w h i t e lines or w h i t e areas. L i n e s vary greatly in thickness a n d are a p p l i e d i n a most slovenly m a n n e r . D e s i g n areas seem m a i n l y curvilinear, b u t a f e w simple a n g u l a r w h i t e designs occur. T h e w h i t e p a i n t is poorly m a d e a n d is very u n e v e n in strength a n d application. It m a y b e w o r t h y of note, suggesting a s o m e w h a t u n c o n v e n t i o n a l m e t h o d of ceramic classification, that a large sample of this w a r e h a d to b e r e m o v e d f r o m o u r L i m a l a b o r a t o r y to the o p e n air because of its strong a n d unpleasant odor. It seems possible that vessels of this w a r e were used as containers of u r i n e or e x c r e m e n t for either sanitary or u t i l i t a r i a n reasons. A t Pachacamac, the closest affiliations of this U n p o l i s h e d White-on-red ware seems to b e w i t h the B l a c k a n d w h i t e o n red style. (Examples: Fig. l o f ; p r o b a b l y U h l e , 1903, Fig. 66.) 2. Polished (Fig. 12a, l a b ) . — I n size a n d g e n e r a l a p p e a r a n c e this type is similar to the a b o v e , a n d there is some o v e r l a p p i n g b e t w e e n the two. T h e r e are, however, w h a t a p p e a r to b e p o l a r differences. I n general, in the Polished White-on-red type the w a r e is better m a d e , the paste is finer, the grit t e m p e r inclusions are smaller, a n d the w h i t e p a i n t is better a n d covers larger areas. I n fact, m u c h of this w a r e c o u l d be called white-slipped, bei n g externally all white. H o w e v e r , i n other cases the w h i t e covers only certain areas of the o r i g i n a l dark r e d slip. W h i t e slip is m o r e c o m m o n o n exteriors b u t does o c c u r occasionally o n interiors. B o t h the red a n d the w h i t e surfaces are smooth a n d polished, in m a r k e d contrast to the p r e c e d i n g type. Designs are generally l a c k i n g i n P o l i s h e d White-on-red, c o m p l e t e or areal coverage w i t h w h i t e b e i n g characteristic. H o w e v e r , some vessels w i t h p a r t i a l w h i t e slip h a v e t h i n black lines borderi n g the r e d areas, a n d some w i t h w h i t e o u t e r coverage h a v e similar black lines or simple g e o m e t r i c motifs (Fig. 12b). (It has n o t seemed necessary to describe this as a n e w b l a c k a n d w h i t e o n r e d type since the p r e d o m i n a t i n g two-color ware, P o l i s h e d White-on-red, is otherwise identical a n d seems the basic type.)

I n f o r m , this type includes large pots or ollas w i t h flaring necks (often w i t h two small aryb a l l o i d l i k e lugs), large bottle necks, smaller l o w - r i m m e d bowls, straight-walled clay dishes (with h o r i z o n t a l handles), plates, a n d lids. T h e great m a j o r i t y of these are definitely I n c a i n form. T h e r e is some fire b l a c k e n i n g , a n d the type appears to h a v e b e e n largely utilitarian. T h e affiliations of the type are obviously w i t h Inca. (Examples: M a c h u P i c c h u , B i n g h a m , 1930, Plates 86h, 118a, n g d , 119c; P a r d o , 1939. W h i l e a p p a r e n t l y c o m m o n , the type does not seem to be w e l l illustrated.) B L A C K A N D W H I T E O N RED S T Y L E ( F I G . 1 2 C - f )

T h i s style is characterized by a rather coarse, red-slipped w a r e decorated w i t h black a n d w h i t e designs; o f t e n w h i t e designs are outl i n e d w i t h black. It is usually three-color, b u t occasionally f o u r colors are e m p l o y e d , includi n g two shades of red. T h e w a r e is r o u g h to the touch, soft, a n d s o m e w h a t c r u m b l i n g i n texture. I t is o f t e n tempered w i t h a w h i t e grit visible to the n a k e d eye. B o t h the slip a n d the decorative paints are coarse. T h e w h i t e is " l i k e w h i t e w a s h , " as U h l e has described it (Uhle, 1903, p. 35). Designs, a c c o r d i n g to type, r a n g e f r o m geometric to c u r v i l i n e a r a n d seminaturalistic. T h e y are a p p l i e d b o l d l y b u t are o f t e n irregular a n d the effects a c h i e v e d are crude a n d w i t h o u t finesse. M o d e l i n g also is crude, a n d shapes are o f t e n s o m e w h a t irregular. A simple h u m a n effigy p o t w i t h slightly m o d e l e d face a n d m o d e l e d or p a i n t e d hands a n d arms is very typical ( U h l e , 1903, Plate 13, Nos. 4 - 7 ; Figs. 81-83). W h i l e the coarse surface of this style is characteristic, a few of these sherds are decorated w i t h designs occurr i n g o n the B o l d geometric type (Inca Polyc h r o m e style). T h e s e Inca designs i n c l u d e the " w e e p i n g e y e " (Fig. 12c; U h l e , 1903, P l a t e 13, N o . 4) a n d a f e w rectilinear motifs. T h e B l a c k a n d w h i t e o n red style is a p p a r e n t l y U h l e ' s latest red, white, a n d b l a c k coastal style of the " t i m e of the I n c a " at P a c h a c a m a c (1903, Plate 13, Nos. 4 - 7 ; Figs. 81-83). ^ is also a p p a r e n t l y related to the C h a n c a y Blackon-white style (Kroeber, 1926b, Plates 80, 81).

FIG. xa. I N C A - A S S O C I A T E D S T Y L E S H E R D S : W H I T E - O N - R E D S T Y L E , TYPE; BLACK AND W H I T E

ON

RED

STYLE

POLISHED

CERAMIC STYLES AND TYPES

62

EARLY PACHACAMAC PACHACAMAC

INTERLOCKING 5 ; FIGS.

STYLE

(PLATES

4,

13-17)

T h e Interlocking style was first clearly recognized by Uhle at Chancay in 1904 (see Kroeber, 1926b p. 295). It has been best described by Kroeber (1926b, pp. 279-90) in analyzing Uhle's collections from that site. T h e style is well known for the Lima valley, and collections of which it comprises a considerable part have been described from Cajamarquilla and Nieveria (R. d'Harcourt, 1922; Gayton, 1927). Its occurrence at ruins within the environs of Lima has been discussed by Kroeber (Kroeber, 1926c, pp. 335— 43). A t Pachacamac it was first illustrated by Uhle (Uhle, 1903, Figs. 26-29) but its distinctness and importance were not recognized by him until after his discoveries at Chancay. Later (Uhle, 1908, p. 356) he encountered sherds of this style below the Sun Temple at Pachacamac. Many similar sherds were also secured by Dr. Albert Giesecke in his restoration work at Pachacamac and are now in the National Museum at Lima. T h e Pachacamac Interlocking style, as described here, is merely a local and possibly temporal variant of the general Interlocking style from Chancay and other sites on the central coast. It is given a site name primarily for purposes of specificity. Our collection from Cuts 1 and 2 at Pachacamac is very fragmentary, but a few complete or restorable pieces from the graves in Pit 1 are illustrated (Plate 4). Due to the fragmentary condition of the ceramic material from Cuts 1 and 2 at Pachacamac, criteria of form are hard to distinguish, but references will be made to whole vessels of similar decorative type from other sites. In our Pachacamac Interlocking materials there is a high proportion of plain and a very small proportion of painted ware (see Table 2, p. 92). This condition has hitherto been somewhat masked by the fact that only complete and decorated vessels, probably associated with burials, have been discussed in previous reports. Compared to the relative abundance of decorated sherds in the Inca Polychrome style, the paucity of decorated pieces in the Inter-

locking style at Pachacamac is striking. (This scarcity of Interlocking may arise through the fact that the design area on many vessels is confined to the neck, rim, or upper part of the body.) T h e Pachacamac Interlocking undecorated ware is monochrome, ranging in color from a dark, almost orange, red through light red and brown to smoked gray or black. T h e decorated ware is characteristically three-color—red, white, and black. Occasionally four colors, including two shades of red or purple, are employed. T h e design patterns may cover the entire vessel or be confined to definite areas, the upper body or rim. T h e background and predominating color is red, ranging from a dull orange to a light red or brown. T h e white runs to yellow, buff, or gray, and the black is sometimes grayish or brownish. T h e black and white paints are often coarse and seem of poor quality, having a marked tendency to fade except on some of the better pieces. As Kroeber states: T h e pattern is therefore not salient, and in many cases is impossible to photograph. Its essential trait is an interlocking of the elements, the

engaging

ones of which are in contrasting colors and between them fill the decorative field. T h e s e elements are fishes, or possibly sometimes snakes, with triangular heads, and bodies bent on themselves and sometimes serrated. . . .

In borders there is an inter-

locking fret which seems to be a reduction of the same fish motive. . . . Perhaps related are triangles that suggest much reduced faces. These interlock, but do not contrast in color. . . . Other elements which are more or less worked into the interlocking scheme are zigzag lines . . . .

and rows of dots. . . .

[Kroeber, 1926b, pp. 280-81.]

As will be seen when the types under the Pachacamac Interlocking style are discussed, these motifs all occur on our materials. However, the extreme conventionalization of the style, combined with the fragmentary nature of the sherd material, makes the use of such major motif designations as "interlocking fish" hazardous if not impossible. For this reason, decorative type designations are given on the basis of geometric elements (see Table 2). These are discussed below.

CERAMIC STYLES AND TYPES It may be noted in passing that the Pachacamac Interlocking sherds with painted decoration seem somewhat poorer and more limited in decoration than those of Uhle's collection from Chancay illustrated by Kroeber (Kroeber, 1926b, Figs. 10-24; Plates 88-90). It should also be noted that, whereas photography often fails to bring out the design in the Interlocking style, drawings tend to overemphasize it. This is especially true where, as in the present case (Figs. 13-17), colors must be symbolically indicated. A combination of drawing and photography gives a more realistic impression (compare Plates 4, 5; Figs. 131 7)T h e outer surface of Pachacamac Interlocking, both plain and decorated, is generally slipped and somewhat polished, especially in the case of the decorated pieces. T h e polish is rarely high, and much of the plain ware is coarse to the touch. Inner surfaces, save in those pieces with inside decoration, are usually unpolished and bricklike to the touch. T h e ware is highly variable in thickness, decorated sherds ranging from three to twelve millimeters in thickness, with some plain ware even thicker. T h e paste of the thinner ware, especially the more polished and the decorated, is well made, with tempering material often not visible to the naked eye. In the thicker pieces a white tempering material, apparently grit, is visible and the paste is coarse and sometimes crumbling. Firing in the thinner ware seems complete, giving a brickred cross section, but in some of the larger pieces it is incomplete, leaving a bluish line in mid-section. T h e majority of the painted vessels of this style appear to have been small to medium in size. However, some of the crudely decorated and many plain pieces indicate quite large vessels, such as those with the burials in Pit 1 (Plate 4a, 4d, 4e). As previously indicated, it is very hard to postulate complete form from the sherd material. However, mediumsized, rounded pots with low or slightly flaring lips seem quite common (compare Plate 4b, 4c, 4f, 4g, 4j). Many of these had either horizontal or vertical strap handles. Open bowls, usually of small size, with flat or rounded bottoms are common. Bottle forms

63

occur (like Kroeber, 1926b, fig. 22; Plate 88e, 88f), and heavy spouts suggest forms from Nieveria (compare Gayton, 1927, Plate 95a, 95b, 95c, ggf). Small tapering spouts and spout-and-bridge fragments are very rare in our collection (Table 2). They occur commonly in the selected funerary pottery of this style from Chancay, Nieveria, and Cajamarquilla (see Kroeber, 1926b; Gayton, 1927; and R. d'Harcourt, 1922) but, if the present case is typical, vessels of this form are either absent or were rarely distributed in the ordinary refuse. Our Pachacamac collection of Interlocking sherds from Cut 2 lacks positive evidence also of the painted cylindrical vessels (Kroeber, 1926b, Figs. 10-12; Plate 88c) which seem characteristic at Chancay. Several were found in the test pits, however, and they probably occur in Cuts 1 and 2 but cannot be surely distinguished in the present collection. From Nieveria, Gayton illustrates all the above forms, as well as many spouted, spoutand-bridge, and modeled pieces. R. d'Harcourt's collection shows even more of the seemingly exotic forms (R. d'Harcourt, 1922). An interesting form of pot rest, found at Pachacamac with the burials in Pit 1 (Plate 4h), is shown by Gayton (1927, Plate 96b) from Nieveria. At Chancay, having mentioned the more usual forms of jars, bowls, and mammiform and double-spouted vessels with typical Interlocking designs, Kroeber adds that there is, in addition, a varied assortment of pieces including double spouts, poorly modeled bird and animal jars, mediocre modeled human figures, crude jars with heads, and so on, which apparently pertain to the same general period (Kroeber, 1926b, p. 282). T h e possible significance of the rather high variability in form of what is probably a rather small proportion of the later Interlocking or Early Lima style pottery will be discussed in a later section. T h e Pachacamac Interlocking style includes the following types. 1. Dot (Fig. 13a, 13b).—This is a rather rare but distinctive design type occurring externally on bodies, rims, and large spouts of vessels. T h e most characteristic dots are dull white on red, but black dots on red also occur (Fig. 13b). Plain dots are used as decorations

a

BUCK

RED

•

PURPLE WHITE

1XRED

Cm.

FIG. 13. PACHACAMAC INTERLOCKING STYLE SHERDS: DOT, CIRCLE AND DOT, CROSSHATCHED, GEOMETRIC, AND INNER GEOMETRIC TYPES

CERAMIC STYLES AND TYPES by themselves or within a black geometric or linear frame. Dot decoration most often occurs on rather well-finished pieces, usually medium in size, but it also occurs on a few sherds from large vessels. (Examples: A t Chancay, black dots occur commonly in connection with what we have called here the T r i a n g l e and dot type; circle and dot designs also occur, but the simple dot design here described is not illustrated by Kroeber [1926b]. A t Nieveria the simple Dot type occurs [Gayton, 1927, Plates 91a, 9if, 93j, 93k, 93I, 94c, 94i, 94I, 95d, g6d] often on handled jugs and modeled spout-and-bridge forms.) 2. Circle and dot (Fig. 13c, 13c, i 3 f ) . — T h i s is probably a slightly more complex variant of the above. It includes small heavy circles of white (or "open dots"), dots within concentric circles, and dots (usually black) within larger black and (or) white circles. T h e circle and dot is sometimes an eye design, as in some of the examples cited below. It too appears to be an external decoration usually on wellfinished sherds from small to medium-sized vessels. Occasionally, however, it occurs on larger, cruder sherds. (Examples: Chancay, Kroeber, 1926b, Figs. 12, 19; Plates 88a, 88b, 88e, 88f, 9oh; Nieveria, Gayton, 1927, Plates 91c, 92c, g2d, 93g, 94c, g6j; Pachacamac, Gayton, 1927, Plate 97c, 97d; Uhle, 1903, Plate 5, No. 8.) j . Cross-hatched (Fig. 13d).—This rather rare external-design type consists of crosshatched white lines, usually framed by black lines against a red background. It sometimes occurs on or near the neck of vessels. T h e lines may be fairly even (Fig. 13d) but are often very crude and uneven. Where the lines are crude, the white paint is usually coarse. (Examples: Chancay, Kroeber, 1926b, Fig. 23; Cajamarquilla, R. d'Harcourt, 1922, Plate 5, No. 5.) 4. Triangle and dot (Fig. i5c-f).—This type is named for two elements which are included in what appear to be among the most important diagnostic designs of the Pachacamac Interlocking style. These designs include simple triangles and dots; serrated border designs with or without dots; interlocking fish or serpent heads (Fig. 15c); conventionalized serrated bodies of the same; conventionalized

65

triangle and dot faces; and probably other motifs which would be more often recognizable on complete vessels than on sherds. Interlocking of angles and dots in contrasting colors is the salient fact here. T h e triangle and dot design is merely a common denominator of these. Extensive conventionalization plus fragmentation makes larger pattern interpretation hazardous in the majority of cases. These designs occur usually on what appear to have been small to medium-sized pots and bottles having a considerable variety of rim forms. T h e colors and methods of employing them have already been discussed in relation to the style as a whole. (Examples: Pachacamac, Uhle, 1903, Fig. 27 and Plate 5, No. 12—assigned by Uhle to the Epigonal style?; Chancay, Kroeber, 1926b, Figs. 10, u , 15, 22 and Plates 88c, 88d, 89a, 8gg; Nieveria, Gayton, 1927, Plates 91c, 93a, 93c, 93^ ggh, 95a, 95c, g6d; Cajamarquilla, R . d'Harcourt, 1922, Plate 7, No. 3.) 5. Inner triangle and dot (Fig. i5g-l).— T h i s type is the same as the above save that the designs are applied on the inner instead of the outer surface of vessels. As is the case in some other types of the Interlocking style, those vessels with inner decoration are better finished and the designs better applied and brighter than where the same motif occurs externally. In the present type, the vessels are, with only a very few exceptions, small open bowls, apparently with rounded bottoms. (Examples: Chancay, Kroeber, 1926b, Fig. 18.) 6. Fret (Fig. i4a-e).—As indicated in the above general description, this is another of the important motifs of the Interlocking style. In addition to true frets, the Fret type includes a variety of designs of general rectilinear interlocking type. It is distinctive but not overly abundant in the Cuts 1 and 2 material. It occurs on a number of forms from small open bowls to rather large pots with slightly flaring rims. It is quite frequently used on both the outside and the inside (see next type) of the rims. In such cases it often accompanies other design patterns, such as the T r i a n g l e and dot type. (Examples: Pachacamac, Uhle, 1903, Figs. 26, 28, and Uhle, 1908, Fig. 16; Chancay, Kroeber, 1926b, Figs. 10-14 [note rims], 16, 20; Plate 88b,

CERAMIC STYLES AND TYPES 88c [note rims]; Nieveria, Gayton, 1927, Plate g i d ; Cajamarquilla, R . d'Harcourt, 1922, Plate 3, No. 5.) 7. Inner fret (Fig. 14H1).—The Inner fret type is similar to the above but is on the inside of bowls or rims. T h e r e are only a few examples of this type (or subtype), and these are not technically superior to the Fret T y p e . (Examples: Chancay, Kroeber, 1926b, Fig. 12 [note rim].) 8. Geometric (Plate 4f, 4j; Fig. lgg-i).— T h i s type includes a variety of designs among which are opposed chevrons, concentric diamonds, stripes, angular and curvilinear areas, and many other such designs not always clear on fragments of vessels. T h e criterion of selection, as opposed to the Linear type discussed below, is the occurrence of curves or angles formed by the contrasting bands of decoration, indicating geometric areal design rather than mere linear treatment. In the majority of cases this type is distinguishable from the T r i a n g l e and dot type. Where indications of both were present the latter was given precedence. T h e majority of vessels in this type appear to have been medium-sized pots with low, slightly flaring rims. Some fragments of very heavy vessels occur, and thickness ranges from the usual five millimeters to an extreme of thirteen millimeters. (Examples: Chancay, Kroeber, 1926b, Figs. 13, 20; Nieveria, Gayton, 1927, Plates g i d , g i e , gja-d, 93g, 931, 94g. 94j. 9 6 d . 9 fi g-) Inner geometric (Fig. l j j , 1 3 k ) . — T h e Inner geometric type, similar to the above, is often represented on the inner surface of small, direct-rimmed bowls; geometric designs also occur on inner surfaces of larger flaringrimmed vessels. If these have no outer designs they are classified here. T h e r e seems to be no ware difference between this and the foregoing type. However, no fragments of thick vessels are included in our Inner geometric sample. Complete examples are not at hand. 10. Linear (Fig. i 4 i ) . — T h i s is something of a catchall type for sherds of Interlocking style which show only parallel lines or bands of opposed colors. In some cases these may actually represent completely linear designs or design areas; in others they are merely fragments of larger Geometric or other types,

67

the sherds being too small to indicate the larger motif. Whenever any motif characteristic of another type was distinguishable, the sherd was counted with that type. A considerable number of sherds of this type represent linear rim and neck designs, both vertical and horizontal. (Examples: Chancay, Kroeber, 1926b, Plate 8ge, 8gf, 8gh; Nieveria, Gayton, 1927, Plates 93g, 93k, [neck only], 94k, g6e, 9 6 g-) 11. Inner linear (Fig. i4j, 14k).—A limited type, this is represented mainly by sherds from small, thin, direct-rimmed bowls. These are decorated inside and below the lips with thin black, white, and red horizontal lines. T h e red is usually that of the inner lip. W h i l e linear designs occur inside the flaring lips of larger pots, these usually have an outer decoration and are classified with one of the above types. T h e bowl form and the thinness of the Inner linear type seem distinctive. Complete examples are not at hand. 12. Black-on-red (Fig. i6a-d).—A limited two-color type, this lacks white and is characterized by linear black designs on dark red. These designs are usually external but occur rarely on inner surfaces. In some cases these sherds may once have had white designs which have eroded, or the sherds may be portions of vessels having white on other areas. However, the bold black lines on dark red backgrounds seem distinctive enough to merit tentative type segregation. T h e s e sherds vary considerably in thickness (five to twelve millimeters), and vessels of various sizes are indicated. Complete examples of this type are not at hand. 1j. White-on-red (Fig. i6e-h).—Of this twocolor type lacking black, the very few examples at hand strongly suggest the White-onred style at Chancay (Kroeber, 1926b; Willey, Part 3, present volume.) Probably this should be considered a style rather than a type, but our sample is so small and limited as to criteria other than color that it seemed best to place it in the smaller category. T h e main form represented in our sample is a small, thin, open, or direct-rimmed red bowl with a faint white decoration inside or outside the rim. Faint, eroded, horizontal white lines, groups of short vertical white lines, concentric

BLACK WHITE

REO

FIG. 15. PACHACAMAC I N T E R L O C K I N G STLYE SHERDS: T R I A N G L E AND D O T , INNER T R I A N G L E A N D D O T TYPES

a

m BUCK WHITE REI^ FIG. 16. PACHACAMAC

INTERLOCKING

STYLE

SHERDS:

WHITE-ON-RED, AND FIGURED TYPES

BLACK-ON-RED,

7°

CERAMIC STYLES AND TYPES

half circles in white, and one example of a white chain design (Fig. i6f) occur as motifs. A few thicker body sherds show larger areas or zones of dull white design and come from bigger vessels. T h e surface of these White-onred sherds is generally smooth but not highly polished. T h e background color runs from light orange to dark red. T h e white paint is dull and faded and flakes off, leaving only faint traces in some examples. (Examples: complete examples from Pachacamac are not at hand; for Chancay, see Kroeber, 1926b, Figs. 5 - 9 and Plates 86, 87; Willey, Whiteon-red style, Part 3, present volume.) i j . Figured (Fig. i6i-o).—Large black-bordered designs of red on white or white on red mark this type. Bird figures are represented in two of the designs (Fig. i6i, i6j), and a rectilinear toothed or ray-backed "monster" design appears in two others (Fig. 16 o). Other sherds have curvilinear angles, and some of these suggest conventionalized fish (Fig. 16m). T h e r e is a bold sweep to these designs, but the sherds are too small to permit complete visualization. One large sherd has a black design edged with white on red. Small to medium-sized vessels are suggested, and they range from four to ten millimeters in thickness. T h e thinner sherds (Fig. i6i-k) are smooth and fairly well polished; the thicker ones are more granular in surface finish. Aside from open bowls, forms are not distinguishable. T h e r e is a definite suggestion of the Nazca Y style in certain pieces of the Figured type of the Pachacamac Interlocking style (compare Gay ton and Kroeber, 1927, Plates 15-17). It occurs with the Interlocking style at Chancay (Kroeber, 1926b, Plates 88b, 88e, 88f, god), at Nieveria and Pachacamac (Gayton, 1927, Plates 91b, 97b), and at Cajamarquilla and Nieveria (d'Harcourt, 1922, Plate 2, Nos. 2, 3). J5. Modeled (Plate 5 ) . — T h i s is a distinctive and unusual type with heavy white or black designs on red, or black and red on dull white, which is particularly characterized by crude, heavy modeling. H u m a n forms are usually conventionalized (Plate 5H1), but one red fragment of a nose is well modeled (Plate 5c). T h e head spout and the squat body with arm in relief (Plate 5g, 5I1) are badly eroded

but appear quite similar to complete vessels from Middle A n c o n I (Strong, 1925, Plate 46, particularly n, q). T h e llama head here illustrated (Plate 5i) is crude but graphic. Unlike the other pieces, it is solid. Particularly characteristic of this type is a strange, heavy beak or nose (Plate 5a-d) of which we have ten examples. These are painted black with white and red design areas on the side. In four cases the nose or beak is white. Usually there is a cleft under the tip of the raised or " b e a k " portion suggesting a cleft bill or nostril. T h e crude human head (Plate 5f) has two holes for nostrils. These strange forms suggest the grossly modeled bird-headed pots from the Middle A n c o n I (Strong, 1925, Plate 460) and Middle Supe (Kroeber, 1925b, Plate 74a-f) periods. Judging from the sherds at hand, the Pachacamac examples are even cruder. It is possible that they are anthropomorphic rather than avian, for the present fragments defy specific identification. In addition to the above, this Modeled type includes a considerable variety of odd shapes and forms vaguely realistic (Plate 5k, 5I) but unidentifiable from the material at hand. Heavy handles (Plate 5I) occur rarely. T h e r e are two perforated vertical lugs, and a number of other nubs or projections may have served similar purposes. T w o fragments suggest crude spouts, one of which may have had a bridge. However, both seem to have come to a rounded (closed?) top and may not be spouts. One is completely eroded, the other has coarse black and red stripes and dots on a coarse white background. Complete forms are not indicated, and one heavy conical bottom and two conical fragments with side projections defy analysis. T h e y may be portions of solid tripod legs. W i t h the exception of the well-modeled nose (Plate 5c) which is thin (two millimeters) and has traces of polish, the other sherds tend to be thick and heavy (ranging from four to ten millimeters). T h e well-modeled nose fragment mentioned above suggests Early C h i m u origin or influence. T h e bulk of the material of this type, however, is remarkably crude and simple. T h e designs are generally in the cruder Interlocking style, but there seem to be somewhat larger areas of coarse white paint

CERAMIC STYLES AND TYPES than is generally characteristic. Complete vessels of this type from Pachacamac are not at hand. Specimens illustrated from Nieveria (Gayton, 1927, Plate 95g-m) and Chancay (Kroeber, 1926b, Plate goa-c, goe-h) may represent this general type. T h e material just described from Pachacamac, however, is generally cruder. 16. Miscellaneous.—This category is merely a catchall for sherds, apparently of the Pachacamac Interlocking style, which are too fragmentary or eroded to classify into types. Features of the Pachacamac Interlocking style are as follows: Flaring necks ( T a b l e 2; Fig. i7a-g).—Necks of this sort range in height from three to eight and a half centimeters. T h e taller examples flare slightly, while the shorter examples are often widely flaring. Due to conformation of design and shape, the type is sometimes hard to determine, which fact led to the present form segregation. However, Dot, T r i a n g l e and dot, Geometric, and Linear types are all represented on this kind of neck (Fig. 17). These should have been included in the count (Table 2) of these types, but time was lacking to make this change. In general, bottles or lownecked pots are represented. T h e majority of these neck fragments are from well-made, polished, and relatively thin vessels (five to eight millimeters thick). As indicated (Fig. 17), they are usually well decorated. Complete examples from Pachacamac, Nieveria, Cajamarquilla, and Chancay are not at hand. Kroeber (ig26b, Fig. 22) illustrates a neck fragment of this type from Chancay. Handles (Table 2; Fig. 17I1, 17!).—Handles are not abundant in the Pachacamac Interlocking style collection. T h e majority of those present appear to be rather heavy vertical strap handles ranging in width from two to four centimeters. T h e broadest example has white line decoration where it joins the body and may be of the White-on-red type. T h e majority are plain red or are too eroded to determine the type. One excellent-quality handle of the Geometric type is illustrated (Fig. i7h), as is an unusual forked piece with angular red, white, and black designs (Fig. 1 yi). T h i s last may not be a handle. It is probably significant that the long, round han-

7i

dles connecting a cylindrical or contracting neck with the upper body, which occur at Pachacamac (Gayton, 1927, Plate 97c, 97c, 97f), at Nieveria (Gayton, ig27, Plate 94a, g4c, 94j-l), and at Chancay (Kroeber, 1926b, Plates 89c, god), do not occur in our Pachacamac Interlocking collection. Similarly, the thin, ribbonlike bridge between spouts also found at all the above places (see examples on each of the plates referred to above) is lacking in our Pachacamac Interlocking collection. O n e fragment of a plain brown or eroded spout and bridge occurred in C u t 2 (Block 146) and two came from C u t 1, but this type is rare, and other definite bridge fragments are absent. Spouts (Table 2; Fig. i7j-m).—As stated above, spout and bridge fragments are represented by only one sherd from C u t 2 and two from C u t 1. A l l three are plain brown or eroded. T h e r e are a number of tapering spouts or necks present, however. T h e majority of these are eroded but Dot, T r i a n g l e and dot, and Geometric decorative types are represented (Fig. i7j-m). These appear to have come from vessels with two widely divergent, nonsymmetrical spouts emerging from the wall rather than from the top of the vessel (see examples from Nieveria, Gay ton, ig27, Plate 95a, 95c, gsf). Some of these (Fig. 17jm) may have been single tapering necks (compare Gayton, 1927, Plate 94a, 94j-l) but none shows any handle or bridge attachment, and the majority, judging by the angle of their attachment, appear to have been heavy side spouts without handles or bridges. So far as the present collection is concerned, this seems to be the most common spout form represented in our phase of the Pachacamac Interlocking style. Best examples are from Nieveria (see Gayton, 1927, Plate 95a, 95c, 95^ d'Harcourt, ig22, Plate 7, No. 2, figures only one Circle and dot type vessel of this general form). PACHACAMAC

NEGATIVE S T Y L E AND FIG.

