SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME I [Whole Volume] Valdivia Bay. looking from the Machalilla Phase site of G-110: La Cabuya northward toward the Yaldivia Valley (upper left), showing typical overcast conditions during the garua season. Early Formative Period of Coastal Ecuador: THE VALDIVIA AND MACHALILLA PHASES Betty J. Meggers, Clifford Evans, and Emilio Estrada SMITHSONIAN INSTITUTION Washington 1965 A Publication of the SMITHSONIAN INSTITUTION United States National Museum LIBRARY OF CONGRESS CARD 65-6117z United States Government Printing Office, Washington, 1965 For sale by the Superintendent of Documents, U.S. Government Printing Office Washington, D.C., 10401 Smithsonian Contributions to Anthropology Anthropology was the subject of the first Smithsonian publication, "Ancient Monuments of the Mississippi Valley," by E. George Squier and E. H. Davis. Issued in 1848 as volume 1 of the Smithsonian Contributions to Knowledge, it has become a landmark in North American archeology. Smithsonian interest in anthropology has continued through the years. Several Smithsonian series have been devoted exclusively to anthropology. In addition, works on this subject have appeared regularly in a number of the Institution's other series. Among these (with first or in- clusive dates of such appearance) are the Smithsonian Annual Report (from 1865), Smithsonian Mis- cellaneous Collections (from i860), Explorations and Fieldwork of the Smithsonian Institution (1927-1940), Smithsonian Institution War Background Studies (1942-1945), United States National Museum Annual Report (1884-1904), United States National Museum Bulletin (from 1879), and the Proceedings (from 1878) and the Circular (1883-1888) of the United States National Museum. Among the series devoted exclusively to anthropology, some have been short lived, others have continued to this day. Contributions to North American Ethnology (vols. 1-7, 9) was issued from 1877 to 1895. The first through the forty-seventh Annual Report of the Bureau of American Ethnology (1879- 80 through 1929-30) contained scientific papers, often of monographic length, on the American Indians (the forty-eighth Report, 1930-31, contains an index to these papers). And 16 volumes of the Publications of the Institute of Social Anthropology appeared between 1944 and 1953. In 1896 the first Bulletin of the Bureau of American Ethnology was issued. This series soon be- came a major vehicle for anthropological publication by scientists at the Smithsonian and around the world. In some 200 Bulletins have appeared monographs and shorter papers on the archeology, ethnology, linguistics, and physical anthropology of the New World, and also basic handbooks on the Indians of North America (no. 30) and of South America (no. 143), and on North American languages (no. 40). Within the Bulletin series also have appeared the subseries Anthropological Papers, nos. 1-80 (1938-1965) and River Basin Surveys Papers, nos. 1-39 (1953-1965). In 1964, the Bureau of American Ethnology and the U.S. National Museum Department of Anthropology were replaced by the Smithsonian Office of Anthropology. It consists of a Division of Physical Anthropology and a Division of Cultural Anthropology concerned with archeology, ethnology, and linguistics. It also administers the Smithsonian River Basin Surveys. Unlike the former Bureau, its activities are not restricted to the native peoples of the New World. While continuing work in this field, the Office of Anthropology is expanding its research programs in Africa, Asia, and the Pacific. Such reorganization and expansion make it appropriate to consolidate anthropological publi- cation at the Smithsonian in a new series of worldwide scope: Smithsonian Contributions to Anthropology. This volume inaugurates the new series. With rare exceptions, all future anthropological monographs and papers issued by the Smithsonian will appear in this series. In addition to works by members of the staff, a select number of contributions will be accepted from authors outside the Smithsonian Institution. Copies of Smithsonian Contributions to Anthropology are distributed by the Smithsonian to libraries, to research institutions, and to recognized specialists in the various fields treated, both in this country, and abroad. Further free distribution is made by the Superintendent of Documents to depository libraries in each of the 50 States. Students and other interested individuals may purchase copies from the Superintendent of Documents, as indicated on the opposite page. Richard B. Woodbury, Acting Head Smithsonian Office of Anthropology Museum of Natural History Preface The publication of this report is a monument to the importance of international cooperation in scientific endeavor. The archeological sites and complexes were discovered by Ecuadorians, detailed analysis of the developmental sequences was furnished by North Americans, invaluable information for comparative study was provided by Japanese, and a Chilean prepared the report on skeletal remains. To those of us who are listed as authors, working with all of these people has been a memorable experience not only because the scientific results have been so exciting, but because the context in which they have been derived has been so rewarding. The largest contribution has been made by the many Ecuadorians who have assisted with fieldwork and preparation of the bulk of material for analysis over the years. Some should be singled out for special mention. Dr. Carlos Zevallos Menendez, then President of the Casa de la Cultura Ecuatoriana, Nucleo del Guayas, arranged for permission under the Ecuadorian antiq- uity laws to conduct the archeological field research. Felix Martinez and later Julio Viteri served as foremen during much of the excavation at G-31. During two seasons of work by Meggers and Evans at G-31 and G-54, Francisco Salcedo generously made available a comfortable house near the site as field headquarters. Washing and preliminary sorting of material from G-84 and G-31, Cut J was done by Walter Molina, part-time aide in the Museo "Victor Emilio Estrada." Staff members of the former Division of Archeology, Museum of Natural History, U.S. National Museum who have over the years assisted in the laborious job of washing, numbering and classi- fying Valdivia and Machalilla Phase materials, are Mr. George Metcalf, Mr. Robert C. Jenkins, and Mrs. Willie Mae Pelham. We are indebted to personnel of other divisions for identification of stone, bone and shell remains, including Dr. Harald A. Rehder, Division of Mollusks; Dr. Henry Setzer, Division of Mammals; Dr. E. P. Henderson, Division of Meteorites; Dr. Leonard P. Schultz and Dr. William R. Taylor, Division of Fishes. Mr. Henry Wright assisted one summer with sorting of rocks from G-31: Valdivia into possible and impossible artifacts. Carbon-14 determinations, which confirm the early chronological placement of the Valdivia complex, were made over several years at three different laboratories: the United States Geologi- cal Survey Low Frequency Radiation Laboratory, the University of Michigan Laboratory, and the Smithsonian Institution Carbon-Dating Laboratory. We would like to thank Dr. Meyer Rubin of the United States Geological Survey for his willingness to accept shell samples for dating at a time when this material was considered unsuitable in many quarters, and Dr. Austin Long of the Department of Radiation and Organisms, Smithsonian Institution Carbon-Dating Laboratory for consultation and advice in the evaluation of the entire series of dates, which led to several of the interpretations in the section on dating. Our inferences about the origin of Valdivia Phase pottery would have been poorly supported had it not been for the opportunity to visit Japan during March and April, 1963 to examine col- lections and talk with experts on the Early and Middle Jomon Period. Initial communication with Japanese archeologists was facilitated by advice and introductions from Dr. Chester Chard, University of Wisconsin; Dr. Richard K. Beardsley, University of Michigan, and Dr. Edward Norbeck, William Marsh Rice University. Informed in advance of our general problem, mem- bers of the staff of the Institute of Cultural Anthropology, University of Tokyo, headed by Prof. Seiichi Izumi, laid out a tentative schedule of visits that permitted us to make best use of our lim- ited time. Our ability to accomplish so much was largely because of this generous unsolicited aid by Prof. Izumi and his colleagues, Prof. Shozo Masuda and Prof. Toshihiko Sono. Through their PREFACE advice, we were accompanied on trips outside the Tokyo area by one of their senior graduate students, Mr. Hiroaki