(PLATE

6

l8)

T o the best of our knowledge, this distinctive style has not been previously described or illustrated. It was apparently first encountered by Dr. Giesecke in his restoration work at Pachacamac. These sherds are now in the collections of the National Museum at Lima.

BLACK

WHITE

RED

H

O

•

Cm

5

FIG. 17. PACHACAMAC INTERLOCKING STYLE SHERDS: FLARING NECKS, HANDLES AND SPOUTS; DOT, T R I A N G L E AND DOT, GEOMETRIC, AND LINEAR TYPES

CERAMIC STYLES AND TYPES In our excavations, Pachacamac Negative style sherds occurred rarely in the intermediate and lower portions of Cuts 1 and 2 and in the test pits. Its rarity is indicated by the fact that only thirty-four sherds of the style were encountered in Cut 2 (Table 2). One fragmentary but restorable bowl was encountered in Pit 1 (Plate 6). W e know of no complete vessels in this style. T h e outstanding characteristic of the Pachacamac Negative style is the occurrence of bold negative designs on the inner surface of open, direct-rimmed bowls (Plate 6 and Fig. 18). Only two sherds in the entire collection show negative designs on the outer surface. T h e body color of the sherds is a light red or, occasionally, dull brown. Some of the sherds show smoke-blackening on the basal and lower portions, but the majority do not. T h e paste is fine, and tempering is not distinguishable to the naked eye. A medium thickness characterizes the sherds, and variation in this regard is not extreme (four to six millimeters). Basal portions are thickest, lips thinnest. Firing is generally complete, but a few sherds show a blue median cross section from inadequate firing. A l l appear to have received a complete red slip and partial polishing or smoothing, particularly on the inner surface. Design areas on the inside were then covered with wax or a similar material, and the interior was painted with black. O n firing, the black paint and guarding material burned off the inner design areas, leaving them a light red or orange. T h e black inner background color ranges from a very dull to a very positive shade. T h e drawings (Fig. 18) stress the latter, since exact shading is impossible. However, they give the correct effect, which is one of striking light designs against a wholly black background. T h e fading of this black appears to be due to variable erosion through time rather than to any original differences in the black tones employed. In regard to form, there are no indications of any other than an open, direct-rimmed, flat-bottomed bowl (Plate 6). These bowls were apparently small. T h e restorable piece (Plate 6) has a diameter of approximately 16.5 centimeters and a height of 5.5 centimeters. Judging from the other fragments, this appears to be reasonably typical.

73

Some of the bowls may have been roundbottomed, and totally different forms may have occurred, but there are no evidences of the latter in our limited sample. T h e r e are no examples of handles, spouts, or feet pertaining to this style in our collections. As indicated by the type designations (Table 2), the designs employed appear to be mainly geometric. However, the majority of sherds are so small that only a very inadequate idea of total design is obtainable. T h e most common decorative type in our collection is the Inner line and dot (Fig. 18a, 18b, i8e, i8f). W h i l e the dots are reasonably rounded, the lines are variable and irregular. Scrolls, curves, and parallel lines are indicated, but what their total configuration might be is rather obscure. T h e Inner triangle and dot design, best illustrated on the restorable bowl fragment (Plate 6), is more definite and suggests the T r i a n g l e and dot type of the Interlocking style (compare Fig. 15). T h i s resemblance is further enforced by the occurrence in both styles of triangles containing three dots, suggesting a face (compare Fig. i8j and Kroeber, 1926b, Plates 88d, 8gg). T h e Inner scroll type of the Pachacamac Negative style (Fig. i8e), like various miscellaneous designs, is more obscure. A t Pachacamac, the Negative and Interlocking styles have some stylistic elements in common, but they are generally distinctive in form and technique. A n attempt to relate the Pachacamac Negative style to other Peruvian negative painting styles is difficult but opens up some interesting problems. T h e ornate Recuay style from the Callejón de Huaylas is particularly characterized by negative painting (Schmidt, 1929, illustrations pp. 231-42). These designs are often complex and occur in association with elaborate modeling. T h e Pachacamac Negative style is simple in form and decoration but does appear to have stylistic connections with the Interlocking style. T h e stylistic relationship of the Recuay negative and three-color styles to the Interlocking is much more striking and specific. T h e interlocking fish designs on some Recuay vessels are identical with those from Chancay and Pachacamac (see Schmidt, illustrations pp. 232, 233, 241). T h e r e is obviously a strong relationship between

FIG. 18. PACHACAMAC NEGATIVE STYLE SHERDS: INNER LINE AND DOT, INNER DOT, INNER TRIANGLE AND DOT, INNER SCROLL TYPES

CERAMIC STYLES AND TYPES the more elaborate Recuay and Interlocking painted styles and very probably some relationship between the simpler Recuay negative and the Pachacamac Negative under discussion. T h e Recuay styles have not as yet been stratigraphically placed. Bennett has described a negative style from Gallinazo in the lower Viru valley (Bennett, 1939, Fig. 12) which is likewise related to Recuay. In many ways the simpler Gallinazo negative pieces (Bennett, 1939, Negative A, p. 73 and Fig. 12) resemble the Pachacamac Negative sherds in the emphasis on lines and dots and the irregular nature of much of the line work. These resemblances are hardly specific, but they are suggestive. O n the basis of a few underlying sherds which he considers Epigonal, Bennett (1939, Fig. nr-t) regards the Gallinazo Negative as partly contemporaneous with Epigonal. Personally, the present senior author doubts that these particular sherds are Epigonal, since the most distinctive of them (Fig. n t ) suggests to him a style allied to the Interlocking rather than to the Epigonal. T h i s particular problem, like that of the original spread of the interlocking fish design from a southern center such as Nazca, as Uhle suggests, or from Recuay and a " M a n t a r o " center, as T e l l o believes, must be considered later. A l l that is intended here is to indicate that there are some evidences, both direct and indirect, of a cultural connection between the negative painting technique at Pachacamac and that in the Callejón de Huaylas and on the north coast. T h e Pachacamac Negative style must be much better known than at present before its exact relationships can be determined. T h e Pachacamac Negative style includes the following types. 1. Inner dot (Fig. 18c, i 8 d ) . — T h i s appears to be the simplest design type in the Pachacamac Negative style. Because of the small size of sherds, it is not certain whether there

75

is any regularity in the placing of these dots. 2. Inner line and dot (Fig. 18a, 18b, i8e, i 8 f ) . — T h i s type is characterized by round dots and irregular paired, curved and branching lines. A few of these are so irregular that the term "line" seems inaccurate (Fig. 18a, 18b). Here, again, total design configurations are obscure. 5. Inner triangle and dot (Plate 6 and Fig. i8g-k).—This is the most distinctive design element in this rather amorphous decorative style. It includes paired elements and interlocking triangles enclosing dots (Plate 6), as well as triangles enclosing three dots suggesting extremely conventionalized faces (Fig. i8j). Other designs utilize the same elements but are obscure as to pattern. 4. Inner scroll (Fig. 18m).—This design element is sometimes a scroll and sometimes an interlocking scroll, and our classification probably also incorporates fragmentary concentric circles and other curvilinear motifs. It occurs with line and dot (Fig. i8e) and with dots alone. A number of sherds with this type of design suggest that the entire inner surface of small bowls was filled with scroll or at least curvilinear designs. T h i s type is inadequately illustrated here and might prove to be an important decorative unit for the style. 5. Miscellaneous (Fig. 18I).—This group includes a very few sherds decorated by negative painting in black over red slip which do not fit the above simple design categories. T w o sherds, the only ones in the collection, have curvilinear designs on their outer surface. One is very dull, the design suggesting two concentric circles edge to edge. T h e interior of this sherd is unslipped or eroded. T h e other has a bright, yellow-brown background, with curving line design, against a rather shiny black. These two sherds indicate that exterior decoration by negative painting was not unknown in the Pachacamac Negative style.

VARIOUS STYLES A N D TYPES Unclassified material.—This category ineludes a number of unusual sherds which cannot be fitted into any of the above styles or

types, as well as a few figurines, panpipes, and steaming vessel or sieve fragments. T h e material comes from both C u t 1 and C u t 2. It

76

CERAMIC STYLES AND TYPES

will be briefly described and listed here, and significant pieces which can be stratigraphically placed will be mentioned again in this connection. T h r e e solid figurines, possibly mold-made, came from the upper or chala layer of Cut i. Each one has two perforations through the shoulders for suspension and each has its arms folded on the chest. T w o have very large circular earplugs; the third lacks a head. T h e style is apparently Inca or Inca-associated. T h r e e small panpipe fragments came from C u t 2, one from the lowest deposits on the eastern end, two from just below the chala at the eastern end of the main midden. One of the upper specimens is polished brown; the other two lack paint or polish. A l l three occurred in association with Pachacamac Interlocking style pottery. Several plain brown perforated sherds suggest steaming vessels or sieves. T w o from C u t 2 occurred in association with Pachacamac Interlocking style pottery. Identifiable trade sherds from a distance seem rare in the present collection. A n exception is the occurrence of two indubitable Middle lea II (Kroeber and Strong, 1924b, Plate 32g, 321) rim sherds from the lowest portions of the chala in C u t 1. T h e chronological significance of this occurrence will be discussed later. T h r e e polished, light brown sherds with heavy angular purple designs with black edges and dots seem unique. T h e two of these from C u t 2 occurred just below the chala in the western end of the deposit. A somewhat similar sherd with a triangular red design on a white background came from Block 103 in C u t 2. T h e r e are eight thin, highly polished red sherds from C u t 1, and three of these from C u t 2 came from below the chala, one of them occurring in the stratum at the eastern end next to the bottom (Block 213). A similar polished red sherd with a black-line design came from the same block. Four thin, polished, orange-colored sherds from Cut 1 have decoration of dark red and black lines and dots. T h e i r exact provenience is unknown. Seven sherds from Cuts 1 and 2 have two shades of red, the darker forming a geometric

design, and three of these have a white-line edging on the dark red. T h e four sherds from C u t 2 characterized by two shades of red came from deep in the layers below the chala, in association with Pachacamac Interlocking style pottery. It is possible that these seven sherds belong to the White-on-red type of the Pachacamac Interlocking style. Eight thin, gray to black, highly polished sherds came from Cuts 1 and 2. These suggest an early smoked and highly polished type. T h e y seem to be broken from simple open bowls. A l l six from C u t 2 came from beneath the chala, the majority occurring in the lowest layers containing the Pachacamac Interlocking style pottery. Punctate (and incised) type (Fig. 19).—This plain brown or red pottery with punctate, incised, and applique decoration proves to be an important type. It is characterized by crude modeling of animal or other forms with hollow-reed punctate eyes (Fig. 19a, 19b, ige); by raised snakelike bands or ridges with incised, punctate or hollow-reed punctate decorations (Fig. igd, igg); and by a variety of crude modeling, relief, and punctate or hollow-reed punctate decoration (Fig. igf, igh, igj-1). Neat incising occurs (Fig. igi) but is not commonly represented. Strange protuberances or lugs, marked by very numerous small punctations, and a few fragments suggesting animal heads with punctated head bands in relief and teeth indicated by cutting and modeling also occur. T w o hollow fragments of this type from C u t 1, suggesting crude figurines, have faint rectilinear white-paint designs indicating a relationship to the Whiteon-red type of the Pachacamac Interlocking style. As suggested earlier, the early Whiteon-red at Pachacamac is probably a style rather than a type. If this is so, the Punctate (and incised) type, described here, is actually a type pertaining to the White-on-red style as represented at Chancay (Kroeber, 1926b, and Willey, Part 3, present volume, Fig. 9). Plain style.—Only one undecorated, or truly plain, type was segregated, and this is the T h i n bowl plain. T h i s classification is not at all satisfactory, for it is based solely on the selection of rim or other fragments from thin, hard or polished, direct-rimmed bowls

FIG. 19. PUNCTATE (AND INCISED) TYPE SHERDS

78

CERAMIC STYLES AND TYPES

(Table 2). These appeared to be more numerous in the lower strata of Cut 1, and Table 2 shows that the type increases in the lower levels of Cut 2. However, the form is a very simple

and common one, and more exact criteria are needed to objectify the selection and demonstrate the significance of its apparent distribution.

Stratigraphie Distribution of Ceramic Styles (Cut 2)

T

HIS correlation between ceramic styles and soil strata is based on all the sherd material from the stratigraphic blocks in C u t 2. T h e r e are 24,038 of these, of which 19,919 (83 percent) are plain and 4,119 (17 percent) are decorated. Since the plain ware has been only superficially studied, the present classification is based primarily on the distribution of decorated sherds.

T h e exact distribution of styles and types in the stratigraphic blocks and strata of C u t 2 is presented in T a b l e 2. T h e distribution of styles is presented diagrammatically (Fig. 20).1 T o facilitate reference and comparisons between Figures 5 and 20 and T a b l e 2, each horizontal half-meter stratum has been designated by a letter (A-Z on Fig. 20) and the vertical block columns have been indicated by superimposing the number of the top block (Nos. 2-178 on Fig. 20). In the present discussion we have grouped one series of styles (Plain red painting, Polished black, Modeled brown, White-on-red, and Black and white on red) under the stratigraphic or period term Inca-associated (see T a b l e 2). T h i s has been done for purposes of convenience, since the limited amounts representing certain of these styles would have made graphic presentation difficult (Fig. 20). T o anticipate, it may be pointed out that the 1 In this diagram (Fig. 20) all blocks are arbitrarily made the same size. Percentages are figured to the nearest two percent except in cases where the amount present was less than ten percent, when for purposes of visibility it was arbitrarily raised to that amount, with corresponding changes in other percentages. T h e following blocks show the same percentages because they were combined in excavating: Blocks 1, 2; 68, 78; 81, 91; 218-221 inclusive. T h e sherds from Block 158 were lost. T h e original stratigraphic block numbers are given at the end of Table 2. Catalogue numbers conform to this original numbering which was changed into more orderly sequence here.

stratigraphic interrelationships between the Inca Polychrome style and those here termed Inca-associated amply justify this convenient but perhaps slightly premature introduction of a stratigraphic or temporal designation (Fig. 20). It should be remembered, however, as indicated in our description of certain of these Inca-associated styles, that several of these may be non-Inca in cultural affiliation despite their obvious stratigraphic relationship. T h i s point has been discussed in an earlier section (p. 56). T h e soil strata and structures revealed in Cuts 1 and 2 have already been discussed. T h r e e major types of deposit were noted. A t the top is the great Inca refuse heap or chala midden (Fig. 5) and next, the sloping d u m p or slides underlying the chala. T h i s latter was designated as the intermediate main stratum ("gravel and dirt," "dust, detritus, and rocks": Fig. 5). Below this are the consolidated layers of adobe habitation levels, the ash lenses, and the adobita retaining wall (Fig. 5). W e will consider the uppermost of these three major divisions first. T h e great chala midden (Figs. 5, 20) is characterized throughout by Inca Polychrome and Inca-associated pottery. Horizontal Strata A - D (Fig. 20), which practically cap the refuse heap, contain sherds of only these styles. A t the extreme top Inca-associated is somewhat more abundant, but in Stratum D the Inca Polychrome has increased to 63 percent of the total. Strata E and F are similar to the above save that one block (20) on the extreme west lacks decorated pottery. A t this point the upper chala layer impinges on the highest portion of the second or intermediate stratum (Fig. 5). T h i s point of juncture is marked by the uppermost depositional occurrence of Pachacamac Interlocking style sherds. T h e one

8o

DISTRIBUTION OF CERAMIC STYLES

Interlocking sherd isolated in Block 17, like the Pachacamac Negative sherd in Block 51 (Fig. 20), was, apparently, accidentally gathered from older refuse and included with later Inca rubbish deposited on the heap. Inca Polychrome and Inca-associated sherds predominate in the entire chala area, which slopes abruptly down from west to east. O n the west these styles attain their greatest depth in Stratum H (Fig. 5). Inca Polychrome runs out at this point (a single sherd in Block 76 is probably intrusive. T a b l e 2), but Inca-associated is found in the two deeper strata (I, J on Fig. 20). In the central portion of the deposit (Row 34 on Fig. 20) these styles extend down to Stratum N , with the Inca-associated persisting down into the stratum below. O n the extreme east of the midden proper, Inca Polychrome and Inca-associated thin out with the tailing off of the midden deposit (Figs. 5, 20). In and near Block 150, Stratum S (Fig. 20), both occur, and again the Inca-associated extends down into the stratum below at several points. A few sherds have rolled down the hill. T h e entirely Inca Polychrome block (Block 179 in Fig. 20) in Stratum U is so classified because of the occurrence of two decorated sherds, both of that type. A little Inca-associated is indicated in the east end of Stratum V, but this is due to the occurrence of a few polished black sherds which are not entirely diagnostic. In general, it can be said that the Inca and Inca-associated styles occur together, with the latter continuing down to slightly greater depths throughout. T h e Inca Polychrome is most abundant toward the center of the chala deposit, and the proportion of Inca-associated increases markedly toward the bottom (Fig. 20). It is apparent that these styles are not only contemporaneous but were intimately associated during the entire period when the chala midden was being built up (Figs. 5, 20). Since the Inca-associated styles occur beneath the Inca Polychrome and preponderate in the lower strata where they occur together, it may be that they (or some of their types) have a slight priority in time. It is apparent from the midden contour (Fig. 5) that the refuse of the period marked by these two ceramic style groups was dumped down from the

higher ground to the west until the rounded midden was gradually built up. T h e second or intermediate major soil stratum apparently consists of slide materials from the high ground of the temple area to the west (Figs. 5, 6). T h i s layer is ceramically characterized by limited numbers of Pachacamac Interlocking and Negative style sherds (Fig. 20). T h i s intermediate deposit differs markedly from the overlying chala midden in that it lacks nearly all vegetal, textile, and other perishable remains. It is apparently the result of dumping or sliding from the high ground to the west probably as a consequence of temple-building or other activities in the later period. Deposits containing similar ceramic types beneath the present floor of the top terrace of the Sun T e m p l e above the area of our cuts have already been mentioned. Since both the chala midden with its Inca and Inca-associated pottery and this second or intermediate layer with its Pachacamac Interlocking and Negative ceramic styles slope down sharply from the west, the horizontal stratigraphic blocks cut the line of juncture at an angle (Figs. 5, 20), thereby including both upper and lower ceramic types in the same squares. In addition, the soil of this second main stratum is generally loose and there has been considerable rodent work tending to break down exact lines of cleavage. However, on the basis of observations at the site, as well as the actual ceramic distribution and lack of stylistic continuity (Fig. 20), it seems probable that there was a definite time break between the deposition of the upper Inca and Inca-associated and the lower Interlocking and other ceramics. T h e marked change in the nature of the deposits also argues for a considerable time interval between the two periods of deposition. Perishable materials are abundant in the upper Inca layers and almost absent in the Interlocking style layers. As can be seen from T a b l e 2, pottery (and this was also true of other cultural remains) was somewhat less abundant in the "dust, detritus, and rocks" or the "gravel and dirt" (Fig. 5) of this intermediate main stratum than it was in the concentrated chala midden above or the consolidated living levels below.

< I m ! u l û l UJ M

o l i i - I '">1*1-il 5 1 Z i

Ola-lolo:|(fl|i-lDl>l$lx|

H

S U c/3 W •J

h c/} U

S < Pi

w u ÍÜ o o H-1

H D S 2

h C/3 S

o w

o < h

z M

U

Pi w ft* ta O S < Pi

O
o c t j OJ

bO ci

.S

e

^S

8

e-S3 • 3 *

o h

o H

« V)

00

SHERDS

MISSING

H D U,
«) co W

en «S

H 2 W u

00 co

c • «

01 If

•«f i f } «

CT.

to

°

co 0*

O o

W

-H

Ci CO

m o CT)

mo ->

H

OO co

o O

_ 1 Uh

M u IH CS «Lì (U c e

U es l> •S r3 S H -I

5«

u b « c

T3

OJ

H

C ê O 9 - d

(u

ö S s a; G - O t í •S- J 3 T 3 i 3 es t>0 o u r •S S a o S E K m M

U

«

es n

r H

S y ^

a

S3 ai

o « u es O G ¡ g - a a

. . . . U O) 11 c P-c •fi a

2

h U c a

h CJ

bo es 1 h V

u U

a a .s a «g

a

_

s

S

u

H

•o u

C/5 w hJ S " CO h W tzi 0
T O O O CTI « 00

CO c cooo

— ^r

O Tf O O) C. oo oo

R E M O V E D 00

CO C •«f o

io m o

co

o

-< co o co ~ W W I ^

oo

rTf (

Tf

t£> CO © flO IO

N O ~ -

« o

OC to

m co

CO

Ci c ^

X

ce

CO TP COC)

o oc




R E M O V E D

o •a o

J"* eoo o

13 a> h T3 § 3 rt Si « ctu V * - u I S "0 • S s S ' 5 8 9 T3 C •H £ ^¿i o Ù W Ti a «h 5 u cj 3 ¡5 i l JJ 13 e T3 t i flj 5 4J QJ u ^ Ji 3-0 u C ni bcO

§ •

Ui I—I ^

HH t—] £ 3

t

2

PMffi(n

V bo a s«

s ¡3 o ^ v r h Ai

bc

I s s i .a s 1 § aS ou TTt « 2 •5" «Ä o _ s •S S ivo srt u V u sqj gWTjtj O O boS rt SC3O*^ u•Si I s •»-» Vajh Ó ^"m Í.S-S¡2 »QJ rj " 3 u OH < sVskrlo S ü IT® ÄU 'oC rh < U «-y R*fecotoco^ j-i« X¡i> ;I¡ to M

5

o t o t^oo

¡2 V E S c ^ M u (j

•o

C/5 M J ¡ x w> H W V I O*

. S Ô 2 u

a

•S

" « S a

js «

r h

o co —

QJ — ' OJ

ë

Ö

-S

£ ' 3 *0 J) F-H u JO O í g .S § S

8 S .3

e

a

J3 ' S

CJ bo w

« « o u 4J •o

a qj u Ö

.S

e

" n

2 IH (y

.2

§ ' 3 H 0 Ö

o

h

o

h

•3 S

S -a

*l

- 00

O CTS to IO CTS CO

ITS to CO

t^ ITS O ITS •f CTS Oí

CO 00

-i CTS CTi

O oí co OíCT)CTS

OO

00

CO O Ci O

""f CTS

oí CTs

o

o •o ^ "O O ™S •O O "O ^ •a u T3 a •a "5 * § ä -a bjo «-' u u V S

u c

-Ö •O S u 'S«3 aj

«

Û.S~ c » g.H S "tí O -O--«u jQ s y f3,

g § H O Q Í3 A. E

3 V

60 a

ö S

C 60

«

u _o P3 3

•a g û - 0 c 73 ^ - B h 'Sb o o o h h h

g

TABLE

3

PROVENIENCE OF P O T T E R Y Plate 4 a. Pit 1 d. Pit 1 g. Pit 1 j- Pit

b. Pit 1 e. Pit 1 h. Pit i

c. Pit 1 f. Pit 1 i. C u t 1

b. C u t 1 e. C u t 1

c. C u t 1 f. C u t 1

ILLUSTRATED

Plate 5 a. C u t c. C u t e. C u t g. C u t i. Cut k. C u t

2, Block 2, Block 2, Block 2, Block 2, Block 2, Block

194 183 155 170 184 102

b. L». d. f. h. j. 1.

C u lt U U Cut Cut Cut Cut Cut

2, Block AJ1UI.K 2, Block 2, Block 2, Block 2, Block 2, Block

104 J U ^ 193 136 137 144 io2

Plate 6: Pit 1 Figure 7 a. C u t 1 d. C u t 1

Figure 8 a. C u t 1 d. Cut 1

b. C u t 1 e. C u t 1

Figure 9 a. after Uhle (1903, Piate 18, No, 1) b. C u t 1 c. C u t 2, Block 71 d. C u t 1 e. C u t 1 f. C u t 1 g. C u t 1

Figure 10 a. C u t 1 d. C u t 1 f. C u t 1

Figure 11 a. C u t 1 d. C u t 1 g. C u t 1

Figure 12 a. C u t 1 b. C u t 1 d. C u t 1 e. C u t 1 f. C u t 2, Block 69

Figure 13 a. C u t 2, Block c. C u t 2, Block e. C u t 2, Block g. C u t 2, Block i. C u t 1 k. C u t 2, Block

b. C u t 1 e. C u t 1 h. C u t 1

171 153 140 205

c. C u t 1 f. C u t ì

b. d. f. h. j.

Cut Cut Cut Cut Cut

2, Block 1 2, Block 1. below 2, Block

154

b. d. f. h. ; 1.

Cut Cut Cut Cut cut Cut

2, Block 47 2, Block 206 1 1, below Inca chala 1 1

206 Inca chala 124

66

Figure 15 a. C u t i c. C u t 2, Block 196 e. C u t 1 g. C u t 2, Block 171 i. C u t 1 k. C u t 1

c. C u t 1 f. C u t 1

b. C u t i c. C u t 1 e. C u t 2, Block 15

c. C u t 1

Figure 14 a. C u t 2, Block c. C u t 2, Block e. C u t 2, Block g. C u t 2, Block i. C u t 2, Block k. C u t 2, Block

101 197 66 48 195 46

b. d. f. h. j.