Okada, who served as an efficient guide and interpreter, and an amused informant on Japanese inns and outs. Our search for Valdivia-like pottery led up a few blind alleys and into several productive fields, and we gratefully acknowledge guidance and information from the following individuals: Prof. Sugao Yamanouchi, and Prof. N. Watanabe, Department of Physical Anthropology, University of Tokyo; Prof. Sosuke Sugihara, Department of Archeology, Meiji University; Prof. Teruya Esaka, Department of Archeology, Keio University; Mr. Chosuke Serizawa, Tokyo; Prof. J. Edward Kidder, Jr., Archeology Laboratory, International Christian University; Prof. Kyoichi Arimitsu, Department of Archeology, University of Kyoto; Mr. Fuku- hara and Mr. and Mrs. Shirakiba, Department of Archeology, Tenri Museum; Mr. Yoshimasa Kamaki and Mr. and Mrs. T. Macabe, Kurashiki Archeological Museum; Prof. Teigo Yoshida, Institute of Comparative Education and Culture, University of Kyushu; Prof. Morimitsu Ushijima and Mr. Mitsuhiko Higashi, Kumamoto Municipal Museum; Prof. Matsumoto, Department of History, University of Kumamoto; Prof. Sadanori Kawaguchi, Goyokuryu High School, Kago- shima; and Mr. M. Furuta of Shimabara. Prof. Ichiro Yawata, Archaeological Laboratory, Tokyo University of Education, led us on a memorable visit to an inland Middle Jomon site near the town of Oomiyama. The warm welcome and open generosity of all these people in providing us with advice, assistance and freedom to take notes and photographs of anything and everything is beyond the power of words to acknowledge. We hope that they will receive some satisfaction from seeing how significant has been their contribution to the conclusions in this report. Financial support for the research has come from a number of different organizations, whose contribution we gratefully record: the American Philosophical Society for Penrose Fund Grants 2012 and 2370; the National Science Foundation for Grants G-9055 and G-15641 to the Institute of Andean Research for a three-year program entitled "Interrelationships of New World Cultures", under which we were included as Project J: Coastal Ecuador; and the National Science Founda- tion Cooperative International Science Activities Program (supplemental funds to Grant GS-37), for sponsoring the trip to Japan. Throughout the various periods of field investigation from 1957-1961, a large portion of the field expense was borne by the Museo "Victor Emilio Estrada". Individuals who deserve special thanks for aid in preparation of the monograph are Miss Judith Hill, Secretary of the former Division of Archeology, United States National Museum, who skillfully and uncomplainingly deciphered the rough drafts, improved the consistency of the style and format, and typed rapidly, neatly and efficiently the final copy of the manuscript; Mr. George Robert Lewis, Scientific Illustrator, of the former Department of Anthropology, United States National Museum, who produced the beautiful and accurate line drawings; Mr. Jack Scott, Head, Museum of Natural History Photo Lab, for production of excellent enlargements from negatives taken under varying conditions over several years; and Prof. Kazuo Terada, University of Tokyo, who translated statements from Japanese publications. As the first of a new format, this volume presented special problems to the Editorial and Publica- tions Division, Smithsonian Institution. We wish to express our gratitude to Mrs. Joan Horn and Mr. John S. Lea for their constructive suggestions, careful editing for consistency and accuracy, and forebearance with our many demands. To the Government Printing Office, we offer a word of admiration for the remarkably error-free setting of the text and tables, their speed of execution of each phase of the work, and their high quality reproduction of a wide variety of photographs into excellent plates. We have left until last the recording of our indebtedness to those Ecuadorian colleagues with whom we shared the excitement of discovering the early Formative cultures of coastal Ecuador and of reconstructing from fragments of pottery, stone and shell, long forgotten historical events: Francisco Huerta Rendon, Carlos Zevallos Menendez and Olaf Holm. The years we worked together under the leadership of Emilio Estrada are treasured memories to all of us-golden years beyond