C u t 2, Block 102 C u t 2, Block 136 Cut 1 Cuti C u t 2, Block 92

Figure 16 a. C u t 2, Block c. C u t 2, Block e. C u t 2, Block g. C u t 2, Block i. C u t 2, Block k. C u t 2, Block m. C u t 2, Block o. C u t 2, Block

126 103 194 185 172 205 193 184

b. d. f. h. j. 1. n.

Cut Cut Cut Cut Cut Cut Cut

2, Block 2, Block 2, Block 2, Block 2, Block 2, Block 2, Block

205 159 187 154 187 213 205

TABLE 3 ( C o n t i n u e d ) PROVENIENCE OF P O T T E R Y Figure 17 a. C u t i, below Inca chala c. C u t 2, Block 46 e. C u t 2, Block 102 g. C u t 2, Block 140 i. C u t 2, Block 103 k. C u t 2, Block 152 m. C u t 2, Block 193

Figure 18 a. C u t 2, Block 47 b. C u t c. C u t 2, Block 90 d. Cut e. C u t 2, Block 90 f. Cut g. C u t 2, Block 213 h. C u t i. C u t 2, Block 100 j. Cut k. C u t 2, Block 154 I. C u t aln m. C u t 1, below Inca chala

b. C u t 1 d. f. h. j. 1.

Cut Cut Cut Cut Cut

2, Block 1, below 2, Block 2, Block 2, Block

102 Inca chala 104 121 165

Figure 19 a. C u t 2, Block c. C u t 2, Block e. C u t 1 g. C u t 2, Block i. C u t 2, Block k. C u t 2, Block

143 215 131 201 215

ILLUSTRATED

b. d. f. h. j. 1.

Cut Cut Cut Cut Cut Cut

1 2, Block 2, Block 2, Block 2, Block i

213 213 213 95

2, Block 2, Block 1 2, Block 1 2, Block

57 54 136 214

PLATE 2.

EXCAVATIONS BELOW THE TEMPLE OF THE SUN

b

PLATE 3. CROSS S E C T I O N S OF C U T 2 S H O W I N G I N C A CHALA ADOBITA WALL

AND LOWEST

PLATE 4.

P A C H A C A M A C I N T E R L O C K I N G S T Y L E VESSELS, P I T 1; M O L D E D A D O B E , C U T 1 (all scales calibrated in centimeters)

HAND-

PLATE

5. PACHACAMAC INTERLOCKING STYLE SHERDS, MODELED TYPE, CUT 2

PLATE 6.

PACHACAMAC NEGATIVE STYLE BOWL FRAGMENT, INNER TRIANGLE AND D O T TYPE, P I T 1

EXCAVATIONS IN THE GHANCAY VALLEY By GORDON R. WILLEY

COLUMBIA STUDIES IN ARCHEOLOGY AND ETHNOLOGY, Vol. I, No. 3

Contents Introduction

127

Excavation at Cerro de Trinidad, Site E

130

The Pottery from Cerro de Trinidad, Site E

142

The Artifacts from Cerro de Trinidad, Site E

159

The Stratigraphy at Cerro de Trinidad, Site E

167

Baños de Boza

184

Conclusions: Comparisons and Sequence

189

PLATES 1.

White-on-Red Style Pottery from Site E

2.

White-on-Red Style Pottery from Site E

3.

White-on-Red Style Sherds from Site E

4.

White-on-Red and Plain Red Pottery from Site E

5 a. White-on-Red and Plain Pottery from Site E b.

Sherds from Baños de Boza

'6.

Interlocking Vessels and Sherds and "Intermediate" Vessels from Site E

7.

Interlocking Type Sherds from Site E

8 a.

Minority Types of the Interlocking Style, Site E

b. 9.

Various Decorated Types from Site E Vessels from Baños de Boza FIGURES

1.

Map of Cerro de Trinidad, Site E

132

2.

Profile of Pits IV and VII, Site E

136

CONTENTS 3-

Burial Diagrams, Pit VII, Site E

138

4-

Common Vessel Forms from Site E

144

5- Rim Profiles from Site E 6.

148

Border Designs from Interlocking Style Vessels

150

7-

Various Decorated Types from Site E and from Baños de Boza

154

8.

Artifacts from Site E

160

9-

Adornos or Figurines from Site E

162

ÎO.

Pottery Types for Pit IV, Site E (by Percentage)

167

xi.

Pottery Types for Pit VII, Site E (by Percentage)

168

12.

Pottery Types for Pit VI, Site E (by Percentage)

170

Pottery Types for Pit V, Site E (by Percentage)

171

TABLES 1.

Provenience of Pottery Illustrated

156

2.

Provenience of Artifacts Illustrated

166

3-

Grave Association Lists: Pits IV and VII

173

4-

Grave Association Lists: Pit V

173

5- Grave Association Lists: Pit VI

173

6.

Grave Association Lists: Pit II

174

7-

Pottery Types by Occurrence and Percentage for Each Level, Pit IV

177

8.

Pottery Types by Occurrence and Percentage for Each Level, Pit VII

178

9-

Pottery Types by Occurrence and Percentage for Each Level, Pit VI

180

io.

Pottery Types by Occurrence and Percentage for Each Level, Pit V

181

il.

Classification of the Uhle Chancay Collections

183

12.

Pottery from Banos de Boza

186

13- Culture Sequence for the Central Coast of Peru

196

Introduction PROBLEMS: T H E E A R L Y PERIOD IN T H E C H A N C A Y V A L L E Y jAt t h e t u r n of the nineteenth century f \ Dr. M a x Uhle conducted investigations X A . along the Peruvian coast that were to form the basis of a systematic archeology for a large part of the Andean area. A t Pachacamac (Uhle, 1903), in the L u r i n valley south of Lima, he established Tiahuanaco, or Tiahuanaco-Epigonal, as a generic Peruvian style; on the north coast the concepts of a Proto- or Early C h i m u 1 and a Late C h i m u are Uhle's (Uhle, 1913); and in the south he recognized what is now commonly referred to as Nazca, or Early Nazca, besides, in the lea and Chincha valleys, a later series of culture periods (Kroeber and Strong, 1924b). For both the north and the south he postulated the appearance of T i a h u a n a c o i d ware as postEarly C h i m u and post-Early Nazca. A m o n g the early or pre-Tiahuanacoid cultures on the central coast he found a greater diversity: the early incised pottery horizons at A n c o n (Uhle, 1912) and Supe (Uhle, 1925); the early period of painted pottery (White-on-red and Interlocking) in the Chancay valley; and, possibly, Early L i m a in the Rimac valley (Uhle, 1908).2 Of these, it was in the Chancay valley that Uhle first formulated his concept of a developed painted-pottery period for the central coast that equated with the early cultures of the north and south. In the Chancay valley in 1904 Uhle found Late Chancay Black-on-white ware at four sites. A t one of these sites he excavated graves also yielding pottery of the Epigonal and Black, white, and red geometric styles (Kroeber, 1926b, pp. 266-67). O n the basis of these finds he established for the Chancay a sequence comparable to the last three Necropolis periods at A n c o n (Strong, 1925) and, with the exception of Coastal Inca, at Pachacamac 1 Following Kroeber, 1927,1 am using the term "Early C h i m u , " throughout, rather than "Proto-Chimu." 2 Kroeber, 1927, has a summary of this.

(Uhle, 1903). T h e n , on the southern slopes of the Cerro de Trinidad, at what he designated as "Site E , " U h l e discovered the Whiteon-red and Interlocking styles (Uhle, 1908). He recognized the Interlocking pottery as similar to sherds he had found in the fills of the T e m p l e of Pachacamac and of the T e m p l e of the Sun at Pachacamac (Uhle, 1908, p.356 and Fig. 5). H e then regarded the type as postTiahuanacoid, but the occurrence of the style in the Chancay valley, unmixed with T i a h u a naco or Epigonal material or with any of the later wares, suggested its priority to the Tiahuanacoid period of the coast. T h e interlocking fish patterns found at Site E on vessels (Uhle, 1908, Figs. 4, 10; pp. 35357), sherds, and a wall painting (Uhle, 1908, Fig. 6), Uhle related to the stylized fish design in Early Nazca (Uhle, 1908, Fig. 7). Vessel forms and the use of small, handmade adobes in construction were further evidence for strengthening the relationship with the Early Period in the south. He considered the threecolor and, occasionally, four-color Interlocking ware the result of foreign influences from the south, bringing to the central coast its first high civilization. These southern influences, he felt, had stimulated the "primitive" fisherfolk of the shell mounds into altering their techniques for decorating pottery from incising to painting. T h e result of this fusion was the White-on-red style which then replaced the interlocking fish designs of the Nazcalike ware (Uhle, 1908, pp. 352-53). T h e White-onred style, because of its simple but unique decoration, Uhle thought of as a resurgence of the "primitive" pottery-makers, and he later referred to the White-on-red occupation of the Cerro de T r i n i d a d as one of the five recorded sites of the early shell-mound or fishing culture, stating that it was only "a little different" (Uhle, 1908). T h e story in the ground at Site E from

128

INTRODUCTION

which he derived this sequence of first, Interlocking, and second, White-on-red is complicated. In excavating graves, Uhle found flexed burials covered with large sherds of the Interlocking style; and in these same graves were White-on-red vessels. In some of the graves he also "occasionally found unbroken vessels of the older type [Interlocking]." From the evidence of Site E, Uhle argued that the Whiteon-red people had exhumed pottery from the earlier Interlocking graves and reburied it with their own dead, often using broken Interlocking jars to cover the burials (Kroeber, 1926b, pp. 296-98). Kroeber, in a paper dealing with the Uhle collections and with a consideration of the relationships of the two styles at Site E, advances Uhle's scheme with reservation and points out the important weaknesses (Kroeber, 1926b, pp. 283-84). T h e transference from incised pottery to painted, as from Ancon-Supe ware to White-on-red, seems very abrupt. There is actually little resemblance between the Early Ancon design motifs and those of the Whiteon-red. Furthermore, the interposition of the considerably more spectacular Interlocking style between the incised and the White-onred, with few or none of its traces being left as an incorporated part of the later style,

makes such a transition on the spot much less than probable. Technically, the derivation of a sequence based upon the possibility of a later people robbing earlier graves is also extremely insecure. T h e problems for investigation of the Early Period in the Chancay valley are formulated from this work of Uhle's at Cerro de Trinidad, Site E. T h e major objective for research is the determination of the relationship between the White-on-red and Interlocking ceramic styles at that site or at any other site in the valley. This larger problem can be broken down into three subdivisions: (1) Is the relationship one of time difference, and if so is Uhle's sequence correct? (2) If the relationship is one of sequence is it a developmental story or is there a hiatus between the two styles? (3) A more refined point, what is the significance of the various substyles, or types, within the Interlocking group? 8 Secondary objectives are those present for any archeological exploration: the recording of data on all features and artifacts found and the relating of these, as culture traits, into groups or periods on the basis of association or style. T h e essentially historical nature of the problems, coupled with abundant occupational rubbish at Site E, made the stratigraphic method the obvious approach.

P R O B L E M S : SEQUENCE F O R T H E C E N T R A L C O A S T OF P E R U It is the general archeological opinion, and it has also been specifically formalized (Kroeber, 1925a, pp. 231-32), that the development of Peruvian coastal cultures is encompassed in four great periods: (1) an Early or pre-Tiahuanacoid; (2) a Middle or Tiahuanacoid; (3) a Late I or postTiahuanacoid; (4) and a Late II or Inca. This division has been observed for the north, south, and central coastal regions. T h e Early or pre-Tiahuanacoid Period for the central coast is less well demonstrated than Early Chimu or Early Nazca. T h e early or "primitive" shell mound culture of Ancon and Supe may fill a part of this time gap for the central coast and equate with a part of Early Nazca, as Uhle claimed (Uhle, 1920, p. 11). A stylistic similarity between the Interlocking of Chancay and Nazca, as Uhle (Uhle, 1908, pp. 353-

57) and Kroeber (Kroeber, 1926b, pp. 287-89) have pointed out, also would suggest a partial contemporaneity of Interlocking with the Early Period of the south. Other data bearing on the Early Period for the central coast, and its more widespread implications, were obtained by Uhle in the Rimac valley after his work in Chancay (Uhle, 1908, pp. 359-67). T h e style and culture which were found at several places near Lima he termed "Proto- (Early) Lima." Since Uhle's work on the problem of the Early Period in the Chancay-Rimac area, Kroeber, Means, Tello, Strong, and Lehmann and Doering, among others, have examined the data and ventured opinions. Going from 3 Kroeber, 1926b, pp. 279-82, refers to some of these in his " b " and "c" divisions and also, in part, in his "a" division of the Interlocking style.

INTRODUCTION the specific problems of the Early Period in Chancay to the general ones of a central coast sequence, the second major problem of this report is to consider Uhle's and these later opinions in the light of any new data from Chancay. Using the recent Chancay excava-

129

tions and results as a base datum, we propose to refer to it the Supe, Ancon, Rimac, and Pachacamac culture sequences. Correlations farther afield than the central coast are made when these outside influences serve as markers in the local time scale.

M E T H O D OF P R E S E N T A T I O N T h e greater part of the body of this report is devoted to the excavations at Cerro de Trinidad, Site E, and an analysis of the materials found at that site. T h e first section deals with Site E, the actual operations of excavation, and a description of the larger archeological features such as strata, structures, and burials. T h e second section is a descriptive analysis of the pottery, vessels and sherds, from Site E. T h e third section is a similar treatment of the remaining artifacts from Site E. T h e fourth section is an exposition and interpretation of the cultural stratigraphy at Site E as based upon the pottery types. When

possible, this stratification is related to physical strata, structures, burials, and small artifacts. As such, the section is the concluding one for Site E. T h e fifth section briefly summarizes excavations and cultural materials found at Baños de Boza, a small Early Period site in the Chancay valley. "Conclusions: Comparisons and Sequence" is, in the first part, a summing up of the Early Period in the Chancay valley. In the second part, it is a comparison of the Early Period Chancay cultures with cultures of other valleys of the central coast and a consideration of the problem of culture sequence.

Excavation at Cerro de Trinidad, Site E T H E S I T E AND E N V I R O N M E N T

T

C H A N C A Y , the first valley to the north of the converging Rimac and Chillón valleys, is separated from these by forty kilometers of desert hills. It is a triangular patch of green, twenty kilometers broad on the coast and narrowing sharply to an apex as it goes inland. It is nourished by the Rio Chancay and by its irrigation system, which drains out of the Sierra, beginning some eighty or ninety kilometers back from the coast near the Lima-Junin departmental border. The valley and the river are quite typical for the central coastal region of Peru. In size the valley is less than those of Lima, Cañete, and Chincha; the Chancay has a total catchment basin of 2,200 square kilometers and an average annual runoff of 605 millions of cubic meters (Kroeber, 1930a, p. 76). Today it is an important agricultural area, mainly for staples such as flax, although corn and other foodstuffs are grown in small quantities. A little south of the center of the mouth of the valley an isolated land mass of rocks and sandy hills separates the lower Chancay into two parts. The smaller and more southerly is the valley of Pasamayo, through which the Rio Chancay flows to the Pacific. The larger arm, or the Chancay valley proper, in which the pueblo and port are situated, is to the north of this dividing land mass. Fronting the isolated group of hills and separating them from the beach, which is about half a kilometer away, is a narrow strip of cultivated flat lands, the salinas. These extend from the Pasamayo arm almost through to the Port of Chancay. Cutting them off from the port and the Chancay valley proper are two large hills which extend out toward the sea irom this isolated hill group. The larger and more inland of these two interposed hills blocking HE

the salinas from the port is the Cerro de Trinidad; the other, La Punta, extends out into the sea, a rocky guano crag on its seaward face. In 1904, an old railroad cut ran through the low saddle of small hills and ridges which connects the Trinidad with La Punta. Today, the modern Pan-American Highway, running from Lima up the coast, follows the same cut. From Uhle's description (Kroeber, 1926b, pp. 294-96) of Site E, on the south slope of the Trinidad, there has been little change in the last forty years. Sherds, adobes, and deep refuse are visible all along the road cut now as then. On the north side of the Cerro de Trinidad is the large, much-looted Late Chancay cemetery which Uhle called Site A (Kroeber, 1926b, pp. 266-67). This cemetery covers several acres, extending on both sides of the highway. A short distance along the highway toward the Port of Chancay a large huaca has been sliced by road-building operations. Late Chancay Black-on-white sherds in this cut suggest that this structure was related to the Site A cemetery. In the opposite direction on the highway, a kilometer away from Site E, towards Limaj there is a large Late Chancay site on the inland side of the highway. Walls, courtyards, and mounds show this to be an occupation rather than a burial site. This site was not mentioned by Uhle, but recent excavations conducted here by the Institute of Andean Research 4 revealed Late Chancay sherds and rectangular, brick-form adobes. On the east side of the cerro, overlooking this big Late Chancay site, is a large rectangle or enclosure, several hundred meters in each di* These excavations, conducted by the author of this paper and Mr. J. M. Corbett, will be reported upon at a future date.

EXCAVATION A T CERRO DE TRINIDAD mension, built of small, rough rock. Cultural affiliations of this structure are unknown. Across from Trinidad, on the landward slope of L a Punta, are smaller rectangles of stones, apparently the same remarked upon by Uhle (Kroeber, ig26b, p. 295). T h e r e are but few sherds scattered among these L a Punta structures, but in one exploratory walk on the stone-covered slope the writer picked u p two Interlocking style fragments. A s to Site E itself, the sherds found in the road cut were of the Interlocking and Whiteon-red styles with an occasional fragment of Late Chancay Black-on-white. T h e site extends from the lowest slopes of the Trinidad, which are in some places on the seaward side of the highway, back u p the hill for 150 to 200 meters. East and west the extent of scattered rock and sherds is nearer 400 meters.

Actually, the occupation area is south-southeast of the hill instead of due south. A series of weathered and irregular natural terraces marks this side of the cerro. (They are indicated on the map, Fig. 1, in approximate twometer contour intervals.) These terraces are in some places covered with scattered rock and rock-filled pits. Sandy patches, relatively free from rock, appear at intervals. T h e r e are few definite surface evidences of walls on Site E. T h e rectangle of stones, presumably an old structure, indicated on the map (Fig. 1), was the only one noted. Near the western end of the occupation, in a large sandy area, is a series of drift-filled trenches, Uhle's old excavations. It was impossible to tell how deep these had penetrated, but their extent and location is marked on the map, with reference to the recent work, as accurately as possible.

EXCAVATION T o determine culture stratification, seven pits were excavated on the south slope of the Cerro de T r i n i d a d . A l l pits were made 3 by 3 meters at the top. Each was excavated to sterile soil. In two cases pits were excavated contiguously in order to obtain more material and data from important locations. As it is difficult to dig a small hole directly into deep rubbish, extensions or subsidiary excavations were necessary in one instance to facilitate dirt removal and for safety. Sherd material was saved from .50- or .25meter levels, depending upon the particular conditions. From two pits every sherd was saved for later laboratory study. From the others plain body sherds were discarded in the field, tabulations being made per level. A number of burials were recovered from the pits. In nearly every instance these were discovered in the sand below the rubbish, and the graves had not disturbed the overlying soil. Burials and features such as walls and house floors which were encountered in either the pits proper or their extensions will be descriptively treated in this section and later discussed with relation to the stratification.

structure. Spoil dirt from the road excavations has been heaped back over the bank, but a few meters north of the road the ground appears undisturbed. Pit I was located in this area and oriented to the four cardinal points. Level i, 0-.50 meter deep, revealed loose dirt, sand, rocks, and sherds. Sherds were very weathered and sand-blown. T h e y were abundant in the level, and included Interlocking, plain white, and Chancay Black-on-white types. In Level 2, .50-1.00 meter deep, the soil changed to almost pure sand. Sherds became more infrequent with increasing depth. Burials 1 and 2 of Pit I were found at .50-.80 meter below surface in this level. Burial 1, an adolescent, was located .70 meter from the northwest corner; and Burial 2, a small child, was almost directly beneath the northeast corner of the pit, with part of the burial extending outside the area of the pit. T h e r e were no accompanying artifacts with either burial. Below 1.00 meter there were no sherds or evidences of occupation. T h e pit was carried to a depth of 1.90 meters. P I T III

PIT 1

A l o n g the lower slope of the cerro the road cut exposes deep rubbish, rock, and adobe

Because of the similarity of the location of Pits I and III the latter is discussed before Pit II.

EXCAVATION A T CERRO DE TRINIDAD

133

Pit III was located 45 meters west and a little to the north of Pit I. Both of these pits are on the lowest main terrace, or shelf, above the highway. T h e present surface of this terrace is sandy and relatively free from the loose rock that is so common on the slope of the cerro. Level 1, 0-.50 meter deep, was extremely sandy and only a few sherds were recovered, but in Level 2, splintered rock and sherds of Late Chancay Black-on-white, Interlocking, and White-on-red began at a depth of .65 meter. A few stray broken human bones appeared in the rocks and debris. A t about .95 meter below surface a sand and dirt layer was struck. Sherds continued throughout Level 3, 1.00-1.50 meters deep. At the bottom of Level 3 a large, broken, plain red olla was found. Sherds from the olla extended down into Level 4, but all fragments were saved separately from the level collection. In Level 4, 1.50-2.00 meters deep, there were sherds in the upper few centimeters. T h e lower part of the level was of clean sand, devoid of any cultural material.

adobes were turned up, but these were not a part of any immediate structure. T h e adobes were odontiform or circular with one flat side and domed or spheroid on the other. Level 4, 1.00-1.25 meters deep, marked a lessening of broken rock and sherds. A t the top of Level 5, 1.25-1.50 meters, rock fragments abruptly left off and natural sand of the cerro was encountered. A few sherds came from Levels 5 and 6, but in contrast to the rich detritus above, the change was striking. In Level 6, 1.50-2.00 meters deep, an infant burial was found in the sand on the east side of the pit. T h e body had been wrapped in textiles, now almost entirely decayed. T h e position of the body was not determinable. Associated with the burial were two small plain red vessels and a small gourd bowl. From the profiles of the pit it appeared that the burial had been made in the sand previous to the deposition of the rubbish. Below 2.00 meters a test hole of an additional .60 meter was put down as a check for deeper rubbish, but only clean sand was revealed.

PIT N

Pit IV, together with Pit VII and their several extensions, was situated on a small ridge 35 meters west and slightly south of Pit II. T h e ridge is a part of a contour that is approximately 2 meters higher than the area of Pit II (see Fig. 1). T o the west, on the same elevation, are the old trenches excavated by Uhle. Pits IV and VII are conjoined excavations, and they were made, in part, at the same time. After Pit IV had proceeded to a depth of 3.50 meters below surface, Pit VII was begun to serve as a "step" for dirt removal from Pit IV. For convenience and clarity, they will be discussed separately at this time. This also applies to the extensions which were eventually excavated completely around the two pits. Levels 1 and 2 of Pit IV, each .50 meter deep, passed through sand and adobe in the south end of the pit and compacted splintered rock and sherds in the north side. As in Pit II, sherds were extremely plentiful in the mass of rock splinters. Interlocking, Whiteon-red, and one or two pieces of Chancay Black-on-white were observed during the digging of these two levels. In Level 3, however, the lens of rock chips and sherds on the north

Pit II was located 160 meters northwest of Pit III, on a higher terrace than that of Pits I and III. It is in the center of a small level area surrounded on three sides by higher ground. Rock splinters cover the surface at this point, but there are few large boulders. Surface sherds are more numerous than on the lower terrace of Pits I and III. Level 1, 0-.50 meter deep, yielded a tremendous number of sherds. In fact, the first three levels of Pit II were tightly packed sherds and rock fragments with very little soil. Some shell and ash were also present in these levels. In Level 1, Interlocking, White-on-red, and Black-on-white types appeared together. Fragments of human bone, badly weathered, were also mixed through the sherds and rock. Owing to the superabundance of sherds, the following four levels were taken down by .25meter intervals. There was no change in the physical nature of the pit through Levels 2 (.50-.75 meter) and 3 (.75-1.00 meter). A few fragments of dark brown decayed textile came from Level 2, and more weathered human bone was noticed. In Level 3 some loose

PIT IV

134

EXCAVATION AT CERRO DE TRINIDAD

side of the pit was exhausted. I n the main, this level showed a dusty, sandy rubbish intermixed with small fragments of wood. These last, occurring in greatest concentration at the bottom of the level, suggested perishable house walls or a roof which h a d collapsed. T h i s supposition was made more credible by the discovery of a clay floor, 10 centimeters thick (E on Fig. 2), exactly on the bottom of the level. Also in the west profile a puddled adobe wall, .85 meter thick (L on Fig. 2), was revealed in cross section. Sherds in Level 3 n u m b e r e d about one half of those in either of the two preceding levels. I n excavating Level 4, 1.50-2.00 meters deep, it was necessary to cut through the clay floor. T h e adobe wall, which did not extend below the 1.50-meter mark, was clearly related to the floor. Immediately below the floor were a few centimeters of sandy refuse; next came a layer of large, rough boulders interspersed with small bits of broken rock. T h e boulders may have served as a foundation to the floor. Sherds of both the Interlocking and White-on-red styles, as well as large quantities of plain ware, continued through the level. Level 5, 2.00-2.50 meters deep, was m u c h the same as Level 4, with irregular boulders, small rock, and sherds. Level 6, 2.50-3.00 meters deep, was marked by a termination of the boulders and small rock and the beginning of compact, graybrown detritus with only scattered small rocks. I n the southeast corner of the pit there was a small pocket of animal bones. T h e first extension for Pit IV was cut on the south wall of the pit after the termination of Level 6. T h i s extension was made 1.50 meters broad, 2.00 meters deep, and 5.00 meters wide along the south edge of the pit, so that it extended 1.50 meters beyond the original southeast a n d southwest corners. I n order to speed u p the work, sherds were not saved by level from any of the extensions. U p o n completion of this south extension, excavation in Pit IV proper was resumed in Level 7, 3.00-3.50 meters deep. Conditions in this level were very similar to those in Level 6. I n the u p p e r portion of Level 8, 3.50-4.00 meters deep, another prepared clay floor (G on Fig. 2), very similar to the one at 1.50

meters below surface, was encountered. Beneath the floor, in compact rubbish, and definitely not intrusive through the floor, a burial was f o u n d in the southeast corner of the pit at a depth of 3.90 meters. T h e skeleton extended back into the profile with only portions of the leg bones showing in the pit. T h e body, that of a n adult, h a d been wrapped in textiles and placed on some form of wooden litter or platform made of poles. T h e r e was some evidence of a crude boulder tomb a b o u t the burial, b u t the scattered rocks through the soil leave this open to d o u b t o n present evidence. T o judge from the leg bones, the skeleton was in an excellent state of preservation. It was impossible to be sure, b u t the position was probably flexed. N o grave artifacts were found, although it is possible that such existed farther within the profile. T h e condition of the bones a n d the m a n n e r of burial made removal of the feature tempting, b u t this would have been impossible without cutting down an additional block of earth almost 4.00 meters deep. Sherds f r o m the immediate area of the burial were plain or of the White-on-red painted types. More important, n o Interlocking style sherds were noted from this level, a fact verified by later analysis. Level 9, 4.00-4.50 meters deep, had the same dust-gray refuse as the level above. White-onred style sherds were the only decorated types noticed. Near the bottom of the level, in the center of the south profile, was a huge olla, cracked apparently by the pressure of earth a n d rocks. About one t h i r d of the vessel extended into the scope of the pit. W i t h i n the olla were rocks and several of the small handmade, odontiform or dome-shaped adobes, identical with those f o u n d in the rubbish of Pit II. Here, as with the burial above, it was impossible to remove the vessel or its contents without danger of a cave-in or without time-consuming excavation. T h e vessel was flattened globular in shape, very thick a n d coarse, red-slipped, and smeared in places with a white paint. More extensions, this time on the east and west, were excavated before carrying Pit IV down farther. These were also 1.50 meters broad and 2.00 meters deep. T o g e t h e r with the extension on the south, they formed a

EXCAVATION A T CERRO DE TRINIDAD "step" or ledge running around three sides of the pit. In the east extension at a depth of 1.25-1.50 meters below surface another huge, coarse red olla was encountered. It had seemingly been left on the clay floor. T h e floor was the same one recorded at 1.50 meters down in the pit proper and extended at least an additional 1.50 meters east of the limits of the pit. Like the olla at the deeper level, the one above had very carelessly painted white areas over the red slip. It was 53 centimeters in diameter, flattened globular in shape, with a flaring neck and two handles. Decayed vegetal remains and two Interlocking style beakers were found inside. T h e western extension uncovered nothing of interest except the thick, puddled adobe wall which continued running in a westerly direction. Levels 10 and u , 4.50-5.00 and 5.00-5.50 meters deep, produced decorated sherds of White-on-red style, none of the Interlocking. A t the bottom of Level 10 a rather hard, thin, ash-gray lens was broken through. Beneath this was clean, light-colored sand. T h e lowest sherds recovered were taken from the upper half of Level 11. PIT

VII

Pit V I I was excavated contiguous to Pit I V on the north. Surface dimensions were the same, 3 by 3 meters. Unlike Pits I, II, III, and IV, no sherd counts and discards were made in the field, but all material was saved for laboratory study. Level 1, 0.-.50 meter deep, was mostly fractured rock and sherds, comparable to the top level of Pit IV. Beginning with Level 2, all levels were taken down in .25meter intervals. Level 2, .50-.75 meter deep, was mostly sandy rubbish. A dozen or more dome-shaped or odontiform adobes came out of the northwest corner. Level 3, .75-1.00 meter deep, exposed a wall (N on Fig. 2) of small domeshaped adobes extending diagonally across the northwest part of the pit in a northeastsouthwest direction. For the time being, the wall was left in situ. In the lower portion of the level, abundant wood fragments, shell, and matting showed up. Level 4, 1.00-1.25 meters deep, continued through sandy rubbish. Sherds

x

35

were equally numerous in Levels 2, 3, and 4. T h e small adobe wall, which had been temporarily left in place, had its foundations on the clay floor at 1.50 meters deep. T h i s floor was disclosed in Level 5, 1.25-1.50 meters, and was the same one encountered at the same depth in Pit I V and designated as E on Figure 2. In the southeast corner of the pit in Level 5, on this same floor, a large Interlocking style pot was found smashed in situ. A small bed of humus, decayed wood, decayed vegetal matter, and animal bone was around the broken vessel. O n the completion of this level a large extension was added onto the north of Pit VII. Like the south extension of Pit IV this addition was 3.00 meters wider, east and west, than Pit VII, and thus extended 1.50 meters beyond the sides of the pit proper. T h e northsouth dimension of the extension was 4.00 meters. It was excavated rapidly, and only exceptional sherds or specimens were saved. T h e wall of dome-shaped adobes was further followed in its course by this north extension. A t 1.10 meters to the northwest of the adobe wall, and paralleling it, was a wall of large, rough stones (J on Fig. 2). T w o wooden posts, about 15 centimeters in diameter and .50 meter high, had been set in the ground, through the clay floor level, between the two walls and adjacent to the stone wall. T h e posts were about 2.00 meters apart. Near the extreme northwest corner of the north extension, behind or on the northwest side of the rock wall, fragments of two restorable Interlocking pots came to light. These vessels appeared also to have been broken in situ on the Floor E level (1.50 meters below surface). T h e y were enclosed by rocks of what may have been a part of the collapsed stone wall or of a stone-lined cist (K on Fig. 2) or grave. N o bones were found in the area within the limits of the extension, and further troweling among the boulders revealed none. A f t e r making a photographic record, the sections of the adobe and rock walls crossing through Pit V I I and the north extension were removed. T h e dome-shaped adobes are circular, flat on one side, domed on the other, and show handprints. T h e y range from 14 to 18 centimeters in diameter and 12 to 15 centi-

136

EXCAVATION A T CERRO DE TRINIDAD

FIG. 2. PROFILE OF PITS IV AND VII, SITE E Line A-B, approximate division of Pits IV and VII; C, D, E, F, and G, clay floors; H and I, wooden posts; J, crude rock wall; K, cist or corner of room; L, puddled adobe wall; M, rubble-filled wall of crude boulders; N, wall of small, odontiform or dome-shaped, hand-made adobes meters in thickness. T h e wall had been constructed by setting a double tier of the adobes with the flat side down in mud plaster. T h e coursing is even and the wall well made. Extensions 1.50 meters wide were made also on the east and west sides of Pit VII. These connected with the east and west extensions of Pit IV and formed a "step" or gallery 1.50 meters wide on the two sides of the combined pits. A second clay floor 1.30 meters below surface was observed in the west profile of the north extension of Pit VII. This floor, marked D on Figure 2, extends south to about the dividing line between Pit VII proper and the north extension of that pit. A third similar floor showed in this same profile at .30 meter below surface (C on Fig. 2). This one was

limited in extent to the northwest corner of the north extension. It had been built above the stone wall and apparently above the adobe wall. It is evident from the structural detail of Pits IV and VII, and extensions, that these excavations were located diagonally to a building or group of buildings showing different types of masonry. T h e puddled adobe wall (L on Fig. 2), which also runs southwestnortheast, the wall of small adobes (N on Fig. 2), and the rough rock wall (J on Fig. 2) all appear to be related to the clay Floor E at 1.50 meters below surface. It is certain that none of the walls go below this floor. Whether any of the three walls belongs to the clay floors above E could not be determined within the limited area of excavation. T h e resumption of level excavation in Pit

EXCAVATION A T CERRO DE TRINIDAD V I I disclosed, immediately below the clay floor, a small shattered jar of the White-onred style in Level 6, 1.50-1.75 meters deep. Level 6 consisted mainly of sandy rubbish, as did the level above. T h e r e were many sherds in the level, and, as in Levels 1 to 8 inclusive, of this pit, Interlocking and White-on-red were the principal styles. Another clay floor (F on Fig. 2) was cut through in Level 7, 1.75-2.00 meters deep. It is quite probable that the vessel described from Level 6 had been broken on this floor. Levels 8, 9, 10, and 11, 2.00-3.00 meters deep, were relatively unproductive of sherds. T h i s was because a huge block of solid adobe, perhaps a base for Floor F, took u p a large part of the area of the pit for more than three levels. In Level 10 in the northeast corner, not in the hard adobe mass, two very interesting finds were made. T h e s e were small, spherical, unfired clay vessels. In Levels 12 and 13, 3.00-3.50 meters deep, compact, coarse detritus and rock replaced the consolidated clay or adobe. A thin clay floor appeared at the very bottom of Level 13. Sherds were more frequent again, below the clay. A n absence of Interlocking types, noted first at Level 9, was still consistent. Below the thin clay floor (G on Fig. 2) the refuse of Levels 14 and 15, 3.50-4.00 meters deep, was identical in texture and appearance with that of Levels 12 and 13. A marked increase in sherds was noted for Level 15. A part of the north extension was cut down deeper for ease in throwing dirt from the main pit. A t a depth of between 1.50 and 3.00 meters below surface a third sealed vessel of unfired clay was unearthed. T h e depth below surface was approximately the same as for the two previous ones. Level 16, 4.00-4.25 meters, had a large number of sherds, equal to that of Level 15. Soil or refuse content did not change for either Level 16 or Level 17. In the east and west profiles, Floor G and a rubble-filled stone wall almost a meter wide could be clearly discerned at Level 16. T h e floor joined the wall (M on Fig. 2) on its south side. T h e wall was surmounted by the big adobe block referred to in the description of the preceding levels.

»37

Refuse in Level 18, 4.50-4.75 meters deep, grew slightly sandier and looser. T h e beginnings of a sand level, sloping to the south, could be made out in the north part of the pit. In the east profile, center, the top of a large White-on-red style olla, of the same crude, heavy ware as those from Pit IV and extensions, was noticed. T h e olla extended on down into the next two levels, 19 and 20, to a total depth of 5.25 meters. Against the east profile also, but a little to the north and farther within the profile, a second similar olla was discovered while clearing the first. These were designated (Fig. 3) Olla A (the first and more southerly one) and O l l a B (the second and more northerly). T h e s e two ollas proved to be burial units accompanied by large amounts of grave goods. Sherds were saved for Levels 19 and 20. T h e r e were not many in either level, as the rubbish ran out between 4.75 and 5.00 meters below surface. A f t e r the removal of the graves, an additional level, 5.25-5.50, was taken down into the sterile sand. In the course of the work in Levels 19 and 20, the top portions of Ollas A and B were cleaned off and the sand and rubbish was scraped back from around the edges of the two huge vessels. Both were badly cracked, but appeared, in part, to have been cracked in situ by the weight of the earth and rocks. Neither was originally complete when used for the burial purpose. T h e y were not burial urns but presumably halves or large portions of heavy, White-on-red utilitarian vessels forming a shield for the bodies. Placed around the outer edge of each olla were a number of smaller vessels (Fig. 3, left). Many of them were broken, but all were in situ grave offerings and not scattered sherds. A l l were of the White-on-red style. A f t e r taking off a few of the heavy covering sherds of Olla A , the skull of Burial 1 was exposed. Further removal of the sherd tomb, piece by piece, revealed other bowls and jars of the White-on-red style within the covering. These interior offerings must have been closely packed into the grave, for most of them were badly broken and the pieces were mixed. T h o s e vessels which were more nearly whole, both inside and outside of the sherd

EXCAVATION A T CERRO DE TRINIDAD

i38

NORTH PROFILE

NORTH PROFILE

PIT VII DIAGRAM OF

DIAGRAM OF

BURIAL DLL AS

BURIALS

"A*

IN OLLA 'A"

" NTT

I AND 2

A AND B

AFTER REMOVAL

AS

OF SHERDS AND

FIRST

EXPOSED

POTTERY

500-525 METERS DEEP olla'A'

5.25 5.SO METERS DEEP

FIG. 3. BURIAL DIAGRAMS, PIT VII, SITE E covering, contained decayed vegetal matter. A number of bowls had smaller bowls or cups over them, evidently to serve as lids. With the removal of the interior grave goods and most of the big sherds, a second burial was found just to the north of, and partly covered by, Burial 1. Burial 1 was in a kneeling position with the face down and the hands under the head. Burial 2 was also flexed but lying on its right side, facing east, with legs extending under the east profile (Fig. 3, right). In bending forward, the head of Burial 1 almost touched that of Burial 2. Both burials still showed some shreds of the textiles in which they had been wrapped. Some wooden poles, about 8 centimeters in diameter and smaller, had been employed to form either a substructure to the sherd shelter or a litter for the dead. Some of the fragments of these poles showed binding marks suggesting that they had been lashed together. As found in the grave, one pole extended out from the east profile horizontally over the bodies but underneath the large sherd covering; another pole had been set in the ground and crossed the first.

Besides the pottery offering, Burial 2 had associated with it a number of other items of grave furniture. There had been a head wrapping of fabrics and a metal mask or face covering. Shell necklaces with several forms of beads, rather elaborate stone beads or spindle whorls, and an incised clay bead or spindle whorl were the principal artifacts. Burial 2 seemed to be the important personage. There were no specific offerings with Burial 1 other than the pottery. A few restorable pots were taken from the inside of Olla B, where they were mixed with dirt and rock. These were also White-on-red pieces. A human femur was seen within the matrix of dirt, sherds, and rock, establishing Olla B as another sherd tomb similar to Olla A. It was decided not to remove the rest of Olla B because of the difficulties presented by the overhanging dirt and rocks; however, during the night following the discovery and removal of Olla A and its contents, vandals risked gouging back into the profile and tore out most of the Olla B grave. This caused a small cave-in, and it was difficult to see whether all of the grave goods and the burial

EXCAVATION A T CERRO DE TRINIDAD had been looted. Large sherds of White-onred vessels had been scattered around the floor of Pit VII. These were sacked and some of them later restored. In clearing away the debris, a hardwood pointed stick or lance fragment was brought to light. T w o other pieces, possibly of the same artifact, were also picked up in the spoil dirt. All of them probably came from the Olla B burial. PIT V

»39

Levels 7, 8, 9, and 10, from 1.75 down to 2.75 meters deep, were in rubbish similar to that in the lower strata of Pits IV and VII. Large and small rocks were scattered through the dirt, sand, and refuse. It should be pointed out that the refuse below Level 6 and beneath the adobe capping tends to be bedded on a slope to the south. This slope is much more abrupt than that of Pits IV and VII. Near the bottom of Level 10, 2.50-2.75 meters deep, a hard, irregular mass of adobe was encountered. Burial 1 of Pit V was removed from the sand beneath this lowest adobe layer in Level 1 1 , 2.75-3.00 meters deep. T h e burial was located slightly east of the center of the pit. It was a child's skeleton in extremely poor condition of preservation. It had been closely flexed but lacked the skull. Burial 2 of Pit V was found in the southeast corner of the same level in sterile sand. This second burial was also that of a child and similar to Burial 1 except that the skeleton was relatively intact. Burial i was accompanied by a single vessel; Burial 2 by two fragmentary jars. All three pots were of the White-on-red style. From Level 1 of Pit V, 0-.50 meter deep, over 1,500 sherds were recovered. Of these, 617 were discarded. For the next five levels the sherd count was much lower, averaging about 150 sherds per level. Below Level 6 another concentration of sherds comparable to that of the surface level obtained. T h e pit was exhausted in the twelfth level, 3.00-3.25 meters deep, from which only four sherds were taken. Interlocking and White-on-red types were present in almost all levels. Interlocking sherds were more abundant than in any of the other pits and occurred in appreciable quantities from top to bottom of the excavation. A few Chancay Black-on-white pieces came out of the o-.so-meter level, but none was found below this depth.

Pits V and V I were located on the top of a steep sandy mound, 50 meters east-northeast from Pits IV and VII. Pit I I lies midway between these two principal sites. T h e mound into which Pits V and V I were excavated is 25 by 18 meters, rectangular, and flat-topped. It rises 3 or 4 meters above the surrounding terrace. T h e terrace is connected to, and the same height as, the ridge on which Pits IV and V I I were situated. Both pits were 3 by 3 meters in surface dimensions and oriented to the cardinal directions. Pit V, which was excavated first, was on the west, with Pit V I adjoining on the east side. Each pit was taken out in .25-meter levels with the exception of the top level, which was .50 meter deep. Sherd counts and plain body sherd discards were made from Pit V, but from Pit V I all sherds were saved. Levels 1 to 6, inclusive, 0 - 1 . 7 5 meters deep, were virtually pure adobe. Layers could be distinguished in the adobe mass, which was thoroughly cracked, and there were differences in the texture of the various layers. Some were sandy; others had an admixture of small pebbles. Old pits, observed in profile, had been excavated into these adobe layers from the present ground surface. These ranged in depth from a few centimeters to almost a meter. It was impossible to tell how recent these holes were, but they were all filled with sherds of all types and other rubbish. T h e clay layers may have been house floors or they may PIT VI have served as surfacing for a flat-topped pyramidal structure. On one of the adobe levels, T h e eastern side of Pit V was chosen for at 1.25 meters below surface, a number of the addition of another 3-by-3-meter stratilarge rocks, some mixed with ashes and other graphic test, Pit VI. On the west and south evidences of fire, were uncovered. These may sides of Pit V the refuse sloped at a 45-degree have been crude fireplaces related to a specific angle away from the mound. T h e north side structural level. of Pit V was ruled out because of old adjacent

140

EXCAVATION A T CERRO DE TRINIDAD

excavations. From all indications it appeared that the possible center and the most evenly bedded strata of the rubbish lay to the east of Pit V. Level i, 0-.50 meter deep, yielded almost as many sherds as the top level of Pit V. This may have been due to a few refuse-filled shallow pits in the top half-meter, similar to those in the first level of Pit V. Likewise, Levels 2, 3, 4, and 5, from .50 to 1.50 meters deep, consisted of layers of cracked adobe. This adobe capping contained a considerable number of sherds, although, as in Pit V, these levels showed a smaller amount of pottery than either the surface level or the refuse strata below the adobe. Near the bottom of Level 5, 1.25-1.50 meters deep, were the remains of some reeds which had been bound together to form matting. At about the same depth, in another part of the pit, a layer of hard, burned clay appeared, presumably a hearth. Immediately above the hearth area was a lens of ash, sherds, vegetal remains, shell, and midden. T h e burned clay stratum continued for 15 to 20 centimeters into Level 6. These evidences of occupation correspond in depth to the ash and fire-blackened rock found in Pit V. Levels 6 to 10, inclusive, were excavated through loose to semicompact rubbish mixed with large and small rocks, lumps of adobe, and small pockets and beds of ash. T h e number of sherds per level for these five levels remained roughly constant at an average of about 800, four times as many per level as in the cracked adobe above. In the southeast corner of Level 10, 2.50-2.75 meters deep, the facial half of a child's skull was found without immediate associations. In Level 1, about 50 Late Chancay plain and decorated sherds were observed in the field. No Late Chancay material was found below this level. It seems likely that these sherds were in the surface sand or in the rubbish-filled pits discussed above. Although White-on-red decorated sherds were found in small quantities throughout Pit VI, Interlocking types obtained in greater numbers in most of the levels. However, a proportional decrease in the Interlocking was noted from

top to bottom. In Level 12, 3.00-3.25 meters deep, no Interlocking sherds were present among the 126 fragments recovered. In the lower part of Level 11, 2.75-3.00 meters deep, a burial of a child (Pit VI, Burial 1) was uncovered in the natural sand underlying the refuse. T h e facial fragment from Level 10 may have belonged to this individual. T h e skeleton was enclosed by a fragment of a large white-slipped jar. T h e jar had been broken into several pieces: a bottom portion had been used to cover the head and upper part of the body; the middle portion lay around the feet of the burial; and the upper part of the vessel lay off to one side. Inside the sherd covering, directly over the head of the child, was a medium-sized black and white on orange bowl. T h e skeleton lay on its back, with arms folded upon the chest, and legs drawn up in a flexed position. It had been wrapped in textiles which were now brown and black with decay. Orientation was northeast-southwest with head to southwest. Besides the black, white, and orange bowl, a smaller bowl with similar decoration was found with the burial. Neither of these vessels is typical of the three-color Interlocking style or of the White-on-red. In addition to the pots, a bee or bird effigy and a human figure of pottery, shell beads, and some round, hollow lumps of unfired clay were found as grave furniture. In the basal sand, but in Level 12, 3.003.25 meters deep, Burial 2 of this pit was found under similar circumstances. Burial 2 was on the western side of the pit and covered and underlain by large sherds of a White-onred style vessel. T h e body, that of a small child, had been wrapped in textiles and probably had been flexed. It was oriented eastwest with head to the west. Directly below the huge sherds serving as a floor to the grave was a small White-on-red style bowl. Another bowl of the same style was encountered just to the south of the burial. In the same level and sand stratum, Burial 3, another child, was found in the center of the pit. This burial was covered with sherds from a large, plain, red-brown olla. Oriented north-south, with head to south, it had prob-

EXCAVATION AT CERRO DE TRINIDAD ably been interred in a flexed position and textile-wrapped. No artifacts were in association. Deeper in the sand, at 3.25-3.50 meters below surface, Burial 4, an infant, was located extending under the south wall of the pit near the southeast corner. The bones were in the lower half, or side, of a large White-on-

141

red flattened globular bowl. The other half of the bowl formed a cover. A second Whiteon-red style bowl had been placed with the burial. No sherds, burials, or any other evidences of human occupation were discovered below 3.50 meters deep, and the excavation was terminated.

The Pottery from Cerro de Trinidad, Site E

T

from Site E is easily divided into two major styles, the Chancay White-on-red and the Interlocking. These were terms used by Kroeber to describe two lots of specimens obtained by Uhle from the site (Kroeber, 1926b). These lots, according to Uhle, represented a "first" and a "second" period (Kroeber, 1926b, p. 275). T h e salient features of the vessels which Uhle had segregated into his "first," or E i , period (Interlocking) were the use of three colors (red, white, black), the interlocking fish pattern and its variations, and cylindrical bowl forms. T h e "second," or E2 period (White-on-red), was characterized by crude, simple designs of white on a red-slipped background, and jar and bowl forms differing from those of the E i period. These standards, however, were not consistent for all pottery from each of the periods. In the E2 period Uhle listed specimens of three colors (Kroeber, 1926b, p. 278), and there were a number of vessels designated as E i which did not show the interlocking fish motif in any of its variations (Kroeber, 1926b). In other words, style did not conform completely to period. Lacking grave association for the specimens, Kroeber proceeded on the basis of the E i and E2 classification of Uhle. T o quote Kroeber (1926b, p. 276): HE POTTERY

Plausible as Dr. U h l e ' s interpretation of the situation at site E is, it is accordingly a subjective one. However,

the o b j e c t i v e facts as to association

of

specimens in the g r o u n d a n d w i t h i n separate burials not having been

obtainable

or b e i n g

no

longer

available, the collection will be e x a m i n e d as d i v i d e d by Dr. U h l e i n t o lots E i

a n d E2. T h e s e two des-

ignations therefore d o n o t . . . refer to graves; n o r are they e m p l o y e d with unreserved acceptance of Dr. U h l e ' s v i e w as to the lots b e i n g temporally distinct. T h e y are used as e n f o r c e d groupings w h i c h

obviously c o n f o r m in the m a i n to a real distinction of some sort w i t h i n the site E ware.

T h e present use of the concepts of these two styles differs in some degree from that previously employed. T h e Chancay White-on-red style has been made to include only that pottery showing the White-on-red decoration or other closely related types of vessel ornamentation and surface treatment. T h e Interlocking style has been restricted to include only those vessels and sherds which are decorated with some form of the interlocking fish or commonly allied design motifs and elements. Each style has been subdivided into types based upon surface treatment or decoration. Outside of the range of these two styles, other types and styles have been designated. Some of these, as the types Coarse plain red and Smooth plain red, are related in form and quality of ware to the two major styles. Both of these types are numerically important in the sherd counts. Other types, such as a Negative painted ware, Polished black ware, and Late Chancay Black-on-white, bear no close relationship to the White-on-red or Interlocking. T h e first two are quantitatively trivial and probably represent trade influences. Chancay Black-on-white ware is clearly later in time. In describing the pottery, whole and restorable vessels from graves or other in situ proveniences are treated separately from the sherds, although the same classification applies to both. In some of the types a much greater range of decoration and rim form is observed in the sherd material. Also, the numerical strength of a type arrived at by a sherd count may vary greatly from that derived from a count of whole specimens. T h i s may be explained, first, by the relatively few graves ex-

POTTERY FROM CERRO DE TRINIDAD cavated and, second, by the fact that a pot of the type White-decorated, for instance, could shatter down into rubbish sherds which would be classified under more than one type. Plain fragments would be considered as Smooth plain red, white bottom pieces as White-slipped, and only the area of decoration itself, which is relatively small, as Whitedecorated. There is no satisfactory way to

143

obviate this difficulty. T h e fineness of the typological breakdown is not a descriptive exercise but a means to facilitate stratigraphic studies. T h e type is completely illustrated only by the vessels which are aggregates of modes (Rouse, 1939) in decoration and pottery construction. T h e study, stratigraphically, of sherds is actually a tabulation, per level, of modes rather than of whole types.

T H E W H I T E - O N - R E D STYLE There are 38 complete or partially restorable White-on-red vessels from Cerro de Trinidad, Site E, in this collection. These divide into five main decoration types. T h e vessels of this style have been classified according to vessel form following Kroeber's original shape classification (Kroeber, 1926b, p. 277). T h e decoration types do not correspond exactly with the form classification, although there are some partial correlations. T h e ware of the White-on-red style is hard and fired a dull red. T h e walls are medium in thickness, ranging from 4 to 7 millimeters, with an average closer to the lower figure. W i t h few exceptions, there is little variation in thickness among rim, side, and bottom areas. No evidence of differential firing is observed in paste cross section. T h e paste is fairly even and sandy, and coarser particles of grit have been added as a temper. T h e exterior surfaces are smooth but not polished. They are red-slipped, and the red varies from a bright orange-red to a fire-clouded brown or black. T h e form classification for the style is given as follows (see Fig. 4): I. Bowls A. Lipless 1. Low, flaring 2 2. Low, incurved or vertical-walled o 3. Low, incurved, 2 handles o B. With lip 1. Incurved 1 2. Incurved, 2 handles 12 II. Jars A. Broad mouth 1. No definite neck o 2. Two to four bulges o B. Vertical or flaring neck 1. Vertical or flaring neck, no handles 6

15

21

2. Vertical or flaring neck, 2 handles or knobs 3. Vertical neck, 1 flat or round handle on neck C. Mammiform 1. Small 2. Large D. Cylindrical, with or without rim of knobs E. Small mouth, large knobs or handles F. Double spout G. Bird III. Eccentric forms

1 o 8 6 o o o o

TOTAL VESSELS

2 38

Although the complete form classification of the Uhle Ei collection is not necessary for the present material, it has been given for comparison. T w o small modifications have been made: II C, "Mammiform jar," has been subdivided into "Large" (Fig. 4d, 4e) and "Small" (Fig. 4b); and III, "Eccentric forms," has been added for the recent collection. Summations of the vessel forms for each decoration type are listed under the following type discussions. TYPE:

WHITE-DECORATED

Vessels.—The ten vessel specimens of the White-decorated type may be summarized under the following forms: I I I II II

A 1. B 1. B 2. B 1. C 1. III.