repetition or recall. The unexpected death of Estrada in November 1961, shortly following the final season of fieldwork, brought an end to many dreams, but one at least has developed in a manner he would have loved to see-the verification of his correlation, timidly proposed many years ago, between Valdivia and Jomon. His co-authorship of this report is not simply a tribute it is a position fully earned. BJM _ ., . . . . CE Smithsonian Institution Washington, D.C. June 22, 1964 Contents Introduction 1 Theoretical approach to analysis and classification 1 Theoretical approach to interpretation 5 Environmental characteristics 9 The Valdivta Phase 15 Description of sites and excavations 15 G-25: Punta Arenas 15 G-31: Valdivia 16 G-54: Buena Vista 20 G-84: Posorja 21 G-88: Palmar Norte 21 G-L-2 21 G-L-3 21 G-L-27 22 Data from other investigations 22 G-115: San Pablo 22 G-117: LaLibertad 22 The site sequence and its implications 23 Description of artifacts 26 Stone artifacts 26 Abraders 26 Blades or knives 26 Bowls 26 Choppers 26 Cores 27 Gravers 27 Grinding stones 27 Hammerstones 27 "Jaketown perforators" 28 Paint stones 28 Pebble polishing stones 28 Polished axes 28 Reamers 29 Saws 29 Scrapers 33 Sinkers 33 Fireburnt rocks 33 Chronological distribution of stone artifact types 34 Shell artifacts 37 Abraders and polishers 37 Beads 38 Bowl or cup 38 Disks 38 767-841-65 CONTENTS Page no Drilled clam shell pendants Drilled and shaped pendants Drilled pecten pendant Fishhooks 39 Fishhook blanks 40 Scoops, spoons or spatulas Unclassified worked shell Chronological distribution of shell artifact types Bone and teeth artifacts Deer antler awls Antler tip projectile point Fish bone awls Fish vertebra Awls of teeth from saw fish Chronological position of bone and teeth artifacts 42 Pottery artifacts 42 Pottery type descriptions 42 Punta Arenas Incised 43 Puntas Arenas Plain 43 San Pablo Plain 45 Valdivia Applique Fillet 45 Valdivia Broad-line Incised 47 Valdivia Brushed 51 Valdivia Carved 53 Valdivia Combed 54 Valdivia Cord Impressed 54 Valdivia Corrugated 56 Valdivia Cut and Beveled Rim 57 Valdivia Embossed 57 Valdivia Excised 58 Valdivia Fine-line Incised 60 Valdivia Finger Grooved 61 Valdivia Fingernail Decorated 62 Valdivia Incised 63 Valdivia Modeled 66 Valdivia Multiple Drag-and-Jab Punctate 67 Valdivia Nicked Broad-line Incised 68 Valdivia Nicked Rib or Nubbin 69 Valdivia Pebble Polished 70 Valdivia Plain 72 Valdivia Polished Plain 74 Valdivia Polished Red 76 Valdivia Pseudo-Corrugated 80 Valdivia Punctate 80 Valdivia Red Incised 81 Valdivia Red Zoned Punctate 81 Valdivia Rocker Stamped 82 Valdivia Shell Stamped 34 Valdivia Striated Polished Plain 84 Valdivia Zoned Incised 35 Unclassified Decorated 37 Trade pottery of Machalilla Phase types 37 The seriated ceramic sequence and its implications 37 Figurines 95 Type descriptions 95 Palmar Plain 05 Palmar Notched o^ Palmar Incised og Valdivia o< San Pablo 97 CONTENTS XL Page Buena Vista 98 Punta Arenas 100 Unclassified 100 Figurine stools 101 Chronological distribution and evolution of figurine types 102 Miscellaneous 107 Worked sherds 107 Unfired clay objects 107 Diagnostic features and period subdivisions of the Valdivia Phase 107 The Machalilla Phase 110 Description of sites and excavations 110 G-110: LaCabuya 110 G-112 Ill M-28: Machalilla Cemetery Ill Data from other investigations Ill Description of artifacts 112 Stone artifacts 112 Abraders 112 Blades or knives 112 Choppers 112 Cores 112 Gravers 112 Grinding stones 112 Hammerstones 112 "Jaketown perforators" 112 Paint stones 112 Pebble polishing stones 112 Reamers 112 Saws 112 Scrapers 112 Miscellaneous stone 112 Chronological distribution of stone artifacts 113 Shell artifacts 113 Abraders and polishers 113 Bead 113 Bracelet 113 Disks 116 Fishhooks 116 Fishhook blanks 116 Pendant blanks 116 Chronological distribution of shell artifacts 116 Bone and tooth artifacts 116 Pottery artifacts 117 Pottery type description 117 Ayangue Incised 117 Cabuya Black-on-White 119 Cabuya Finger Pressed Rim 120 Cabuya Plain 121 Chorrera Incised 121 Chorrera Plain 122 Machalilla Burnished Line 122 Machalilla Double-line Incised 123 Machalilla Embellished and Red Zoned 124 Machalilla Embellished Shoulder 124 Machalilla Finger Punched 126 Machalilla Incised 127 Machalilla Incised and Punctate 127 Machalilla Incised and Red Zoned 