Bowl, lipless, low flaring Bowl, with lip, incurved Bowl, with lip, incurved, 2 handles Jar, vertical or flaring neck, no handles Mammiform jar, small Eccentric form

1 1 4 1 2 1

This is the marker type for the White-onred style as it is found in the Chancay valley. It is easy to recognize even in small sherds. Design units such as dots, triangles, lines or

FIG. 4. C O M M O N VESSEL F O R M S F R O M S I T E E a, incurved bowl with two horizontal strap handles; b, small mammiform jar; c, small jar with vertical neck; d and e, large mammiform jar; f, small jar with flaring neck; g, bowl with out-slanting sides; h, beaker bowl, a - g are essentially White-on-red forms; h is an Interlocking form. Size: a, b, f, h, % natural size; d, e, Yio natural size; g, % natural size

POTTERY FROM CERRO DE TRINIDAD thin bands, diamond forms, and circles are crudely executed. Combinations of nested triangles filled with dots, crisscrossed lines forming a series of diamonds, and fields of dots are frequent. An occasional meandering-line border is noted. Most of the bowls of this type have a marked median ridge running horizontally around the vessel and marking the maximum diameter of the vessel. T h e line and dot decoration is placed on the upper side between the vessel mouth and the median ridge (Plates lb, id, le, 2f, 2i). T h e custom of painting half of the vessel white and leaving the other half red is common in both the White-decorated and White-zoned types. T h e incurved bowls of the White-decorated type sometimes have the lower half below the median ridge slipped white. Broad white bands often extend from the median ridge to the rim in the vicinity of the handles. Handles are of the strap type and always horizontal to the plane of the vessel mouth. T h e small mammiform jars of the Whitedecorated type (Plates la, 2b) are divided vertically with the more protuberant side of the jar white. White dot decoration appears on the opposite side. T h e one "eccentric form" (Plate if) of this type follows the tradition in having the lower half of a lobate or zeppelin-shaped body painted white. T h e upper or red surface is dotted in white. T w o incurved bowls (Plate lb, id), a jar (Plate 2c), and a flaring bowl (Plate 2d) are completely red except for the white ornamentations. An incurved bowl, without handles, which is one of these three, verges upon the jar form and proportions. Actually its diameter of" 17 centimeters slightly exceeds its height. A flaring bowl is a darker red than the other specimens in the White-decorated type or in the White-on-red style. This color difference appears to be in the nature of the slip rather than in the firing. T h e bowl is maroon, and the decoration of the exterior border, which is a series of joined diamonds each enclosing a circle, is of an orange rather than a white cast. There are no other vessels of this color, although a few sherds resembling this ware and decoration were found. Sherds.—Nearly all of the sherds of this type display the same ware qualities as the

145

complete vessels. Only a few coarse, heavy ware fragments with white decoration obtained. A number of new forms, however, are indicated. These include a bowl with incurved walls and a low flaring collar rather than a lip (Fig. 5 o) and a very shallow bowl or plate with an upturned lip or edge (Fig. 5a). T h e following are variations in the Whitedecorated type. Most of these were not found on the complete vessels. 1. Single or double rows of circles around the rims of bowls and jars (Plate 3a, 3c) 2. Fields of white dots (common also on complete vessels) (Plate 3g) 3. Inscribed rectangles (Plate 3p, 3q) 4. Crosshatching (Plate 3I, 3m) 5. Bands of alternating horizontal and vertical lines (Plate 3i, 3j) 6. Volutes or inscribed semicircles extending down from the rim (Plate 3r, 3s) 7. Groups of vertical lines extending down from the rim (Plate 3n, 3 o) 8. Nested triangles filled with dots (common on the complete vessels; not common on sherds) (Plate 3!) 9. Opposed diagonal lines extending down from the rim (Plate 3k) 10. Band of broken horizontal lines or "dashes" around rim (Plate 3h) 11. Combination of rectilinear volutes or serried diamonds with circles (found almost exclusively on maroon ware decorated with light orange) 12. Combination of inscribed triangles, rectangles, and circles (the few coarse ware sherds were of this type) (Plate 3d, 3e) 13. Miscellaneous designs (Plate 3t, 3U) TYPE:

RED-ON-WHITE

Vessels.—The four vessels of the Red-onwhite type may be summarized under the following forms: I A 1. B o w l , lipless, low,

flaring

I I B 1. Jar, vertical or flaring neck, n o handles

i 2

I I B 2. Jar, vertical or flaring neck, 2 handles or knobs

1

This is a minority type which is related in decoration and ware to the previous Whitedecorated type. Over the original red-fired

P O T T E R Y FROM CERRO DE TRINIDAD

146

surface a w h i t e slip was a p p l i e d o n either the exterior or the interior surface or o n both surfaces. D o t or line designs in red are applied on a w h i t e background. A j a r w i t h a flaring neck (Plate 4b) a n d a low flaring b o w l (Plate 4d) have an exterior w h i t e slip as a g r o u n d for dark red dots. A second j a r has a white-slipped interior, a n d the red design is composed of several groups of crudely d r a w n lines e x t e n d i n g d o w n f r o m the l i p for a few centimeters o n the interior. A jar w i t h two n u b handles (Plate 4a) is very carelessly divided, w i t h the u p p e r half or two-thirds of the vessel white, the other portion red. Simple lines a n d cross-hachure b e l o w the rim, o n the w h i t e exterior surface, are done in red. Sherds.—There are sherds of b o t h fine (Plate 3V, 3W) a n d coarse (Plate 3X) ware. R i m sherds of small straight-necked jars are in b o t h groups. T h e shallow b o w l a n d plate forms b o t h occur in the fine ware. O n e coarse sherd is u n d o u b t e d l y a part of the mammif o r m b u l g e of a large jar. Decoration is limited to red dots a n d stripes o n the w h i t e background. TYPE:

COMBINATION WHITE AND RED DECORATED

Vessels.—The vessel of this type may classified according to f o r m as: I B 2, bowl, with lip, incurved, 2 handles

be 1

T h i s specimen is a large incurved b o w l (Plate IC). B e l o w the m e d i a n ridge it is whiteslipped. A b o v e the ridge the red field is decorated w i t h carelessly arranged rows of large w h i t e dots. W h i t e semicircular zones extend above the ridge to enclose the handles o n each side of the b o w l . Dark red dots are placed o n these zones. Similar red dots are o n the interior of the vessel lip, w h i c h is painted white. TYPE: WHITE-ZONED (FINE WARE)

Vessels.—The nineteen vessels of the Whitezoned (fine ware) type m a y be summarized u n d e r the f o l l o w i n g forms: I II II II

B 2. B 1. C 1. C 2. II?.

Bowl, with lip, incurved, 2 handles Jar, vertical or flaring neck, no handles Mammiform jar, small Mammiform jar, large [Fragmentary small jar]

5 3 6 4 1

T h i s type, numerically the greatest, includes all vessels whose only surface decoration is w h i t e slipping of large zones, such as the bottom, or p a i n t i n g of very wide, bold, w h i t e bands over the red-slipped ground. T h e incurved bowls w h i c h b e l o n g to this type h a v e the bottom below the m e d i a n ridge whiteslipped and usually are further m a r k e d by w h i t e zones e x t e n d i n g f r o m the ridge to the m o u t h in the region of the handles (Plate lg). A s has been p o i n t e d out, b o t h these types of zoning also occur o n incurved bowls w h i c h are typed as White-decorated because of the added line a n d dot ornamentation. O t h e r forms of zoning include white slipping of the u p p e r portion of the vessel rather than t h e l o w e r (Plate 2a), b o t h horizontal (Plate i h , li) a n d vertical (Plate 2h) division of small m a m m i f o r m jars into red a n d w h i t e areas, alternating w h i t e a n d red vertical bands (Plate l j ) r u n n i n g the complete height of the vessel, bands encircling the vessel immediately below the r i m (Plate 2e), a n d very carelessly a p p l i e d large irregular bands o n the body. T h e small m a m m i f o r m jar, a f o r m peculiar to the White-on-red style, is c o m m o n in this White-zoned type. T h e y are quite u n i f o r m in size, v a r y i n g f r o m 18 to 20 centimeters in height, 15.5 to 17 centimeters in m a x i m u m width, a n d 12 to 13 centimeters across the smaller width. T h e y have small, straight spouts, a n d two small strap handles are placed equidistant f r o m the spout. In some instances nubs are used in place of handles. O n e side of the j a r is relatively flat; the other protrudes a n d is often p a i n t e d white. T h e protuberance is f o r m e d by placing a single circular a n d conical pat of clay in the w a l l of the j a r as a final step in the construction. T h i s is true of the large as well as the small m a m m i f o r m jars. T h e f o u r large m a m m i f o r m jars of this decoration type were all f o u n d in a very fragmentary condition (Plate 4e). T h e y range f r o m 50 to 30 centimeters in height a n d are almost as b r o a d as high. T h e y all have rather small mouths (10 centimeters in diameter) a n d short straight necks. T h e w h i t e paint is a p p l i e d in irregular zones r u n n i n g more or less vertically f r o m neck to base. T h e m a m m i f o r m protuberance is small a n d b l u n t l y conical, measuring

POTTERY FROM CERRO DE TRINIDAD about 12 centimeters across. It is located o n the center of one side of the body. Sherds.—The sherds of this type conform closely to the complete vessels. T h e incurved, lipped bowl with broad vertical white bands on the sides is represented by numerous rim pieces (Plate 3yy). T h e same is true of large, straight-necked jars (probably mammiform) (Plate 3Z). O n e small spout with an everted lip is in the collection. TYPE:

WHITE-ZONED

(COARSE

WARE)

Sherds.—No complete vessels of this type were recovered, but the broken ollas used as burial covers from Pits IV and V I I are fragmentary examples. Size in some cases is as much as one meter in diameter and height. These huge vessels are globular or flattened globular in shape, with short flaring collars. A m o n g the other vessels are smaller incurved bowls with short collars, large jars (perhaps mammiform) with small straight necks, and small straight-necked jars. Vessel walls average over one centimeter in thickness, and the paste is coarse and tempered with large particles of grit. Exteriors are fairly well smoothed and are red-slipped. T h e white decoration is applied haphazardly. Large white swaths encircle the neck (Plate 3y) and upper part of the vessel, with wide vertical bands joining these and extending down the vessel sides. Vertical white zones are sometimes placed on the necks of jars (Plate 3ZZ). TYPE:

WHITE-SLIPPED

(FINE

WARE)

Vessels.—The three vessels of this type may be summarized under the following forms: I B 2. Bowl, with lip, incurved, 2 handles II C 2. Mammiform jar, large

2 1

T h i s type is completely slipped on the exterior surface with a white paint. T h e white is the same chalky dead white that is observed throughout the White-on-red style. T h e incurved bowl forms (Plate 4c) with two horizontally placed strap handles are the same as those of the White-decorated and White-zoned types. T h e large mammiform jar is over 50 centimeters high and 40 and 30 centimeters broad and has two very small, apparently non-

147

functional, nubs, one on either side of the neck (Plate 4f). Sherds.—Among the sherds, the most common form is the shallow bowl or plate. These are white-slipped on the bottom and redslipped on the interior (Plate s A j ) . T h e r e are two principal variations in the rims of the plate form. O n e is the upturned rim which leaves a definite angle on the outside of the vessel (Fig. 5a). T h e other is more gently upcurved, and on the interior there is a thickening or bulge on the lip (Fig. 5b). Shallow open bowls (Fig. 5I1), bottle forms, short-necked jars (Fig. 5c), and incurved bowls with low collars (Fig. 5 o) also occur in this type. O n e fragment of a hollow stirrup-spout is shown on Plate 5Ai. TYPE:

WHITE-SLIPPED

(COARSE

WARE)

Vessels.—There is only one restorable or partially restorable vessel of this type; it is classified as: III. Eccentric form

1

T h e eccentric form is a restored fragment of an obviously larger vessel (Plate 2g). It is 27 centimeters high and about 13 centimeters in diameter. W h a t is now an orifice appears to have been a connection with a larger vessel. T h e opposite end comes to a point at which four prominent ridges converge. It seems to be a plant effigy, perhaps a cactus. Sherds.—The sherds show this to be a coarse ware very similar to that of the White-zoned (coarse ware) type. It averages well over one centimeter in thickness. Extremes of 2.5 centimeters are noted. T h e white paint is used as an exterior slip. Some sherds appear to have been first red-slipped, but most of them have only a thin white wash over the original surface. T h i s latter flakes off easily to expose the coarse, brick-red, heavily tempered paste. T h r e e main vessel forms are recorded: a bowl with incurved sides, a jar with a short neck (Fig. 5r), and a big open bowl with outcurving sides (Fig. 5q). Of the first, the bowl or olla with incurved sides, there are a number of rim or neck variations. A few rim sherds show no lip or neck. Most common is a short, straight collar (Plate 5Aa). Others are short but outflared (Plate

FIG. 5. RIM PROFILES FROM SITE E Interiors to right, a, c, m, o, White-decorated; b, h, i, n, p, Smooth plain red; d, e, g, j, k, 1, Interlocking type; f, Circle type; q, r, Coarse Plain red. (These same forms are also found in connection with other decoration or surface finish types as explained in text.) Size: ¥2. natural size

POTTERY FROM CERRO DE TRINIDAD 5 A b ) . S o m e are v e r y h i g h a n d s t r a i g h t o r v e r y slightly outslanted. Strap handles are a feature of s o m e of t h e b o w l s . T h e s h o r t n e c k s o n t h e jars are either straight or very slightly everted at t h e t o p . T h e b i g b o w l s are c h a r a c t e r i z e d b y a h e a v y t h i c k e n i n g of the r i m (Fig. 5q). I n this type, a n u m b e r of c u r i o u s r i d g e d sherds s i m i l a r t o the r e s t o r e d p a r t of t h e vessel THE

TYPE:

described above occur. T h e ridges r u n longitudinally o n long, lobate forms (Plate 5Ae). Sometimes the ridges are n o t c h e d a l o n g the edges ( P l a t e 5 A d ) . I n o t h e r instances t h e y a r e i n d e n t e d w i t h d e e p c u p l i k e depressions ( P l a t e 5AC). B e s i d e s the ridges, some vessels h a v e r o u n d knoblike additions similarly indented (Plate 5AI1).

INTERLOCKING

T h e I n t e r l o c k i n g style is d e f i n e d b y t h e use of a c o n v e n t i o n a l i z e d , t e x t i l e - d e r i v e d , interl o c k i n g fish design. T y p e s w i t h i n t h e style a r e denoted by rectilinear a n d curvilinear renderi n g of the d e s i g n a n d b y a n u m b e r of m i n o r e l e m e n t s s u c h as circles, fields of dots, a n d X marks w h i c h are sometimes used in conjunct i o n w i t h the fish m o t i f . T h e d e c o r a t i o n is i n black and white, or black, white, and bright red, o n a r e d - s l i p p e d g r o u n d . A n i n t e r l o c k i n g fish f r e t is f r e q u e n t l y u s e d as a b o r d e r decoration b e l o w the rim. INTERLOCKING

Vessels.—The I n t e r l o c k i n g t y p e p r o p e r is c o n c e r n e d w i t h a l l of the r e c t i l i n e a r v a r i a t i o n s of t h e i n t e r l o c k i n g fish p a t t e r n u s e d as vessel body decoration. Inscribed triangle o n chevron d e s i g n s ( " r e d u c e d f a c e s " ) are also i n c l u d e d in the t y p e . A s d e f i n e d , it i n c l u d e s m o s t of K r o e b e r ' s " a " d i v i s i o n of t h e I n t e r l o c k i n g style, o m i t t i n g e l e m e n t s l i k e r o w s of dots o r circles ( K r o e b e r , 1926b, p p . 2 7 9 - 8 1 ) . T h e five res t o r e d vessels of t h e I n t e r l o c k i n g t y p e m a y b e s u b s u m e d as f o l l o w s , a c c o r d i n g to f o r m : Large beakers, straight or outslanting sides Large beakers, slightly outflared sides Mammiform jar, large

3 1 1

T h e s e b e a k e r s a v e r a g e 20 c e n t i m e t e r s in h e i g h t a n d a b o u t the same i n m a x i m u m d i a m eter, w h i c h is at o r n e a r t h e orifice (Fig. 4I1). O n e of t h e straight-sided b e a k e r s ( P l a t e 6c) a n d the b e a k e r w i t h the s l i g h t o u t f l a r e ( P l a t e 6a) h a v e o n l y the i n t e r l o c k i n g fish f r e t d e s i g n as a b o r d e r , w i t h t h e rest of the vessel l e f t p l a i n . T h e o t h e r t w o straight-sided b e a k e r s h a v e a h i g h l y s i m p l i f i e d f r e t b o r d e r a n d chevr o n or inscribed triangular designs covering the e x t e r i o r w a l l s ( P l a t e 6b, 6d).

149

STYLE

T h e l a r g e m a m m i f o r m j a r ( P l a t e 6g), conf o r m s closely i n s h a p e to those of the W h i t e on-red style. T h i s o n e is a l m o s t 50 c e n t i m e t e r s h i g h a n d 40 c e n t i m e t e r s i n m a x i m u m w i d t h . T h e n e c k is s t r a i g h t , b u t t h e r e is a s m a l l f l a r e d o r e v e r t e d l i p . T h e r e are n o h a n d l e s o r n u b s o n the j a r . C o n s t r u c t i o n a n d f o r m are b e t t e r a n d m o r e c a r e f u l l y d o n e t h a n i n the W h i t e o n - r e d style. T h e j a r is a l m o s t flat o n o n e side a n d f u l l y r o u n d e d o n the o t h e r , a n d a d e f i n i t e a n g l e o r r i d g e m a r k s t h e j u n c t u r e of the flat a n d r o u n d e d sides. T h e p r o t u b e r a n c e is in the c e n t e r of the r o u n d e d side, a n d h a s b e e n f o r m e d b y a single r o u n d p i e c e of clay. T h e d e c o r a t i o n is c o n t a i n e d in a b r o a d b a n d r u n n i n g a r o u n d the u p p e r h a l f of t h e j a r . T h e n e c k , h o w e v e r , is l e f t p l a i n . T h e d e s i g n is, t y p i c a l l y , l a i d o u t in b l a c k w i t h b r i g h t r e d f o r a filler b e t w e e n the b l a c k lines. W h i t e is u s e d t o o u t l i n e the b l a c k . T h e h e a d s of the fish, l a r g e a n d t r i a n g u l a r , w i t h b i f u r c a t e d apices for m o u t h s and smaller triangles or rectangles f o r eyes, are i n t e r l o c k e d in a c o m p l i c a t e d design. Sherds.—Nearly a l l of the sherds of the Int e r l o c k i n g t y p e c o n f o r m i n w a r e q u a l i t y to t h e r e s t o r e d vessels j u s t discussed. T h e r a n g e i n thickness, w i t h a f e w e x c e p t i o n s , is f r o m 3 t o 7 m i l l i m e t e r s . T h e e x c e p t i o n s are sherds of a coarse w a r e , t e m p e r e d w i t h h e a v y g r i t a n d v e r y s i m i l a r to the o t h e r coarse w a r e s des c r i b e d f r o m the site. T h e thickness of these sherds a v e r a g e s 1.5 c e n t i m e t e r s . A g r e a t v a r i e t y of d e c o r a t i o n is r e p r e s e n t e d i n t h e s h e r d c o l l e c t i o n , b u t it is a l l a l l i e d t o the i n t e r l o c k i n g fish m o t i f . D e c o r a t i o n is app l i e d t o e i t h e r the e x t e r i o r o r the i n t e r i o r of vessels, a n d s o m e l o w b o w l s a n d p l a t e s h a v e b e e n d e c o r a t e d o n b o t h surfaces. T h e use of the i n t e r l o c k i n g fish f r e t i n o n e f o r m or an-

150

POTTERY FROM CERRO DE TRINIDAD

RED WHITE BLACK.

i m

a»«£-yaum

FIG. 6. BORDER DESIGNS FROM INTERLOCKING STYLE VESSELS a-f, variations of the interlocking fish fret; a, b, c, d, f, from vessel exteriors; e, from interior of a shallow bowl; g, a sigmoid design used as a border

other is almost universal. Variations in this border are shown on Figure 6 and Plate 6h-l. Body decoration may be analyzed as follows: 1. Bifurcated triangles or fishheads (Plate 7C> 7f"i) 2. Simple triangular fishheads interlocked (Plate 7e, 7j, 7 P ) 3. Serrations and enclosed dots (Plate 7k, 7I, 7m) 4. Rhomboid fishheads 5. Nested triangles or chevrons ("reduced faces?") (Plate 7n, 7q) B 6. Crosshatching (Plate 7s) 7. Parallel lines of alternating colors 8. Step designs 9. Crossbar design (Plate 7 t. 7U) T h e first three design units are the most common, although there are many differences in their manner of execution. Number 3 is s The simplest of these were later reclassified as Threecolor "Intermediate."

often used in conjunction with Numbers 1, 2, or 4. Numbers 6 to 9 are relatively rare. Number 5 often represents a fishhead with the triangle divided by lines or enclosing dots. Vessel forms can be listed after an examination of sherds of the Interlocking type. 1. Low bowl, slightly incurved sides, no lip

(Fig- 5 n ) 2. Plate with gradually upcurved rim (Fig.

5S) 3. Large mammiform jar with neck (Plate 4. 5. 6. 7. 8.

7r) Large olla (a globular bowl or jar) with straight or flared neck (Fig. 5j) Beaker bowl (Fig. 4I1) Beaker bowl or deep bowl with a thickened rims often inturned (Fig. 5f- 5l)rim; . Small jar with straight or slightly outflared neck Small jar with bulbous or swollen neck (Fig- 5^)

POTTERY FROM CERRO DE TRINIDAD TYPE:

THREE-COLOR CURVILINEAR NATURALISTIC

AND

8

Sherds.—Four fifths of the sherds of this type are red-slipped and decorated with broad, sweeping, curvilinear bands of black outlined with white (Plate 8A1). When white bordering lines are not used, small white spots or bars dot the black bands at close intervals. None of the sherds gives the complete pattern of the design, which is apparently large and boldly executed. One element, besides the common ones described, is a black circle filled with white from which black bands radiate in four directions (Plate 8An). A finer, more carefully done design is rectilinear and relatively naturalistic. It is the fish design standing free, and not the typical highly stylized fish conception (Plate 8Am). In only one instance is the interlocking fish fret used as a border. Rim sherds are too few to permit discussion of vessel form. TYPE:

DOT

7

Sherds.—This variant type of the Interlocking style is characterized by the use of small painted dots. These occur in all three colors, red, white, and black. They are used as fillers (Plate 8Ab, 8Ae-g) for line and band designs and also to form the entire design. In nearly every instance where the sherd considered is a rim piece the interlocking fish fret is used as a border below the rim. Dots are commonly associated with inscribed triangle or chevron motifs where they are used as a filler between the bands of the design. Some of these are clearly parts of the interlocking fish pattern (Plate 8Ac). On some sherds, areas of white or black dots on a red ground compose the fish designs (Plate 8Aa). Beaker bowls, shallow open bowls, and an

occasional bowl with incurved walls, with or without a neck, are the forms noted. TYPE:

CIRCLE

8

Sherds.—Small painted circles used as body decoration on three-color pottery denote this type. T h e circles, which average 9 millimeters in diameter, are usually white, red, or orange on a dark-red to blackish background. They are arranged in fields or rows on the body of the vessel and are often interspersed with small red or white dots (Plate 8Ah, 8Ai). Interlocking style affinities are clearly implied by the use of the interlocking fish fret border. On three sherds, the circle decoration on the body is combined with a row of horizontally placed sigmoid figures running around the border. These are the only occurrences of the sigmoid unit on types related to the Interlocking style. One sherd has red and black circles and white dots on an orange field (Plate 8Ad). T h e polished surface and decorative arrangement of this fragment suggests the Early Lima ware of the Rimac valley. Variations of the beaker bowl rim are the only indications of vessel form among the sherds of this type. TYPE:

X-MARK

Sherds.—There are less than a dozen sherds of the X-mark Type. T h e motif is a border of X marks running around the top of the vessel. All of the sherds are rim pieces, and it seems certain that the design arrangement is a rim border only. Three colors are used in decoration. As a rule, white X marks occur on a red or black background (Plate 8Aj, 8Ak). Once black X marks appear on a red ground. Rim profiles suggest low bowls with slightly incurved sides as the usual form associated with the design.

THREE-COLOR "INTERMEDIATE" Vessels.—Two vessels of black and white on red are so different from those of the Interlock6 T h i s type conforms closely to of the Interlocking. See Kroeber, 7 T h i s type designates a part of under Kroeber's " a " division of Kroeber 1926b, p. 2 8 1 .

Kroeber's " b " division 1926b, pp. 2 8 1 - 8 2 . the elements included the Interlocking. See

ing style that they are considered separately. In form and simplicity of design they resemble the White-on-red style; in use of three colors they are more like Interlocking. T h e i r grave 8 See also division " a , " Interlocking style. Kroeber, 1926b, p. 2 8 1 .

POTTERY FROM CERRO DE TRINIDAD and stratigraphic associations strengthen this position of intermediacy. Both specimens are low bowls with incurved walls. Both are without lips and handles. T h e larger (Plate 6e), which is 22 centimeters in diameter and 11 centimeters high, has the pronounced median ridge at the point of maximum diameter, a feature characteristic of White-on-red bowls. T h e rim, in profile, is thinned and pointed. T h e interior is unslipped. T h e exterior has a pale orange slip which has been fire clouded. T h e surface has been well smoothed on the outside, and the ware is identical with that of the White-on-red or Interlocking style. Decoration, which covers the full area between median ridge and mouth, is a series of white rectangles or panels. Each of these is separated by four vertical black lines and a row of black dots between the two middle lines. A single black line runs around the median ridge and connects with the vertical black lines. T h e second bowl has been badly damaged by fire, but it is clearly of the same type (Plate 6f). It measures only 10 centimeters in diameter and 5 centimeters in height. T h e design on the upper half is similar to that of the

larger vessel, although nested triangles, or "reduced faces," of black and white are used instead of rectangles and vertical lines. There is a black dot in each triangle. It is, perhaps, closer to the Interlocking type than the other. Sherds.—In the original scheme, possible sherds of Three-color "Intermediate" type were not classified separately but were counted with the Interlocking type. When the probable significance of the type was recognized, all of the Interlocking type sherds were reexamined to sort out "Intermediate" pieces. Nine sherds were found which seem to fit into this classification (Plate 7 o). All of these resemble in simplicity of design arrangement and color the two whole vessels just described. T h e vessel shape, where it was possible to determine it in the sherds, was also the low, incurved bowl with an unmodified lip. There was, however, considerable difficulty experienced in trying to distinguish in many cases between Interlocking type sherds with the nested triangle or chevron design and the Three-color "Intermediate" sherds. There is little doubt but that there is a relationship, and the types grade very easily from one to the other.

PLAIN RED WARES In sherd count, the Plain red wares make up the greater part of the material from the stratigraphic cuts. There are, however, few whole vessels in Plain red. T h e ones found with burials are all small pieces. It is likely that the huge Plain red ollas were never placed with the dead. Plain red has been divided into two types, both of which relate to the White-onred and Interlocking styles. This is clear from vessel forms, rim forms, paste and ware characteristics, and association. These two types, Smooth plain red and Coarse plain red, follow similar divisions in the White-slipped and White-zoned types of the White-on-red style. Smooth plain red is medium-thin well-fired pottery which has a great deal of variation in color, ranging from orange to mottled brown. In small vessels it is no thicker than the White-on-red painted types; in extremely large ollas or jars the walls and bases average a few millimeters more. Coarse plain red, for sherd

classification, includes thick slipped and unslipped pottery which was not well smoothed and is coarse in paste and surface texture. It is red-fired and has heavy coarse temper, particles of grit which extrude through onto the smoothed surfaces. Walls are between .8 and 1.5 centimeters thick. As in the case of Smooth plain red, it is similar to the Whiteslipped (coarse) or the White-zoned (coarse) type of the White-on-red style. TYPE:

SMOOTH PLAIN RED

Vessels.—The six smooth plain red vessels may be classified according to form as follows: I A 1. I A 2. I A 3. II B 3.

Bowl, lipless, low, flaring (miniature) Bowl, lipless, incurved or vertical walls Bowl, lipless, low, incurved, 2 handles Jar, vertical neck, 1 round handle on neck

3 1 1 1

T h e three small miniature bowls are very similar (Plate 4j). All have been slipped and smoothed both inside and out. They average

POTTERY FROM CERRO DE TRINIDAD 9 centimeters in diameter at the orifice and 4.5 centimeters in height. B o t h lipless, incurved, low bowls are the same size: 13 centimeters in diameter and 8.5 centimeters in height. T h e one with handles is unique for the collection (Plate 4I1). It has a ridge or bevel encircling the bowl just below the lip. Projecting from this on opposite sides are two triangular, slightly upturned lugs. These have been pierced, vertically to the vessel, with small holes, perhaps for suspension. T h e vessel with the handle on the neck is a small jar or "pitcher" (Plate 4i), 14 centimeters high. T h e mouth is 3.5 centimeters in diameter. T h e handle extends to the lip from the body and is round and not straplike. T h e base is round. Sherds.—Because of the overwhelming number of sherds of the Smooth plain red type recovered from the site, it is possible to describe a very representative range of forms from the sherds. It will be noted that a number of these forms have been described for other types, particularly the White-on-red style. Numerically, the two most prominent forms are the plate with upturned or thickened rim and the incurved bowl with a folded or turned-up lip. For brevity in description, these and other forms, and their variations, are listed below. 1. 2. 3. 4. 5.

Plate with upturned rim (Fig. 5a) Plate with thickened rim (Fig. 5b) Shallow open bowl (Fig. 5I1) Deep open bowl (Fig. 5p) B o w l with slightly incurved sides (Fig.

5n) 6. B o w l with incurved sides (flattened globular), no lip (Fig. 5i) 7. Bowl with incurved sides (flattened globular), lipped (Fig. 5m; Plate 5AI) 8. Jar with straight inslanting sides 9. Jar with bulbous neck (Fig. 5k) 10. Jar with flared neck 11. Jar with small bottleneck 12. Large jar with high neck or collar (Plate 5Ak) Bases are usually rounded, but a circular flat base does occur. Horizontally placed strap handles and rounded nubs are common features. Modeled ridges on vessels and small pointed spouts are interesting but uncommon items.

i5B

TYPE: COARSE PLAIN RED

Vessels.—The three vessels of this type can be summarized according to shape as follows: II B 3. Jar, vertical neck, 1 round handle (miniature) Huge ollas with handles

1 2

T h e miniature jar or "pitcher" (Plate 4k) is very carelessly made and may have been a toy. It is 8 centimeters high, with a round base and a round handle connecting lip and body. T w o huge olla forms were not found with burials but as sherds of vessels broken in situ. Both of these have large, flattened globular bodies and short necks. T h e larger, though fragmentary, can be estimated at 60 centimeters in diameter, with an orifice of 18 centimeters. T h i s one has an outflaring neck which tapers into a pointed lip. T h e strap handles, which are horizontally placed, are 20 centimeters long and over 8 centimeters wide. T h e other olla is a little smaller and has a short, straight neck. Its handles are relatively small, comparable to those of ordinary-sized vessels, and are placed about 10 centimeters below the mouth. Sherds.—The rim sherds of this type show generic shape resemblance to the form of the Smooth plain red but, as would be expected, there is a heavy crudeness about rim and lip treatment. T h e incurved bowl with a lip is one of the most usual forms, but more often the lip is a thick straight collar or neck, 1.5 to 2 centimeters high (Fig. 51-). H u g e olla forms, which are essentially the same as the smaller incurved bowls in body outline, have high straight necks or flaring necks. Some of these have a thickening on the lip exterior. T h e r e is also a big jar form with a straight neck comparable to a similar form in the Smooth plain red. Small jars with straight or outflared necks and deep open bowls (Fig. 5q) are less common. A number of ornamental features are like those of the White-slipped (coarse) type. T h e same ridges and knobs with indentations are applied to the vessel surface. Besides both massive and small strap handles, there are small n u b projections, undecorated, which may be functional.

154

POTTERY FROM CERRO DE TRINIDAD MISCELLANEOUS D E C O R A T E D

Sherds.—Most of the sherds of this category are those pieces which were too fragmentary or too eroded to classify under any of the listed styles or types. It is likely that the majority are of the Interlocking style. Some may be White-on-red. Those few sherds that stand out from this group deserve separate discussion. One is a three-color rim sherd from a bowl with very slightly incurved walls and a tapering and pointed lip. T h e decoration is a border band of white outlined in black. Set in the band are black red-filled rectangles connected to the top black band of the border (Fig. 7a). T w o of these rectangles have small black dots in the center. T h i s sherd came from Pit VI, Level 5 (1.25-1.50 meters), an Interlocking period

level. Another rim sherd from a very similar bowl shape has a white paint design on a red surface. T h i s consists of an interlocking fish fret border, done entirely in white, and what may be a rectilinear and naturalistic white design on the body of the bowl (Fig. 7e). T h e r e is no black paint whatsoever. It was the only sherd of this variation found at the site. Provenience is Pit V, Level 1 (0-.50 meter). A third piece, also a rim sherd from a bowl of the same general shape as the other two, has a white-slipped interior and an orange-slipped exterior. T h e design on the exterior is a cat of the Recuay genre done in deep, purplish black (Plate 8B0). It came from Pit VI, Level 1, 0-.50 meter deep.

I

1

WHITE

pana black preiaa owce-YRUW FIC. 7. VARIOUS DECORATED TYPES FROM SITE E AND FROM BAÑOS DE BOZA a, miscellaneous panel design from Site E; b, c, Negative painted sherds from Site E; d, Negative painted sherd from Baños de Boza; e, Interlocking style design in White-on-red painting from Site E; f, the only Interlocking sherd from Baños de Boza. Size: % natural size

POTTERY FROM CERRO DE TRINIDAD

155

INCISED A N D P U N C T A T E D Sherds.—Few of the incised or punctated sherds recovered from Site E are large enough or distinct enough to give an adequate pattern of the design. There are three fragments, two from mixed Interlocking and White-on-red levels and one from a deep, pure White-on-red level (Plate 8Bu, 8Bv), that are indeterminate bits of animal or human effigies. All three show enough painted surface to suggest relationship to the White-on-red style. T h e one from the deep level has long, triangular punctate marks. One of the others has a single hollow-reed with enclosed dot impression, and the last shows a zigzag line of incision. Five sherds are decorated with rows or fields of circular hollow-reed punctations. T h e circular punctations vary from 4 to 8 millimeters

in diameter. On two rim sherds they have been placed in rows around the base of a short vessel collar (Plate 8Bs, 8Bt). All of these punctated sherds came from the Interlocking or Intermediate levels. T w o other sherds have combination smalldot or jabbed punctations and hollow-reed punctations. One of these is a short spout with a flared lip. T h e decoration is on a bulge or swelling around the base of the spout. Both sherds are from Interlocking period levels. T w o small sherds, one with fingernail (?) punctations and the other with incised lines, are also from upper levels. One tiny sherd from a deep White-on-red level has a depressed area or band of little dot punctations bordered by a narrow incised line.

NEGATIVE PAINTED STYLE Sherds.—Six sherds of the Negative painted style came from Site E. T h e ware is red-fired, is tempered with medium-fine sand and grit, and, judging from the few sherds at hand, averages 6 millimeters in thickness. T h e original slip, on the exterior, is a light red. A very dark red paint or slip has been applied secondarily, leaving the decoration in light red. Decorative elements are tear-shaped dots,

round dots, and bars. Bars and dots occur together on two pieces (Fig. 7b, 7c). There is one spout sherd which has a single light-red stripe near the top as the only decoration. T h e spout, which is almost intact, measures 6.5 centimeters in length and 2.8 centimeters in basal diameter and slopes inward towards the opening. A strap handle has been broken from one side near the base.

POLISHED B L A C K (OR " B U C C H E R O " ) W A R E Sherds.—True black ware was very scarce at Site E. A great many sherds and a few vessels were secondarily fire-blackened, but only a few sherds could be considered as smoke-blackened polished pottery. All sherds assigned to this type are black and polished on one or both surfaces. In cross section, or on the interior surface if unpolished, they are gray. Tempering is with medium to fine sand and grit. Thickness varies from 3 to 7 millimeters.

T h e incurved bowl with median ridge running around the vessel, a common form of the Smooth plain red and White-on-red types, is present in the black ware. More common forms are bowls with slightly incurved sides, jars with necks, open bowls with outslanted sides, and the flat plate described for the Smooth plain red and White-on-red types. There are no sherds of this type that specifically suggest the early or late black ware of the north.

LATE CHANCAY STYLE T h e Late Chancay style was not numerically important at Site E. A few sherds were found in the upper .50-1.00 meter of most of the stratigraphie pits. T h e style and its types have

been described elsewhere from collections in the Chancay valley (Kroeber, 1926b, pp. 26770). On the north side of the Chancay valley are two important sites of this culture. Not

POTTERY

156

FROM

CERRO

more than a kilometer to the south of Site E is the large Late Chancay site of Ollada. Site A, excavated by Uhle, on the north side of the Cerro de Trinidad, yielded grave pottery only of this style. TYPE: CHANCAY BLACK-ON-WHITE

Sherds.—The ware is medium to thin, ranging from 4 to 7 millimeters in thickness. T h e paste is fine or medium-coarse, depending upon the specimen. It has been tempered with sand or fine grit and is not unlike the finer ware of the White-on-red or Interlocking styles. Paste color is either a pale yellow or a dark orange-red. T o judge from the small collection at Site E there does not seem to be much intergradation between these two standards. T h e primary color or slip is usually considered white, but it is not the dead white of the White-on-red style. T h e important variations are: a pale yellow, verging on a limegreen; and a buff. It is a characteristic slip, easily recognized, especially out of its present associations with Interlocking and White-onred. T h e decoration is black and is applied to the exterior. Panels filled with stylized bird figures (Plate 8By) or with step designs are a common motif; zigzag or wavy lines between

DE

TRINIDAD

straight lines are another (Plate 8Bz). White circles with a black dot in the center often occur on a black field. Series of small rectangles or interlocking triangles (Plate 8Bw) sometimes run horizontally around the body of the vessel. T h e conceptions are very formalized; and the execution is competent and assured though not inspired. T h e predominant vessel form indicated by the sherds is a low, small bowl with incurved sides. One fragment suggests a large jar with a constricted orifice. Small strap handles and tiny, probably mold-made, animal adornos are additional features. TYPE: CHANCAY WHITE-SLIPPED

Sherds.—This type is closely related to the Black-on-white type, lacking only the black surface decoration. T h e r e are the same variations in the shades of white used for a slip. T h e most common form is a shallow, open bowl with a definite ring base. Sometimes a cross or a field of dots is left as a relief design within the raised ring of the base. T h e bowl form has the slip applied to both sides. T a l l tumbler forms with outcurved sides are also recorded. Besides the strap handles, an intertwined handle, simulating a rope (Plate 8Bzz), occasionally appears.

TABLE I PROVENIENCE OF P O T T E R Y Pit Location in Pit Plate 1, a-j, all from Pit VII, Burials 1 and 2 Plate 2 a b c d e f g h i

VII VII VI VI VI VII VII VII VII

Plate 3 a VII b [Surface] c IV d II e V £ V

Level 6, Floor F Burials 1 and 2 Burial 2 Burial 2 Burial 4 Burials 1 and 2 Level 8 Burials 1 and 2 Burials 1 and 2

Level 13 Level Level Level Level

3 2 1 11

ILLUSTRATED

Plate 3 ( c o n t i n u e d ) Pit g h i

j k

1 m n

0 P q

r s t u v w X

VII IV V II VII VII VII II VI VII VII VI

VI VII IV II IV VI

Location in Pit Level 8 Level 2 Level 7 Level 1 Level 18 Level 15 Level 16 Level 3 Level 12 Levels 15 and 12 Level 13 Level 12 Level 3 Level 14 Level 5 Level 3 Level 7 Level 10

POTTERY

FROM

CERRO

DE

TRINIDAD

TABLE I (Continued) P R O V E N I E N C E OF P O T T E R Y I L L U S T R A T E D Plate 3 (continued) Location in Pit Pit Level 6 VII y z Level 6 VII zz VII Level 3 Level 4 III yy Plate 4 a b c d e f g h i j k

VII VII VII VII VI VI VII II VII VII II

Plate 5A a V b VII c III d IV e VII f II IV g h IV i VII VII j k IV 1 II Plate 5B (Baños de m IV n III 0 I IV P I q r IV s I t I u I V IV w IV X IV IV y Plate 6 a b c

d e

IV IV VII VII VI

Burials 1 and 2 Burials 1 and 2 Burials 1 and 2 Burials 1 and 2 Burial 2 Burial 1 Burials 1 and 2 associated with child burial Burials 1 and 2 Burials 1 and 2 associated with child burial Level 11 Level 18 Level 4 Level 16 Level 10 Level 1 Level 5 Level 9 Level 20 Level 12 Level 6 Level 2

Plate 6 (continued) Pit £ VI VII g h VI i VI V j k VI 1 IV Plate 7 a b c d e f g h i j k 1 m n 0 P q r s t u

IV III VI V VI VII V IV IV IV VI V VI VII II IV IV V VI VI VII

Location in Pit Burial 1 on floor E Level 1 Level 1 Level 7 Level 6 Level 4 Level 2 Level 3 Level 7 Level 1 Level 6 Level 1 Level 4 Level 1 ext. (about 1.50 meters). Level 7 Level 6 Level 1 Level 1 Level 8 Level 1 Level 2 ext. (about 1.50 meters) Level 11 Level 4 Level 2 Level 1

Boza) Level 9 Level 1 Level 1 Level 6 Level 1 Level 2 Level 3 Level 3 Level 4 Level 5 Level 1 Level 6 Level 9

Plate 8A a VII b VI c VI d V e VII f II VII g h VI i II VII j k IV 1 IV m III n

Level 4 Level 9 Level 3 Level 11 Level 1 Level 2 Level 4 Level 7 Level 2 Level 2 Level 7 Level 5 Level 2 Surface

east ext., floor E east ext., floor E north ext., floor E north ext., floor E Burial 1

Plate 8B 0 VI VI P VII q r VII

Level Level Level Level

1 1 17 17

POTTERY

158

FROM

CERRO

DE

TRINIDAD

TABLE I (Continued.) P R O V E N I E N C E OF P O T T E R Y I L L U S T R A T E D

s t u V w X V / z zz

Pit VII II IV IV IV VI VI VII IV

Figure 6 a VII b VI

Plate 8B (continued) Location in Pit Level i Level l Level l Level io Level s Level l Level l Level l Level g Level l Level 9

Figure 6 (continued) Location in Pit Pit Level 2 c IV Level 3 d III e VI Level 6 f Level 5 VI Level 2 VI g Figure 7 a b c d e f

VI VII IV IV V IV

Level 5 Level 13 Level 9 (Baños de Boza) Level 1 Level 1 (Baños de Boza) Level 1

The Artifacts from Cerro de Trinidad, Site E

T

a small but interesting collection of artifacts, other than the pottery vessels and sherds, from the refuse and burials at Site E. These are discussed below under headings based on material of manufacture or obvious function. Wherever possible, cultural affiliations, or probable affiliations, with the White-on-red, Intermediate, or Interlocking periods, as these are defined in the next section, have been given. T h e association of artifacts in graves containing only White-on-red pottery was, as usual, considered good evidence of identification as to cultural period. It was also possible to make period HERE IS

assignments of artifacts with a good degree of accuracy in cases where stratigraphie position and typology concurred, as in the instances of certain White-on-red figurine or adorno fragments found in levels showing only White-onred sherds. Where either one of the factors of typology or significant stratigraphie provenience was lacking the degree of accuracy is lessened. Where neither factor was present, as, for example, in the case of those artifacts not distinguishable as to style and found in mixed levels, it was impossible to make any sort of judgment.

U N F I R E D C L A Y VESSELS T w o complete unfired clay vessels and a fragment of another were recovered from Levels 10 and 11 (2.50-3.00 meters) of Pit VII. These objects range from 9 to 10.5 centimeters in diameter and are 10 centimeters high (Fig. 8 o). T h e walls of the vessels are 2.5 centimeters thick and taper slightly to a blunt lip. Lids or tops have been sealed over the mouths

with a black gummy substance, now dry and hard. T h e lids are of the same clay as the vessels. They are roughly disk-shaped, 7 centimeters in diameter and 1 centimeter in thickness. One of the vessels which was opened contained a few tiny pebbles and some strawlike vegetal particles. T h e provenience is in the Intermediate period rubbish zone.

PANPIPES There are eight pieces of pottery panpipes of which two are slipped or painted. One of these is decorated in red, white, and black on both sides and suggests the Interlocking style (Fig. 8n). Its provenience is Pit IV, Level 5, 2.00-2.50 meters deep, the Intermediate period refuse zone. T h e fragment is only 6 centimeters long and is a corner of the instrument. Only a portion of one of the cylinders is present. T h e other painted fragment came from the south extension of Pit IV at a depth of approximately one meter. It is a complete cylinder with half of another conjoined. T h e

length is 9 centimeters, and the diameter of the complete cylinder is 1.7 centimeters. It has a brown-slipped, polished ground with a small white circle painted on one side and evidences of bright red paint on the other (Fig. 8m). T h e provenience is in the Interlocking period. In Pit VI, Level 6, 1.50-1.75 meters deep, there were three unslipped fragments of a pipe or pipes. T w o of these have three joined cylinders. T h e cylinders average one centimeter in diameter. In Pit VI, adjoining, a similar panpipe fragment came out of Level 6. This piece shows four cylinders and is the base

m

—«•aaa.fM.i^rmrih «t..» Fig. 8. ARTIFACTS FROM SITE E

a, jadeite(?) spindle whorl set with shell; b, c, lapis lazuli whorls; d, incised pottery spindle whorl; g, stone whorl; e, f, h, j, shell beads; i, bone atlatl hook; k, m, n, fragments of panpipes; 1, wooden "plug"; o, sealed, unfired clay vessel. Size: a-j, natural size; k-n, % natural size; o, % natural size

ARTIFACTS FROM CERRO DE TRINIDAD or bottom end of the instrument (Fig. 8k). T h e base slants at a 45-degree angle to the long axis of the cylinders. Along the bottom, on both sides, are impressions of what appear to be textiles. In Pit II, Level 5, 1.25-1.50 FIGURINES, ADORNOS,

There are six other anthropomorphic representations in the collection. T w o of these are heads, perhaps broken from figurines. In cross section they are flat and taper toward the top of the head. T h e faces are broad, with long eyes formed by incised slits set with tiny pellets of clay. These clay pellets have been punched to form pupils. T h e nose is high and arched. Nostrils are made by punctations, and the end of the nose appears to be cut off or mutilated. T h e more complete of the two measures 6 centimeters across the face (Fig. gc). It came from Pit VII, Level 14 (3-50-3.75 meters), a White-on-red period level. T h e more fragmentary specimen (Fig. gb) came from the same pit, Level 1 (0-.50 meter). There are two other very small sherds suggesting heads of the same style. These come from Pit II, Level 1 (0-.50 meter) and Pit IV, Level 1 (0-.50 meter). A third human face is very similar to the first two, with inset and 9

meters deep, a single cylinder with basal portion attached was found. This one had the same textilelike impressions along the bottom. A single cylinder fragment was found in Pit V, Level 1 (0-.50 meter).

A N D MISCELLANEOUS P O T T E R Y

Under this heading are included various human and animal representations in ceramics. Some are obviously parts of pottery vessels; others may have been broken from vessels or may be parts of individual figurines; and a few are toys. One of the most striking pieces is the effigy of a man, or an owl, or perhaps a combination of these two (Fig. ga). It is a hollow fragment with only the head and part of the shoulders intact and may have been broken from a vessel.9 Total height is 6.5 centimeters and breadth is 7 centimeters. T h e nose is beaklike, merging with the top of the head; the eyes have been formed by small bits of clay placed in incised slots; the mouth is a single incision. Painted decoration of white on a red slip is distinctly White-on-red in style, and the provenience is from the White-on-red rubbish of Pit VII. This specimen has also been counted as a White-decorated sherd in the stratigraphic count for Level 17 of Pit VII.

Kroeber, 1926b, Plate gog, suggests such a vessel.

161

OBJECTS

punched eyes and mutilated nose. This one, however, is on a sherd from a White-slipped bowl or jar (Fig. ge). T h e face is on the vessel exterior just below the rim. T h e sherd came from Pit VI, Level 8 (2.00-2.25 meters). Facial mutilation is encountered again on an adorno of a quite different style. This specimen is a small (3.5 centimeters high), solid, mold-made head of a man which appears to have been broken from a vessel (Fig. gf). It was at one time white-slipped. T h e face is well proportioned and realistic. A headdress is shown by a headband and two ridges along the top of the head. Eyes are bulging but not punctated. T w o holes, one in each side of the head, suggest either mutilation or ornamentation. T h e end of the nose has probably been clipped, and there is a deep gash across the lip (not accidental breaking). This head came from Pit III, Level 2 (.50-1.00 meter). A painted human face came from Pit II, Level 1, 0-.50 meter (Fig. gi). This may have been either an adorno or a figurine. About three fourths of the face remains, showing one eye, one eyebrow, the mouth, and the area of the nose. T h e nose was probably appliqué and has been destroyed. T h e face is painted white with the features done in black. Along one side of the face there is some red paint. A second painted specimen is complete and looks like a toy (Fig. gh). It was found with child Burial 1 at the bottom of Pit VI, a grave of the Intermediate period. T h e modeling is extremely crude. Breasts, arms, feet, and head are represented. T h e entire figure is whiteslipped and decorated with black dots. T w o of these dots serve for eyes. Fastened to the back is a little loop handle. Total height is 5.2 centimeters. A third human effigy fragment is a hand which came from Pit IV, Level g (4.00-4.50 meters), White-on-red period. It is red-slipped and decorated with white. This piece was counted as a White-decorated sherd

FIG. 9. ADORNOS

O R FIGURINES F R O M SITE E

a, White-on-red owl or demon head; b, c, fragments of human faces; d, dog (?); e, human face from a White-slipped sherd; f, solid human head adorno; g, monkey (?); h, anthropomorphic figure with handle; i, painted human face; j, a bee (?), complete figure; k, 1, animals (?). Size: all % natural size

ARTIFACTS FROM CERRO DE TRINIDAD in the stratigraphic count. Full length is 3.5 centimeters. Animal figures are even more varied than the human representations. T h e r e are six of these, including a guinea pig or fish (?), a bee (?) (Fig. gj), a dog (?) (Fig. gd), a monkey (?) (Fig. gg), and two even more doubtful species (Fig. 9k, gl). T h e guinea pig or fish is a hollow head portion and may be a part of a white-slipped vessel. Length is 6.5 centimeters. T h e eyes and mouth are boldly but realistically imitated. T h e specimen came from Pit III, Level 4 (1.50-2.00 meters). T h e bee is complete. It was found with the child burial (Burial 1) in Pit VI. T h e effigy is hollow and may be a whistle. T h e single opening is a small hole in the top of the head. Eyes, nostrils, and mouth have been made by incising and modeling. Additional modeled features, perhaps a handle, have been broken from the back and head; a white slip once covered the entire artifact; and black stripes were lined across the back and wings. 10 Length is 7 centimeters and height 3.5 centimeters. T h e monkey is a small adorno, red-slipped, with white decoration. It was taken from Pit VII, Level 17 (4.25-4.50 meters) and has been counted as a White-decorated sherd. It is 4.5 centimeters long and about 2 centimeters high. Little legs have been modeled on each

163

side. Eyes, nostrils, and mouth are done by incision and punctation. T h e dog, or possibly a bear, is a hollow fragment, perhaps from a vessel. Eyes and nostrils are punctated, and teeth and mouth have been modeled along both sides of a pointed snout. Size is 5 by 5 centimeters. It was found in Pit III, Level 3 (1.00-1.50 meters). T h e two more doubtfully identifiable animal heads are probably adornos from White-on-red vessels. One comes from Pit II, Level 1 (0-.50 meter) and the other from Pit V, Level 8 (2.00-2.25 meters). T h e faces and features are combinations of modeling and incising. Bears or dogs may be indicated. Both have been red-slipped and decorated with white, although the paint is now badly flaked. One has been pierced through the ears and has a blunted or mutilated nose. Each measures about 3 by 4 centimeters. T h e r e are two other fragments which may be bodies of men or animal figures. Both are hollow. One is from Pit IV, Level 8 (3.504.00 meters), the other from Level 9 of the same pit. One toy pot of coarse unslipped ware came from Pit VI, Level 5 (1.25-1.50 meters), an Interlocking period provenience. It is 4 centimeters in diameter and 2.2 centimeters high.

SPINDLE W H O R L S Eight artifacts are considered under this heading. T h e r e are three disk whorls, two of them made of white-slipped potsherds. T h e other five whorls are spheroid and made of stone or pottery. T h e smaller of the potsherd whorls is 2.5 centimeters in diameter and 1 centimeter thick. It has a central perforation 8 centimeters in diameter. Its provenience is Pit VI, Level 5 (1.25-1.50 meters). T h e larger sherd whorl is curious in that it has two perforations, one in the exact center and the other about 1 centimeter off the center. T h e disk is 5 centimeters in diameter and 5 centimeters thick. It was found in Pit VI, Level 1 (0-.50 meter). T h e remaining disk whorl is of gourd. It measures 4.5 centimeters in diam10 Kroeber, 1926b, Plate 90b, shows an effigy vessel decorated in similar manner.

eter and 4 centimeters in thickness. It came from Pit II, Level 3 (1.00-1.50 meters). A l l of these come from Interlocking period levels. It is possible that some of the spheroid whorls may be beads. In two instances the holes seem rather small for a spindle shaft. Four of the five are of stone; the other is of pottery. T h r e e of the handsomest stone whorls were taken from Burial 2 at the bottom of Pit VII. One of these was manufactured of a light green stone, perhaps a jadeite (Fig. 8a). It is 1.6 centimeters in diameter and 1.3 centimeters high. T h e sides are rounded and inlaid with small circular pieces of white and red shell. These are set in three rows of eight pieces each, alternating in color, around the body of the whorl. T h e inlays have been accomplished by drilling small holes in the

164

ARTIFACTS FROM CERRO DE TRINIDAD

stone and fastening the shell pieces in these with pitch or gum. T h e other two specimens found in conjunction with the inlaid whorl are of polished lapis lazuli. One is a flattened spheroid (Fig. 8b), 1.6 centimeters in diameter and 1.3 centimeters high. T h e second is also 1.6 centimeters in diameter but is 2 centimeters in height. It has a rounded "pot-collar" at one end (Fig. 8c). All three were found in a White-on-red period grave. Another stone whorl came from Pit VI, Level 4 (1.00-1.25 meters). This is made of

honey-colored quartzite. It is 1.3 centimeters in diameter and 1.1 centimeters in height. One end is rounded, the other flat (Fig. 8g). T h e provenience is of the Interlocking period. A pottery whorl, or bead, came from Pit IV, Level 8 (3.50-4.00 meters). It is spheroid with a diameter of 1.8 centimeters and a height of 1.6 centimeters. An incised ornamentation, covering most of the surface, consists of two concentric circles with two three-line scrolls attached (Fig. 8d). Association is probably with the White-on-red period.

BONE ARTIFACTS A small bone atlatl hook (Fig. 8i) came from Pit VI, Level 3 (.75-1.00 meter). It has an over-all length of 2.2 centimeters and was carved from a single piece of bone. T h e base is straight, with a delicate projection at one end and a prepared nub for hafting at the other end. T h e notch or hook is deep. T h e provenience is of the Interlocking period. A bone awl, made from the leg bone of a

deer, came from Pit II, Level 1 (0-.50 meter). It is 16 centimeters long, and was made by fracturing the bone longitudinally. A blunt triangular point, slightly worn, indicates its use as a tool. An adult human femur with the proximal end broken away was found in Pit II, Level 4 (1.50-2.0 meters). A projection of the broken end has been worn smooth with use.

WOODEN A R T I F A C T S A wooden spear or throwing-stick projectile came from the bottom of Pit VII in probable association with one of the burials at that depth. It is in three pieces with a total length of 61 centimeters. T h e diameter is 1.5 centimeters. Within 20 centimeters of the point it commences to taper. T h e wood is hard, is still in excellent condition, and was probably fire-hardened. Association is with the Whiteon-red culture. Five top-shaped objects of wood were found in the rubbish of the several pits. These vary from 6 to 4 centimeters in length and 3.5 to 2 centimeters in maximum diameter. Four

are very similar as to form (Fig. 81); the fifth has a ragged stubble on the broad end and may be an unfinished specimen. Four were recovered in the Pit V I I excavations: one from the north extension; one from Level 4 (1.00-1.25 meters); one from Level 5 ( 1 . 2 5 1.50 meters); and one from Level 2 (.50-.75 meter). T h e other came from Pit II, Level 2 ( . 5 0 - 7 5 meter). Their use is unknown; perhaps they served as plugs or stoppers. As to cultural association, their consistent appearance in upper levels and absence in lower ones suggests that they are an Interlocking period artifact type.

OBJECTS OF GROUND STONE A circular disk of light-colored stone was found in Pit VII, Level 14 (3.50-3.75 meters). Dimensions are 10 centimeters in diameter and 1 centimeter in thickness. It was ground into general outline, but it is not well

smoothed. T h e edge is beveled as though to form a lid. Cultural association is with the White-on-red style, on the basis of stratigraphic position. T w o small, smoothed stone balls, which

ARTIFACTS FROM CERRO DE TRINIDAD may have been used for pottery smoothers, came from Pit I, Level 1 (0-.50 meter) and Pit II, Level 1 (0.-.50 meter). A crude circular stone, flat on one side and rounded on the other, may have been used as a hammer. It is 16 centimeters in diameter and 5.5 to 6.5 centimeters thick. Pecking or hammer marks show along the edge. None of the surfaces shows grinding or smoothing. Its provenience is Pit II, Level 1 (0-.50 meter). Two fragmentary ground and well-

165

smoothed stone pieces were probably manos. One is 1 1 centimeters long, 6.5 centimeters wide, and 3.5 centimeters thick. One end is blunted from pecking or hammering; the other end is broken off. In cross section, it is plano-convex. Both the flat and rounded surfaces are well ground down. This specimen came from Pit III, Level 2 (.50-1.00 meter). T h e other specimen is a small fragment of a larger artifact; it came from the same pit and level.

CLOTH Around the skull of Burial 2, Pit V I I , a very ample cotton cloth headdress was wound. T h e original color may have been white or a light neutral shade. T h e fragments are now

brown. These show a fabric of medium-fine thread woven into a fine mesh. T h e weave is simple over-and-under technique. Warp and weft are of thread of the same size.

METAL Over the face of Burial 2, Pit VII, with a number of fragments clinging to the cloth headdress mentioned above, was a thinly beaten mask of copper and gold. This mask, or face cover, had crumbled to pieces. Most

of the pieces, which are heavily corroded, are copper. A few thin sheets of gold, or a goldcopper alloy, were lying over the nasal bones. These were the only particles of metal obtained in the excavations.

O R N A M E N T S OF S H E L L Most of the shell ornaments, which are beads, are made from a large red and white bivalve, the Spondylus pictorum. T h i s shell is common in the waters of the Pacific coast to the north, and the southern extent of its present-day distribution is the Gulf of Guayaquil in Ecuador. Presumably the shells were a trade material brought into the valley of the Chancay. One shell, virtually complete, was found in association with Burial 2, Pit V I I . This one, which is typical as to size, is 1 1 . 5 by 1 1 . 5 centimeters. T h e natural ridges on the outer surface of the shell have been ground down and smoothed. A great number of beads of all sizes, made from the Spondylus pictorum, formed a necklace for Burial 2, Pit VII. There are four different types of these beads. T h e largest type, of which there were three in the necklace, are flat, rectangular beads, about 5 centimeters long, almost 2 centimeters wide, and .6 centimeters in thick-

ness (Fig. 8e). A single hole has been drilled through the long axis of each. A mediumsized, rectangular type was more abundant, there being eighteen of these. They vary from 3 to 1.3 centimeters in length and from 1.8 to .6 in width and are .4 centimeter thick (Fig. 8h). Lengthwise, the ends are worked down on both sides to an edge. They are all drilled through the short axis with two holes. Besides the rectangular types, there are two sizes of round, crudely shaped disk beads. T h e larger of these (Fig. 8f) is 1 to .7 centimeter in diameter and .5 to .2 centimeter in thickness. All are perforated with a central hole. There were forty-eight of these included in the necklace. T h e smaller circular or disk beads are very imperfectly shaped and are less than .5 centimeter in diameter. Approximately two hundred of these were found with the others near the skull of the burial. By the wrist of the same burial, seventeen

i66

ARTIFACTS

FROM

CERRO

other beads of the same type of shell were found. Fourteen of these are of the smaller rectangular shape with a double perforation through the short axis of the bead. T h e other three are quite unique. T h e y are 4 centimeters in length, .5 centimeter wide, and .3 centimeter thick. Each has been drilled through the thinnest dimension with ten perforations (Fig. 8j). In a White-on-red style vessel, also with Burial 2, Pit VII, a few beads of the same

DE

TRINIDAD

type of shell were found. There are one twoholed rectangular bead, one large disk bead, and thirteen small disk beads in this lot. A l l of these beads of shell were from a White-onred period grave. T w o simple beads from the child burial (Burial 1) in Pit V I are of a small, spiral type of shell with the top cut off to permit suspension. These two are of a size, 3.5 centimeters in length and 2 centimeters in diameter.

TABLE 2 PROVENIENCE OF A R T I F A C T S ILLUSTRATED Figure 8 Pit a, b, c VII d IV e , f , h VII VI g VI i VII j V k II 1 m IV n IV VII 0

Figure 9 Location in pit Burial 2 Level 8 Burial 2 (necklace) Level 4 Level 3 Burial 2 (bracelet) Level 6 Level 2 ext. (about 1.00 meter) Level 5 Level 11

a b c d e f g h i j k 1

Pit VII VII VII III VI III VII VI II VI II V

Location in pit Level 17 Level 1 Level 14 Level 3 Level 8 Level 2 Level 17 Burial 1 Level 1 Burial 1 Level l Level 8

The Stratigraphy at Cerro de Trinidad, Site E POTTERY AND STRATIGRAPHY

T

S T R A T I G R A P H Y of pottery types in the refuse at Cerro de Trinidad, Site E, provides the primary datum for arranging the pottery, artifacts, structures, and burials in a culture sequence. In the presentation of this stratigraphy, the pottery types will be considered first. A chronological placement of structure, physical strata, and burials will follow. Finally, a definition of the three periods at the site will be outlined. T h e excavation at Pits IV and VII gives the best stratigraphic picture and is the principal evidence for chronology. T h e occurrences and percentages of occurrences per level of the White-decorated and Interlocking types in Tables 7 and 8 and the graphs (Figures 10 and 11) show the reciprocal relationship of these two types. In Pit IV the greatest conHE

0 / 2 3

4 5 6

centration of White-decorated sherds is between 3.50 and 5.00 meters deep, Levels 7 to 10 inclusive. Level 11 (5.00-5.50 meters), the lowest in the pit, is not expressed on the graph, as no decorated sherds, other than the White-slipped type, were found in this level. With the exception of Level 6 (2.50-3.00 meters), some White-decorated sherds were found in every level above the zone of greatest concentration. Both numerically and in percentages, however, the type is lessened in the upper half of the pit. Opposed to this, the Interlocking type has its maximum occurrence in Level 3 (1.00-1.50 meters) and above. Below Level 3 the type continues in diminished percentages to Level 8. No Interlocking sherds are found in Levels 9 or 10. Another type of the White-on-red style,

0 / 2 3 4 5 6

O I 2 3 4 5 6

HCTflH

-*H0° ,00.M

I 1

wHITE DECORATED

I •

W

TYPE

I

4 f k

m m

5 1

•

6F

H

250-100

7

V

J00-}50

3

^

^

^

^

WW/TE

^

^

^

^

Z a t 1 E D

^

^ ^ 1*0*00 mimERLDCKItlB

TYPE

•

8 9 10 H ^ V

TYPE

'-0(W50

2.00*50

W •

I

Wm I Fic. 10. POTTERY TYPES FOR PIT IV, SITE E (by Percentage)

&JOO-4JO 450-500

x 68

STRATIGRAPHY AT CERRO DE TRINIDAD

the White-zoned, including both the fineware and coarse-ware varieties, has been graphed for Pit IV. Although it appears in all ten levels, there is a suggestion of greater frequency in the lower two thirds of the pit. T h e action of this type, however, is not so marked as the White-decorated or the Interlocking. T h e White-slipped type, both fine and coarse ware, follows the same trend as the White-decorated. Above Level 6 it averages about 3 percent of the total number of sherds in each level. In Level 6 and below, this average is stepped up to approximately 8 percent of the total sherds per level for the fine-ware variety, and 16 percent for the coarse ware.

In rim and vessel form, there is a strong resemblance between Smooth plain red and most of the White-on-red types. Many body sherds of Smooth plain red also are undoubtedly from decorated White-on-red vessels. Stratigraphy of Smooth plain red is bound to be blurred, however, as successful distinctions could not be made between plain red, wellsmoothed sherds of Interlocking affilations, and the more abundant Smooth plain red associated with the earlier White-on-red style. T h e r e is a suggestion in this pit of a similar frequency trend for Coarse plain red; but this is not substantiated by Pit V I I and is undoubtedly the result of selective discarding, a matter discussed below.

Of the minor types of the Interlocking style, Three-color curvilinear and naturalistic falls into the pattern of occurrence set by the Interlocking type. Dot and Circle types, although appearing in small quantities, are mostly in the upper part of the Interlocking occurrence. In Pit I V the Smooth plain red shows its stratigraphic affinity to the White-on-red style by its large occurrences in the lower levels.

A m o n g other features of the stratification is the appearance of the two Late Chancay types in Pit IV, Levels i and 2, and not below these depths. T h e superficial position of these types in the deposits is reemphasized in the other test pits. T h e four sherds of the Negative painted style are all in Level 9 (4.004.50 meters), a position corroborated by the appearance of two sherds of the same type in Pit VII, one in Level 13 (3-25-3.50 meters)

O I tevets /1 2I 3 I «I 5

61

Z 3 4 5 6 7 Ö Q

0

12

3 4 5 6 7 a 9

0 / 2 3 4 5 6 7 0 9 Merois

WHITE DECORATED

WHITE

TYPE

TYPE

ZONED

il ô|

to\ "

I

MTERL0EKM5

el

TYPE

/3 151 l& I

3.00-325

3.25-5.50 350-375 3.75-4.00 4.00-4.25 ' 425-450

/7 J

4.50-4,75 4,75-500 5.00-5.25

/al 201

0-.50

.50-.75 .75- (.00 1.00-1.25 125 -I50 ! L50-I.75 175-2.00 2.00-225 2.25-2.50 250-2.75 2.75-3.00

FIG. 11. P O T T E R Y T Y P E S F O R P I T V I I , S I T E E (by

Percentage)

STRATIGRAPHY A T CERRO DE TRINIDAD and one in Level 17 (4.25-4.50 meters). Incised and punctated sherds are rare and are found only at the very top and near the bottom of the pit. T h i s widely divergent occurrence is also marked in Pit V I I and points to a difference in the sherds that have been classified under this heading. A l t h o u g h only a very few sherds are involved, it was noted that nearly all of the hollow-reed type of punctations occur in the upper levels, while the sherds from the bottom of Pits I V and V I I show a fine dot or triangular punctation. "Bucchero," or Polished black plain ware, is scattered through the upper three fourths of both Pits IV and V I I without showing any significant increases or decreases in frequency. Pit V I I has a similar vertical distribution of the key types, White-decorated and Interlocking. Some of the irregularity in the graphs can be explained by the use of .25-meter rather than .50-meter levels as in Pit IV. Some of it is unexplainable, but a smoothing of these irregularities leaves no doubt as to the time story. A l l high occurrences of Whitedecorated in Pit VII, save one, are found in Level 11 and below (2.75-5.25 meters). T h e exception is Level 8 (2.00-2.25 meters), whose medium-high frequency is overemphasized by the fact that no White-decorated sherds were found just below in Levels 9 and 10. Very few sherds obtained in either of these levels, owing to the fact that at this point the pit passed through solid adobe. T h e complete absence of the Interlocking type is also noted for Level g. T h e relatively high percentage of White-zoned in Level 9 seems more reasonable when an occurrence of four sherds of that type out of a total of forty-five is realized as the basis for computing the percentages. T h e White-zoned type, as in Pit IV, appears in all levels. A higher proportion of the type is f o u n d below the top 1.50 meters, or below the concentration zone of the Interlocking type. T h e Pit V I I graphs suggest a m a x i m u m for the type intermediate between that for Whitedecorated and Interlocking. T h e r e is a slight indication of this in the graphs for Pit IV. In Pit VII, White-slipped, like White-zoned, was not divided into coarse and fine ware in the sherd count but classified as one type. T h e combined type does not check very well with

169

the action of either the fine or coarse Whiteslipped type in Pit IV, although a greater popularity for allover white painting in the earlier levels is still implied. Interlocking, in Pit V I I , occurs in quantity down to a depth of 1.50 meters. From 1.50 to 3.50 meters deep there are moderate amounts of the type. Below 3.50 meters, as in Pit IV, there is practically an absence of Interlocking or any related three-color type. Pit V I I resembles Pit I V in the confinement of Dot, Circle, and X-mark types to the upper Interlocking distribution. "Probable Interlocking," a classification set up to take care of small or eroded sherds suggesting the Interlocking type, is confirmed in both Pits IV and V I I by its close concurrence with Interlocking. Smooth plain red, in Pit V I I , is more common in the lower White-on-red levels. Coarse plain red, on the other hand, is irregular but on the whole makes a stronger showing toward the top of the pit. Negative painted and Incised and punctated sherds check closely with Pit IV, as do the Late Chancay types and the Polished black ware. T h e concurrence of the two pits on several major points is gratifying. First, there is a zone below 3.50 meters deep in which n o three-colored sherds are found. In this same zone occur the greatest numbers of the Whitedecorated type. Second, above the 1.50-meter line are to be found most of the Interlocking sherds, although a few White-decorated pieces come from these top levels. T h i r d , there is an intermediate depth of from 1.50 to 3.50 meters below surface in which some Whitedecorated sherds and lessened amounts of Interlocking sherds are found. Fourth, related three-color types conform to the Interlocking distribution in the refuse and are most common in the upper part of that distribution. Fifth, types like White-zoned and Smooth plain red, which are typologically related to White-decorated, tend to be more abundant in the lower levels of the two pits. A n d sixth, Chancay Black-on-white and Chancay Whiteslipped occur only in the top .50 to 1.00 meter. These data enable us to place the cultural materials from the site in a chronological scheme. A s to the relative dependability of Pits IV and VII, both have similar soil strata and

170

STRATIGRAPHY A T CERRO DE TRINIDAD

structures. Both have an adequate number of sherds for stratigraphic techniques, and the sherd count by .50-meter levels is about the same for the two pits. Pit IV does not have the semisterile layer of adobe encountered at Levels 9 and 10 in Pit VII. But from the standpoint of classification, Pit IV is the less reliable of the two. As mentioned in the first section of this report, all sherds from Pit VII were classified in the laboratory, while discards of plain body sherds from Pit IV were made in the field. T h e number of discards per level is given in the table of occurrences for Pit IV. Most of the discards were of Coarse plain red. It is obvious that if all of the discarded sherds of this type were added to the various levels in the Coarse plain red column, the frequency of that type would more closely resemble the high occurrences of Coarse plain red in the upper levels of Pit VII. Under examination in the field before the types were formulated, it is likely that a number of other plain or plain-slipped types were also discarded. If so, this might account for the marked diminution of the Whiteslipped (coarse) type in the upper levels of Pit IV, a phenomenon not so striking in Pit VII. These weaknesses do not, however, affect the main trends of the stratigraphy which are borne out by the key decorated types. Neither Pit V nor Pit V I (Tables 10 and 9; Figs. 13 and 12) presents the complete stratigraphic story evident in Pits IV and VII. Pit 0 / 2 3 4-

0

I 2 3 4 5

6

0

12

VI, from which no discards were made in the field and in which the physical strata were more horizontally laid, is the better of the two. Total depth is 3.25 meters. T h e distribution of types in Pit V I gives the impression that the pit was contemporaneous with the upper 3.50 meters of Pits IV and VII. Occurrences of none or one or two sherds per level of White-decorated are noted down to 1.75 meters. Beginning at 1.75 meters, the occurrences and percentages of this type pick u p a little and remain constant until the twelfth level (3.00-3.25 meters). In this lowest .25 meter there is a considerable increase in White-decorated. There is little to say about the behavior of White-zoned other than that it is present in all levels. T h e Interlocking type makes up a much greater percentage of total sherds per level in the upper part of this pit than it does in Pit IV or Pit VII. There is no steady decrease of the type until Level 7 (1.75-2.00 meters), when it begins to decline; however, the type is present in appreciable strength all the way to the bottom of the pit. T h e Circle, Dot, and Three-color curvilinear and naturalistic types parallel, in a small way, the action of Interlocking but disappear completely a little short of the bottom of the pit. White-slipped, Smooth plain red, and Coarse plain red vary little in frequency throughout the twelve levels of the pit. Polished black ware also occurs from top to bottom with the 3 4- 5 6 7 6 9 10 U 12 13 Ut 15 16

Ltvas

WHITE DECORATED TYPE WHITE zatiED TYPE

IffTERLDCHiriB TYPE

FIG. 12. P O T T E R Y T Y P E S F O R P I T V I , S I T E E (by

Percentage)

METERS 0-.50 .50-. 75 .75-100 100-1.25 1.25-1.50 1.50-1.75 175-2.00 £.00-2.25 2.25-2.50 250-2.75 275-5.00 3.00-325

STRATIGRAPHY A T CERRO DE TRINIDAD e x c e p t i o n s of the first a n d t w e l f t h levels. T h e r e are a rather large n u m b e r of sherds of b o t h the L a t e C h a n c a y types in the t o p .50 meter. P i t V is very m u c h like Pit V I i n type dist r i b u t i o n s e x c e p t for a n u n u s u a l l y large occurrence of the I n t e r l o c k i n g type in the b o t t o m level (2.75-3.00 meters). T h i s sudden increase f o l l o w s a steady decline of the type w h i c h begins at L e v e l 6 (1.50-1.75 meters). W h i t e - d e c o r a t e d is scattered t h r o u g h most of the levels a n d shows a sharp percentage increase i n the b o t t o m level. L e v e l 12 (3.00-3.25 meters) c o n t a i n e d o n l y f o u r p l a i n sherds a n d is n o t considered i n the graphs. T h e b a g of sherds f r o m L e v e l 5 (1.25-1.50 meters) was lost in transit to the l a b o r a t o r y ; hence, the occurrences for this level are u n k n o w n a n d the g r a p h lines h a v e simply been e x t e n d e d f r o m L e v e l 4 to L e v e l 6.

V a n d V I , L a t e C h a n c a y types are f o u n d o n l y in the first level. T h e s e c o n d i t i o n s suggest that the r u b b i s h e x c a v a t e d in Pits V a n d V I dates f r o m that time, d u r i n g the o c c u p a t i o n at Site E, w h e n the W h i t e - d e c o r a t e d w a s a f a d i n g type a n d w h e n the I n t e r l o c k i n g style was g r o w i n g to a position of d o m i n a n c e . I n Pit I I there are six levels of .25 m e t e r each. D i s r e g a r d i n g the b o t t o m level, w h i c h has o n l y twenty-four sherds, some slight shifts in percentages of types in the r e m a i n i n g five levels reflect the sequence of Pits I V a n d V I I . W h i t e - d e c o r a t e d a n d I n t e r l o c k i n g increase a n d decrease respectively in f r a c t i o n a l percentages w i t h increasing depth. T h e five L a t e C h a n c a y sherds i n the p i t are in the t o p .25 meter. Pits I a n d I I I are t h o r o u g h l y m i x e d as far as the White-on-red a n d I n t e r l o c k i n g styles are concerned. T h e r u b b i s h at b o t h locations is q u i t e shallow. In spite of this, it is observed that L a t e C h a n c a y sherds o c c u r o n l y in the u p p e r half of each pit. It is o b v i o u s f r o m the f o r e g o i n g that the sequence at Site E is n o t a " c l o s e d " stratification. T h e c o n t i n u a t i o n of White-on-red style sherds t h r o u g h o u t the m i d d e n implies cult u r a l continuity. E v e n the key type, W h i t e decorated, is f o u n d consistently at all d e p t h s t h r o u g h o u t the d e e p pits, I V a n d V I I . T h e r e are, h o w e v e r , three zones or periods i n the r u b b i s h of Pits I V a n d V I I w h i c h are reasonably distinct f r o m each other in ceramic content a n d w h i c h c o i n c i d e w i t h some structure

T h e evidences f o r stratigraphy in Pits V a n d V I can be s u m m e d u p as follows. First, the I n t e r l o c k i n g type is in all levels, but there is a d r o p p i n g off in percentage of occurrence i n the lower 1.50 meters. Second, White-decor a t e d is f o u n d in small a m o u n t s at all depths, b u t there is a suggestive percentage increase i n the b o t t o m level of b o t h pits. T h i r d , threec o l o r types related to I n t e r l o c k i n g h a v e almost the same occurrences that they h a v e in Pits I V a n d V I I . F o u r t h , White-zoned, W h i t e slipped, a n d S m o o t h p l a i n r e d types, m o r e closely allied to the White-on-red style than to the I n t e r l o c k i n g , a p p e a r w i t h o u t c h a n g e t h r o u g h o u t the pits. A n d F i f t h , in b o t h Pits O

I 2 3 U 5 e

LBtas

O / 2 3 4 5 6

O / 2 3 *

Z £ L

r

WHITE

T

wi WHITE

DECORATED

k

ZDMED

TYPE

W

TYPE

^

171

5 6

7 8

fl

»«.fii5

0-.50 .50-.Y5 .75-LOO ^

^

^

^

^

^ ^ . 2 0 0 ¿ ¿ z L

w

MTERLOCNtlB TYPE

2.25-250

^ ^ ^ ^ ^ ^ i

METERS

H

H

H

2.50-2.75 B

M

B

i

Fic. 13. P O T T E R Y T Y P E S F O R P I T V, S I T E E (by Percentage)

(20%)

Z15-3M)

172

STRATIGRAPHY A T CERRO DE TRINIDAD

levels: (1) There is a bottom White-on-red zone below 3.50 meters deep which was devoid of any Interlocking style sherds; (2) there is an intermediate zone from 1.50 to 3.50 meters below surface in which the Interlocking and White-on-red styles were found mixed;

(3) and at the top, from o to 1.50 meters deep, there is a zone of refuse in which the Interlocking style was definitely predominant. From the stratigraphy of types, certain inferences can be made with relation to the soil profile on Figure 2.

STRUCTURES AND STRATIGRAPHY Floor E, encountered in both Pit IV and Pit VII, is at a depth of 1.50 meters. This floor is one of the most important in extent and is related to three walls, each of a different type of construction: puddled adobe, brick adobe, and rock (J, L, N on Fig. 2). On this floor, four Interlocking restorable beakers and a large Interlocking mammiform jar were found broken in situ. As the concentration of Interlocking sherds is above Floor E, this floor appears to mark the division between the Intermediate and Interlocking periods. T w o clay floors of similar construction were found above E in Pit VII. Floor D may be related to the rock wall, J, but the north profile (see inset, Fig. 2) shows the same wall on Floor E. Floor C, very near the surface, has what looks like the remnants of a wall of small, dome-shaped adobes in association. T h e possibility that any of the constructions above the 1.50-meter depth could have been made by the Late Chancay peoples is not well supported. Floor E is ruled out by the in situ Interlocking vessels. Floor C, the most recent structural level, has superimposed upon it a row of small, dome-shaped adobes, a type of adobe not found in any Late Chancay site in the valley. One-half meter below Floor E is the clay Floor F. O n this lower floor a White-zoned jar was found smashed in situ. Floor G, and

the rock and rubble Wall M, at 3.50 meters below surface, may form the bottom of the Intermediate period rubbish, as it is the lowest depth at which Interlocking sherds were found. Below 3.50 meters there are no floors or walls and no evidences of structures excepting the single wooden post H and the poles in connection with Burials 1 and 2 at the bottom of Pit VII. T h e rubbish between 3.50 and 5.00 meters in depth, above the ash on the old ground surface, contained only sherds of the White-on-red style. T h e association of structure with cultural stratigraphy in Pits V and V I is not as striking as in Pits IV and VII. T h e top 1.00 meter of Pits V and V I is a thick adobe platform or floor. There are evidences of small occupation layers through this adobe. T h e large amounts of Interlocking sherd material contained in it, and extending into part of the rubbish beneath it, make it clear that the platform, or closely packed series of floors, was built during the Interlocking period. Most of the rubbish below the clay, and a thin adobe floor level at 2.75 meters deep, can be attributed to the Intermediate period. There is no stratum representing the "pure" White-onred period at this location, so it is inferred that the inception of the rubbish at this point began sometime in the Intermediate period.

BURIALS AND S T R A T I G R A P H Y All of the burials in Pits IV and VII belong to the White-on-red period. A n unnumbered burial was seen in the profile of Pit IV below Floor G and inclusive in the rubbish of the White-on-red zone. Burials 1 and 2 of Pit VII, Olla A, and the burial, or burials, of Olla B, came out of the sterile sand under-

lying the rubbish. A list of the cultural material from these two important graves is given in Table 3. Burials 1 and 2 of Pit V came from the sand immediately beneath the rubbish of that pit. T h e lowest levels of Pit V are of the Intermediate (or mixed) period. Neither burial

STRATIGRAPHY

AT

CERRO

DE

TRINIDAD

*73

TABLE 3

TABLE 4

G R A V E ASSOCIATION LISTS: PITS IV AND VII

G R A V E ASSOCIATION LISTS: P I T V BURIAL

Catalogue Number Pottery vessels White-decorated 2 small mammiform jars 4 bowls, incurved, lipped, 2 handles 1 bowl, incurved, lipped 1 eccentric form

536, 537

Catalogue Number

Pottery vessels Red-on-white 1 jar, vertical or flaring neck, no handles BURIAL

535- 542. 546,551 545 547

Red-on-white 1 jar, vertical or flaring neck, no handles 1 jar, vertical or flaring neck, 2 handles or knobs 1 bowl, lipless, low, flaring Combination white-and red decorated 1 bowl, incurved, lipped, 2 handles

493

2

White-zoned 1 large mammiform jar 1 jar, small and fragmentary

494 495

TABLE 5 548

G R A V E ASSOCIATION LISTS: P I T VI

556 554

BURIAL 1

560

White-zoned (fine) 6 small mammiform jars 529, 528, 561, 544, 543, 541 2 large mammiform jars 552, 559 4 bowls, incurved, lipped, 2 handles 550, 562, 564, 565 1 jar, vertical or flaring neck, no handles 540 White-slipped (fine) 2 bowls, incurved, lipped, 2 handles 566A, 566B White-zoned (coarse) 2 large jars used as covers for burials (Ollas A and B) 530A, 530B Artifacts 1 cloth headdress 1 copper and gold mask 2 lapis lazuli spindle whorls (or beads) 1 stone spindle whorl (or bead) inlaid with shell 14 shell beads found in vessel 1 large worked shell i bracelet of shell beads (17 pieces) 1 necklace of shell beads (about 270 pieces) 1 wooden spear or dart

1

538 538 533 533 567 534 531 539 532

Pottery vessels White-slipped 1 large mammiform jar Three-color "Intermediate" 2 bowls, incurved, lipless Artifacts 1 clay figure of a bee (?) 1 human effigy (toy) with handle 2 shell beads Hollow unfired lumps of clay BURIAL

Catalogue Number

5!2 5ii.5«3 508 509 510 [no numbers]

2

Pottery vessels Whi te-decorated 1 bowl, lipless, low, flaring (maroon-red) 1 jar, vertical or flaring neck, no handles White-zoned 1 large mammiform jar Pottery vessels Coarse plain red 1 huge olla

BURIAL

3

BURIAL

4

516 515 514

jig

Pottery vessels White-zoned 1 jar, vertical or flaring neck, no handles 1 bowl, incurved, lipped, 2 handles (very large specimen)

520 521

STRATIGRAPHY A T TABLE 6 GRAVE ASSOCIATION LISTS: PIT II Catalogue Number

Pottery vessels Smooth plain red l bowl, lipless, incurved, 2 handles 488 Coarse plain red 1 jar, vertical neck, 1 flat or round handle on neck (a miniature) 487 Artifacts [no number] 1 gourd bowl was found under a sherd cover. Pottery associations with both burials were White-on-red style vessels. See T a b l e 4. T h e four graves in Pit V I were all sherd covered. T h e i r position in the strata is exactly the same as that for Burials 1 and 2 of Pit V: they were found in the topmost part of the sand underlying Intermediate period rubbish. See T a b l e 5. In Pit I, two burials, an adolescent and a child, were found without sherd coverings or accompanying artifacts. There is no means of assigning these burials to any of the three periods. T h e infant burial in Pit II lay in the top of the sand, just beneath the rubbish. There was no sherd covering. T h e only clue to period is the overlying refuse, suggesting the Intermediate period, through which the grave does not appear to have intruded. SUMMARY OF BURIALS

Stratigraphically, the burials can be grouped in three classes: (1) those of the White-on-red period from Pits IV and VII; (2) those underlying Intermediate period rubbish in Pits V and V I and probably of the Intermediate period; (3) and those of indeterminable stratigraphic relatonships from Pit I. T h e burial of Pit II may go with the second class, although the evidence is admittedly weak. There are several similarities and some differences between the first and second classes of burial. Sherd coverings for burials are found in both Pit IV and Pit V I I and in Pits V and VI. In the first-named excavation they have in addition a substructure of poles. A crude rock

CERRO

DE

TRINIDAD

covering was also noted for the unnumbered burial in the White-on-red stratum of Pit IV. T h e more elaborate mode of burial in the earlier period may be due, however, to the fact that the burials of Pits IV and V I I were adults, whereas those of Pits V and V I were all infants or children. A l l burials at the site, except the rock-covered one just referred to, were found in the top of the natural sand. T h i s indicates a custom of shallow burial, for in neither set of pits was there any evidence of intrusion from above through any great depth of overlying rubbish. T h e intact Floor G and Wall M over the burials in Pits IV and V I I preclude the possibility of the intrusion of Burials 1 or 2 from the Interlocking or Intermediate periods and limit the original depth of the graves to no more than a meter, if that much. T h e burials of Pits V and V I contain pottery types which link them with the Whiteon-red graves of Pits IV and VII. These types are the White-decorated, White-zoned, Whiteslipped, and Red-on-White. In addition the graves had two pots of the Three-color "Intermediate" type. Of the two vessels of Whitedecorated (the marker type for the Whiteon-red period) found with the burials in Pits V and VI, one is a low, flaring bowl of the maroon-slipped variety, the other is a lipped or necked jar. T h e small jar form with a vertical or flaring neck, of which there are eight in the White-on-red collection, occurs three times in the White-on-red period graves of Pit V I I and has four occurrences in the graves of Pits V and VI. T h e remaining jar was a White-zoned specimen taken from Floor F (see Fig. 2) in Pit VII, in the Intermediate period refuse. T h e large mammiform jar, like the small, necked jar, is another form whose occurrence in Pits V and V I graves argues for their Intermediate period position. Six of these large jars were found at Site E. T w o of the six were with the White-on-red period burials in Pit VII; three came from graves in Pits V and VI; and one is an Interlocking style painted piece from Floor E (see Fig. 2) in Pit VII. In this connection it is pointed out that neither the small, canteenlike, mammiform jar nor the small, incurved, and lipped bowl is found in any of the graves in

STRATIGRAPHY AT CERRO DE TRINIDAD Pits V and VI. These two small forms seem to be most characteristic of the White-on-red period at the site. T h e only incurved, lipped bowl from a Pit V I grave is a large specimen without the median ridge typical of the smaller bowls of this form from the Whiteon-red period. T h e two Three-color "Intermediate" vessels from Burial 1, Pit V I , have the median ridge but are lipless and without handles. T h e low, incurved, lipless bowl form has been indicated

as resembling, but not typical of, White-onred shapes. T h e use of three colors suggests the Interlocking style, while the design arrangement is more like the White-on-red. Some Interlocking rim sherds from the refuse are from incurved bowls with a similar unmodified lip. Most of these have the nestedtriangle variation of the fish-head motif, a design resembling the Three-color "Intermediate" bowls, especially the smaller of the two vessels, in simplicity and arrangement.

S E Q U E N C E O F PI J O D S A T S I T E E T h e following descriptions of the three periods are derived from the correlations of pottery, artifact, structure, and burial stratification. THE WHITE-ON-RED PERIOD

T h e evidence for this period comes from those rubbish levels and burials in Pits I V and V I I below the depth of 3.50 meters. It is the first known occupation of the site. Stratigraphically, it is defined by the almost complete dominance of the White-on-red ceramic style. T h e period can be briefly described in the following summaries. Pottery.—1. T h e principal pottery types of the White-on-red style are White-decorated and White-zoned. Both of these types enjoy their greatest frequency of occurrence during this period. 2. M i n o r types of the same style are Red-onwhite and White-slipped. 3. T h e most popular forms of the style, and especially the White-decorated type, are the small mammiform jar and the incurved, lipped bowl with horizontal handles. 4. A small, necked jar, a large mammiform jar, and a plate with upturned rim are other forms. 5. Smooth plain red is more popular in this than in the later periods. 6. A n i m a l and human (or demon) pottery adornos, decorated in the White-on-red style, are recorded. 7. Negative painted ware, Polished black ware, and fine-line and dot incision and punctation decoration, each represented only by occasional pieces, are the only foreign types

found in association with the White-on-red types during this early period. Figurines.—There is a single human figurine head of this period. It has punched clay pellet eyes and a mutilated nose. Spindle whorls.—There are four spheroid whorls of this period: two of lapis lazuli, a shell-inlaid greenstone or jadeite whorl, and an incised pottery whorl. Wooden artifacts.—A spear or throwingstick dart, fire-hardened and sharpened, was found. Objects of ground stone.—The one artifact of this class found is a flat, circular stone with a beveled edge. Cloth.—A turbanlike headdress of cotton cloth is the only cloth specimen from this period. Metal.—There is but a single metal artifact, a part-copper, part-gold mask or face covering. Ornaments of shell.—These include a necklace, a bracelet, and some miscellaneous finds of Spondylus pictorum shell beads. These beads vary greatly in size and shape. Structures.—No in situ walls or floors can be assigned definitely to this period. Domeshaped adobes, however, were found in a huge White-on-red olla, below 3.50 meters, in Pit IV. It is probable that structures of these small adobes were erected in much the same fashion as in the later periods. Graves.—Sherds of a coarse-ware Whitezoned pottery were used as covers for flexed burials. Grave goods, of pottery and other artifacts, are abundant. Burials are in the natural sand just beneath the rubbish, as a rule, and do not appear to have been very deep.

176

STRATIGRAPHY A T CERRO DE TRINIDAD

Poles may have formed a frame on which to rest the sherd coverings, or they may have been lashed together to form a litter of some sort for the dead. A crude rock vault, or covering of the burial with irregular boulders, is probably another type of grave of the period. THE INTERMEDIATE PERIOD

T h e Intermediate period is stratigraphically defined by a zone of transition in the refuse between the lower White-on-red period and the upper Interlocking period. T h i s is marked in Pits I V and V I I and implied in the lower half of Pits V and VI. Pottery.—1. T h e White-decorated type is markedly on the decline; White-zoned, on the other hand, shows a tendency to increase. 2. T h e small, necked jar and the large mammiform jar continue through into this period, but the small mammiform jar and the incurved, lipped bowl with a pronounced median ridge have declined. 3. T h e three-color Interlocking type first appears, in relatively small amounts. 4. T h e Three-color "Intermediate" type appears. 5. Smooth plain red decreases slightly, and Coarse plain red is on the increase. 6. T h e r e are a few sherds of Polished black ware. Unfired clay vessels.—These objects appear only in rubbish of this period. A few unfired clay fragments from Burial 1, Pit VI, further substantiate their Intermediate period position. Panpipes.—A three-color fragment of a panpipe and fragments without paint were found in Intermediate period strata. T h i s is the earliest appearance of panpipes at the site. Figurines or adornos.—There are a human face adorno with mutilated nose and an unidentifiable animal head from Intermediate period rubbish. A human figurine or toy came from Burial 1, Pit V I . Ornaments of shell.—Two beads of small spiral shells occur. Structures.—The rubble-filled rock wall and adjoining clay floor are, in Pits I V and VII, associated stratigraphically with the Intermediate period. So is a higher thin clay floor situated on a thick mass of clay. A thin clay

floor near the bottom of Pits V and V I can also be related to this period. Burials.—All of the burials of the Inter, mediate period are infants or children. Four of these, however, have the coverings of heavy sherds used in the burials of the White-onred period. T w o burials do not have these coverings. T h e burial position, where determinable, is flexed. White-on-red style pottery and Three-color "Intermediate" type pieces are found as burial furniture. In one instance other artifacts were placed with a burial. THE INTERLOCKING PERIOD

T h e Interlocking period refuse is defined as that zone overlying the clay floor at 1.50 meters deep in Pits I V and V I I . T h i s checks with the upper half of the rubbish in Pits V and VI. Pottery.—1. T h e Interlocking type is the predominant decorated ware, and beaker bowls are most common in this period. 2. Related three-color types also are f o u n d in small percentages, tending to be slightly later in the period. 3. Some White-on-red style pottery also is found. Most of this is White-slipped or Whitezoned. 4. In plain wares there is a greater occurrence of coarse than of fine pottery. 5. Polished black and incised and hollowreed punctated sherds occur in very small amounts. Panpipes.—Panpipes, which came in during the Intermediate period, are found during Interlocking times. One painted, with a white circle, and unpainted fragments are f o u n d in the upper levels of the refuse. Figurines or adornos.—Fragments of human figurine heads, similar to the one from the White-on-red period, occur. A number of adorno fragments, including a mold-made human head, come from doubtful stratigraphic proveniences but may belong to the Interlocking period. Spindle whorls.—Two pottery disk whorls, a gourd disk whorl, and a bead-shaped whorl of stone were found. Bone artifacts.—A bone throwing-stick hook occurs in this period. Wooden artifacts.—There are "plugs" or "tops" of wood.

spxaqs

S

¡v}0£

spxvjsip 9ivin uwjj ÍVDUVl{3

oo ^

«o Ü . to ^

« m o> «2

00

S aM. * .S

3JVJ

3]ii{in-uo-y3v]g ivjUVlfQ 3}V7 T¡3vjq

"^'S. «n 2n

S* '8

paifsijoj

8

-8

-

p3]3pOUl •3Stf\¡ p3]v]3unj puv P3S1JUJ 'JSiIV P3JVX033P 'JflJV

p3)Utv4

s co

5

« ~

S' c

" 8

s

»• -Sä o ¡¡ü

•o ^í S-O B * £ « . 5 ¡e

¿ 60

0

]V)OX

00

s

t»

to

... or, ° °° S2 ® o

— «„ So- -

to

p344ip-31lllM &V3Um¡3 3}VJ g)ii/cn-uo-y3V¡g Kvouvijj a)07 yovjq

on

p3ijSi]oj

psppojy

'jsíjv

psjvjDunj psspuj

puv "3sty¡

p3}VX033p

-3SIJM

pax uiv¡4

3SXV03

to

8

„ • * SO

to et o

o> Ol Tf c

i¡tooui$

• Ci

CT>

^

m IO " o

O

r

on r l l - ^ 0O « M

on 0 0 « «

q

M

o

1

^

-t

CO

« ' S« on

- on

eo

M I>°— Son í,

O

pax uiv¡4

« "- • »' S2o

«

» d i ? -"

r

•AS rN

t> eo

&£

O o

-

0«

T*


« 1 Ï

o>to o

XO o

-

5

f

"

a

f

^s

So

1

o

« " 2 " o

" o

H

3iii¡5 Nieveria, Interlocking ware from, 62 ff. passim; Early L i m a style or period, 87; Early L i m a collections from, 193 Northern survey, 6-10; personnel, 6 Nose mutilation, trait of, 192 Ocucaje, Hacienda, 21, 22, 23 Ocucaje valley, Necropolis and Cavernas cultures, 21 if. Ollada, Late-period habitation and ceremonial site, Cerro de T r i n i d a d , 8 Olleria, gorge of, 19 Pachacamac, survey trips to south of, 18-25; ceramic sequence, 27-31; main ruin, 32-33; pre-Inca period, 34; one of most famous archeological sites in Peru, 31; plan of main ruins, f. 32; adobita, 45, 79, 82, ill. 43; ceramic styles and types: Interlocking, 62-71; Negative, 71-75, 80, 82, 85, 87, g i , tabs. 93-117, 120; various, 75-78; ceramic styles and periods, 86-90; suggested sequence, 91, tab. 120; occurrences, with style percentages in blocks and strata (Cut 2), tab. 93-117; provenience of pottery illustrated, tab. 120-21; see also Interlocking ceramic style, Pachacamac; Negative ceramic styles and types, Pachacamac T e m p l e of: temple of chief creator god Pachakamaj, 32; Uhle's excavations below, 34-36; Inca and Late Coastal burials, 34; cross section of excavation, ill. 35; ceramic sequence, 36; ground plan, ill. 38; blending of Inca and Coastal cultures, 39 T e m p l e of the Sun, 31, 32, 37-39; Inca in origin, 32, 39; excavations below, 40-48; orientation, 40; ground plan of present excavations, 41; test pits, 41, ill. 41, 43, 46; Cut 1, 42; Cut 2, 44, ill. 43; strata and structural superimposition, 44-48, ill. 43, 46; ceramic styles and types, 49-78; see also under names of styles and types Pan-American Highway, 6, 9, 130 Panpipes, 159 Paracas site, 21 Parr6, Alberto Area, 203« Pasamayo valley, 130, 184 Peru, first demonstrable case of cultural stratification in, 31; styles of coast seen in Institute of Andean Research collection, 202 central coast: northern survey, 6-10, map, 7; excavations in the vicinity of Puerto de Supe, 11-14; An-

INDEX con shell mounds, 15-17; southern survey, 18-25; Interlocking period, 89; concept of a developed painted-pottery period first formulated, 127; cultural sequence, 128, 195, tab. 196; incised pottery, 189-91; cultural isolation, 190; comparisons: White-on-red sytle, 191; Negative painted style, 192; Interlocking style, 192-95; convergence of strains and styles at beginning of Middle Period, 194 Pisco, 21 Pisco valley, 20 Plain Polished black ware, 56, ill. 59, tab. 92-116 Plain red painting ceramic style, 56, ill. 59, tab. 92-116; stratigraphic distribution, 83; see also Inca-associated ceramic styles Plain red wares from Cerro de Trinidad, Site E, 152; types: smooth, 152; coarse, 153 Plain style pottery, Pachacamac, 76, 82, tab. 93-117 Polished black ceramic style, 56; types, 56, ill. 55, 59, tab. 92-116; stratigraphic distribution, 83; see also Incaassociated ceramic styles Polished black (or "Bucchero") ware from Cerro de Trinidad, 142, 155, 169 Polished White-on-red ware, 60, ill. 61, tab. 92-116 Pottery, see Ceramics Pressed or relief-decorated red ware from Supe valley, 10 Pressed relief ceramic type, Modeled brown ware, 58, ill. 57, tab. g 2 - n 6 Proto-Lima ceramic style, 87 Proto-Nazca culture, links between Interlocking culture and, 36 Puente Piedra, cemetery, 6 Puerto de Supe, excavations in vicinity of, 11-14, !5> Aspero, 11 ff.; Lighthouse site, 11, 13, 16; Li-31, 13 Punctate (and incised) ceramic type, Pachacamac, 76, 82, 88, ill. 77, tab. 93-117 Punta Pegota, 10 Recuay ceramic styles, Negative styles, 22, 87; related to Pachacamac Interlocking and Negative styles, 73, 75; not yet stratigraphically placed, 75 R e d and black on white ware from T e m p l e of Pachacamac, 34 Red-on-white ceramic type from Cerro de Trinidad, 145 R e d painting ceramic style, plain, 56, ill. 59, tab. 92-116 Relief ceramic type, Modeled brown ware, 58, ill. 57, tab. 92-116 Relief-decorated red ware from Supe valley, 10 Respaldiza, José, 18, 19 Rimac valley, Interlocking and Early L i m a ceramic style, 193 R i m profiles from Cerro de Trinidad, ill. 148 R i o Chancay, 130 R o m á n y Zamora, Friar Jerónimo de, 48; quoted, 37 Saenz, Moisés, 19, 20 "Salinar" culture period, 192 San Marcos, University of, 31 San Nicolas Hacienda, quebradas facing fields of, 11, 12 Santos site, or Site A, see under Callango Seminegative design, 54, ill. 53, tab. 92-116

221

Serrated pattern, 52, ill. 51, tab. 92-116 Serried diamond pattern, 52, ill. 51, tab. 92-116 Shell-mound ceramics, see Early Ancon-Supe Shell-mound culture, Early Ancon-Supe, 12 Shell mounds, at Ancon, 6, 15-17; at mouth of lea river, »4 Shell ornaments, 165 Site A cemetery at Cerro de Trinidad, see Cerro de Trinidad Site A or Santos site, see Callango, Site A Site B, Callango, see Callango, Site B Site C, Uhle's, 192 Site E, see Cerro de T r i n i d a d Skulls, deformed, of Cavernas-Necropolis culture, 22; coastal or " C h i n c h a " type, 34 Smooth plain red ceramic type, 142, 168, 169 Southern Survey, 18-25 Spindle whorls, 163 Spoon, carved bone, 17 Spouts, Pachacamac Interlocking style, 71, ill. 72, tab. 93-117; stratigraphic distribution, 85 Squier, E. George, 37 Stratigraphy, distribution of ceramic styles (Cut 2 below Sun Temple), 79-82, ill. 81; of ceramic types, 83-85; at Cerro de Trinidad, Site E, 167-83, ill. 167, 168, 170, 171, tabs. 177-81, 183; structures, 172; burials, 172; grave association lists, 173 Strong, William Duncan, northern survey, 6-10; Middle Ancon I Incised pottery, 190; incised pottery Early Period at Ancon, 195; formulated sequence of periods at Ancon, 201; occurrence of forty traits in each Necropolis period listed by, 203; intercorrelation by periods, 203, tab. 205 and Corbett, John M., " A Ceramic Sequence at Pachacamac," 27-121 and Willey, Gordon ft., "Archeological Notes on the Central Coast," 1-25 Structures and stratigraphy at Cerro de T r i n i d a d , 172 Style, term, 4g Sub-Chancay ceramic style, 6, 9, 90 Sun T e m p l e , see Pachacamac, T e m p l e of the Sun Supe valley, 10 T a m b o Colorado, 20 Teatino, cemetery, 9; incised ware from, 191 T e l l o , Julio C., 6, 9, 18 If. passim, 3g, 75, 191, 195; survey trips under guidance of, 18; excavations at Pachacamac, 31: excavations at Convent Mamacona, 32; work at Chavin and at Casma, 190 T e m p l e of Pachacamac, see Pachacamac, T e m p l e of T e m p l e of the Sun, see Pachacamac, T e m p l e of the Sun T e x t i l e design character of Interlocking fish motif, 35, 149. 193. 194. 195. ill- 150 Textiles, 16, 20, 21, 44, 134, 138, 140, 141; Tiahuanaco tapestry poncho, 34; Coastal Tiahuanaco style, 35; Inca, 47; cotton cloth headdress, 165 T h i n bowls plain type, Pachacamac, 76, 82, tab. 93-117 Three-color Curvilinear and Naturalistic type, Interlocking style, from Cerro de T r i n i d a d , 151, 168 Three-color geometric period, see under Black, white, and red

222

INDEX

Three-color "Intermediate" style from Cerro de T r i n idad, Site E, 151 Tiahuanaco-Epigonal style, 127 Tiahuanacoid ceramic style, 36, 86, 88, 89; sherds in Supe valley, 10; established as a generic Peruvian style, 127; see also Coastal Tiahuanaco Tiahuanacoid influence on Early Lima, 193, 194 Tiahuanacoid period, 89; see also Coastal Tiahuanaco ceramic style; Epigonal ceramic style Tiahuanaco tapestry poncho, 34 Triangle and dot decoration, Pachacamac Interlocking stvle, 65, ill. 68, tab. 93-117; stratigraphic distribution, 84

Truell, Pablo, collection of Cavernas-Necropolis style materials, 21, 22 Type, term, 49 Uhle, Max, 9, 15, 16, 21 if. passim, 39, 48, 50, 54, 60, 75, 87, 8g, 90, 191, 201; excavations at Cerro de Trinidad, 6, 8; excavations in vicinity of Puerto de Supe, n , 13; pottery collections from lea valley at University of California, 25; reported first demonstrable case of cultural stratification in Peru, 31; simplified plan of main ruins at Pachacamac, f. 32; excavations below the Temple of Pachacamac, 34-36, ill., 35; quoted, 35« 37» 182, 20in, 207; discovery of Interlocking style below the Temple of the Sun, Pachacamac, 36; simplified ground plan of Temple, 38; Interlocking style first clearly recognized by, at Chancay, 62; collection of Interlocking sherds from Chancay, 63, 193, 194, tab. (classification), 183; styles encountered in graves, Temple of Pachacamac, 86; one of his earliest styles, 91; investigations along Peruvian coast, 127 if.; objective for research at Cerro de Trinidad the determination of relationship between White-on-red and Interlocking ceramic styles, 127, 128; salient features of vessels of his first and second periods at Cerro de Trinidad, 142; Site C at La Calera de Jecuan, 192; expositipn of relationship between Tiahuanacoid and Early Lima, 194 Unfired clay vessels, 159

Unpolished White-on-red style pottery, 58, ill. 57, tab. 92-116

Valcárcel, Luis E., 18, 19 Walls, cyclopean, at Cerro de Trinidad, 8; rock-walled structure at Puerto de Supe, 11, 12; at Ancon, 16; La Muralla, 20 White-and-red decorated ceramic type, 146 White-decorated ceramic type, 143 White-on-red ceramic style, Chancay, 6, 8, 76, 82, 87, 88, 91, 142, 191; in sequence of styles and periods at Pachacamac, 91, tab. 120 Early Ancon-Supe, 191 from Cerro de Trinidad, 127, 133 ff. passim, 139, 140, 141, 142, ill. 144, 148; relationship between Interlocking and, 127, 128; salient features, 142; classification, 143; types, 143-49; i n sequence of periods, 175 Inca-associated, 58, ill. 57, 61, tab. 92-116; Polished, 58; stratigraphic distribution, 84; comparisons, 191; see also Inca-associated ceramic style White-on-red ceramic type, Pachacamac Interlocking style, 67, 76, 82, 88, ill. 69, tab. 93-117 White-on-red period, possibility of isolating, 88; Baños de Boza represents a single occupation of, in Chancay valley, 187 White-slipped ceramic type; Chancay, 156 White-on-red style: fine and coarse ware, 147 White-zoned ceramic type, fine ware, 146; coarse, 147 Whorls, spindle, 163 Willey, Gordon R., 19, 87, 88, 91; northern survey, 6-10; "Excavations in the Chancay Valley," 123-96; "A Supplement to the Pottery Sequence at Ancon," 197-211; see also Strong, William Duncan, and Willey, Gordon R. Wooden artifacts from Cerro de Trinidad, 164 X-mark ceramic type, in Interlocking style, 151, 193; in Tiahuanacoid-influenced ware, 194 Zapallar cemetery, 6 Zárate, Agustín de, 34

P U B L I C A T I O N S FOR I N S T I T U T E OF A N D E A N

THE RESEARCH

PROJECTS IN LATIN AMERICA UNDER THE SPONSORSHIP OF THE COORDINATOR OF INTER-AMERICAN AFFAIRS PROJECT I . N O R T H E A S T E R N M E X I C O

1 a) Ekholm, Gordon F. Excavations at Tampico and Panuco in the Huasteca (Anthropological Papers of the American Museum of Natural History, Vol. 39, Part 1). PROJECT 2 . NORTHWESTERN M E X I C O

2 a) Kelly, Isabel. Excavations Ibero-Americana, No. 25).

in Western Jalisco (University of California

PROJECT 3. C E N T R A L COAST OF PERU

3 a) Strong, William Duncan, and Gordon R. Willey. "Archeological Notes on the Central Coast" (in Archeological Studies in Peru, 1941-42, Columbia Studies in Archeology and Ethnology, Vol. 1, No. 1). 3 b) Strong, William Duncan, and John M. Corbett. " A Ceramic Sequence at Pachacamac" (in Archeological Studies in Peru, 1941-42, Columbia Studies in Archeology and Ethnology, Vol. 1, No. 2). 3 c) Willey, Gordon R. "Excavations in the Chancay Valley" (in Archeological Studies in Peru, 1941-42, Columbia Studies in Archeology and Ethnology, Vol. 1, No. 3 ) . 3 d) Willey, Gordon R. "A Supplement to the Pottery Sequence at Ancon" (in Archeological Studies in Peru, 1941-42, Columbia Studies in Archeology and Ethnology, Vol. 1, No. 4). PROJECT 4 .

CHILE

4 a) Bird, Junius. Excavations in Northern Chile (Anthropological Papers of the American Museum of Natural History, Vol. 38, Part 4). PROJECT 5 . V E N E Z U E L A AND THE W E S T INDIES

5 a) Howard, George D. Excavations at Ronquin, Venezuela (Yale University Publications in Anthropology, No. 28). 5 b) Osgood, Cornelius. Excavations at Tocoron, Venezuela (Yale University Publications in Anthropology, No. 29). 5 c) Osgood, Cornelius. The Ciboney Culture of Cayo Redondo, Cuba (Yale University Publications in Anthropology, No. 25). 5 d) Osgood, Cornelius, and George D. Howard. An Archaeological Survey of Venezuela (Yale University Publications in Anthropology, No. 27). 5 e) Rouse, Irving. Archaeology of the Maniabon Hills, Cuba (Yale University Publications in Anthropology, No. 26). PROJECT 6 .

COLOMBIA

6 a) Bennett, Wendell C. Archaeological Regions of Colombia: a Ceramic Survey (Yale University Publications in Anthropology).

PUBLICATIONS FOR THE INSTITUTE OF ANDEAN RESEARCH 6 b) Ford, James A. Excavations in the Vicinity of Cali, Colombia (Yale University Publications in Anthropology, No. 31). P R O J E C T 7 . SOUTHERN H I G H L A N D S OF P E R U

7 a) Kidder, Alfred, II. Some Early Sites in the Northern Lake Titicaca Basin (Papers of the Peabody Museum of American Archaeology and Ethnology, Harvard University, Vol. 27, No. 1). 7 b) Rowe, John Howland. An Introduction to the Archaeology of Cuzco (Papers of the Peabody Museum of American Archaeology and Ethnology, Harvard University, Vol. 27, No. 2). 7 c) Tschopik, Marion Hutchinson. Some Notes on the Archaeology of the Department of Puno, Peru (Papers of the Peabody Museum of American Archaeology and Ethnology, Harvard University, Vol. 27, No. 3). PROJECT 8. SOUTHERN C O A S T O F P E R U

8 a) Newman, Marshall T . Some Indian Skeletal Material from the Central Coast of Peru (Papers of the Peabody Museum of American Archaeology and Ethnology, Harvard University, Vol. 28, No. 1). 8 b) Tello, Julio C. Paracas (University of San Marcos, Lima). P R O J E C T 9 . N O R T H E R N PERU AND ECUADOR

9 a) McCown, Theodore D. Pre-Incaic Huamachuco: Survey and Excavations in the Northern Sierra of Peru (University of California Publications in American Archaeology and Ethnology, Vol. 42, No. 1). 9 b) Collier, Donald, and John V. Murra. Survey and Excavations in Southern Ecuador (Field Museum of Natural History, Anthropological Series, Vol. 35)PROJECT 10. C E N T R A L A M E R I C A

10 a) Longyear, John. Excavations in Salvador (Memoirs of the Peabody Museum of Archaeology and Ethnology, Harvard University, Vol. 9, No. 2).