128 Machalilla Plain 129 Machalilla Polished Plain 129 TABLES Page Machalilla Polished Red 13ø Machalilla Punctate 132 Machalilla Punctate and Red Zoned ! Machalilla Red Banded 134 Machalilla Red Incised 136 Machalilla Striated Polished Plain 137 Chorrera Phase types 139 140 Rocker stamped Zoned red Zoned red and black Zoned punctate * " Unclassified decorated ' !' Trade pottery of Valdivia Phase origin * 4U Pottery of probable trade origin 141 The seriated ceramic sequence and its implications ' Figurines 144 Machalilla 144 Figurines of trade origin 144 Worked sherds 144 Disks 145 Scrapers 145 Diagnostic features and period subdivisions of the Machalilla Phase 145 Relative and Absolute Dating of the Valdivia and Machalilla Phases .... 147 Origin and Affiliations of the Valdivia and Machalilla Phases 157 Literature Cited 179 Appendix 1: Tables 1-17 183 Appendix 2: Skeletal remains from sites of the Valdivia and Machalilla Phases 219 By Juan R. Munizaga. Plates 235 Tables TEXT A. Correlation between vessel shapes of Valdivia Phase pottery types and generalized forms recognized for the Valdivia Phase 91 B. Temporal distribution and frequency of techniques of Machalilla Embellished and Red Zoned decoration 124 C. Temporal distribution and frequency of techniques of Machalilla Embellished Shoulder decoration 126 D. Combined temporal distribution of minor motifs of Machalilla Incised and Red Zoned and Machalilla Punctate and Red Zoned 129 E. Temporal distribution and frequency of techniques of Machalilla Punctate decoration . 134 F. Correlation between vessel shapes of Machalilla Phase pottery types and generalized forms recognized for the Machalilla Phase 142 G. Carbon-14 dates for complexes of the early Formative Period 149 H. Carbon-14 dates for complexes of the late Formative, Regional Developmental and Integration Periods 153 APPENDIX I 1. Frequency of species of mollusks in levels of G-31, Cuts A, F, and H, of the Valdivia Phase 183 Frequency of shell artifacts in Valdivia Phase excavations 186 Frequency of stone artifacts and natural stone in Valdivia Phase excavations .... 188 Frequency of kinds of faunal remains from Valdivia Phase excavations 189 ILLUSTRATIONS Page Frequency of bone and tooth artifacts in Valdivia Phase excavations 189 Frequency of pottery types in surface collections and stratigraphic and text excavations at sites of the Valdivia Phase 190 Frequency of decorative motifs of Valdivia Broad-line Incised 206 Frequency of shell artifacts in Machalilla Phase excavations 206 Frequency of decorative motifs of Valdivia Fine-line Incised 207 Frequency of decorative motifs of Valdivia Incised 207 Frequency of tetrapod and concave base forms and lobed rims in Valdivia Phase pot- tery irrespective of type 208 Frequency of decorative motifs of Valdivia Excised 208 Frequency of Valdivia Phase figurine types and alternative kinds of body and leg treatment 209 Frequency of generalized vessel forms of the Valdivia Phase 210 Frequency of generalized vessel forms and unusual appendages in sites of the Macha- lilla Phase 214 16. Frequency of pottery types in surface collections and stratigraphic excavations at sites of the Machalilla Phase 216 17. Frequency of stone artifacts and natural stone in Machalilla Phase excavations ... 218 APPENDIX 2 1. Measurements, indices, and module of individual crania comprising the Buena Vista series 225 Frequency of independently variable traits in the Buena Vista population 225 Measurements, indices and module of individual crania comprising the San Pablo series . 226 Comparative distribution of cranial index in populations of Cabezas Largas and Buena Vista 227 5. Distribution of individuals by two categories of cranial index in five coastal Andean series 227 Illustrations FIGURES 1. Northwestern South America showing location of geographical features and selected modern towns 10 Location of sites of the Valdivia and Machalilla Phases 12 Sketch map of G-25: Punta Arenas, a Period D site of the Valdivia Phase 14 Sketch map of G-31: Valdivia, occupied during Periods A-C of the Valdivia Phase . 17 Sketch map of G-54: Buena Vista, a Period C site of the Valdivia Phase 18 Profile of the bank at the north edge of G-54 showing layer of sterile dirt overlying the refuse deposit 20 7. Sketch map of G-88: Palmar Norte, a Period A-B site of the Valdivia Phase, showing extent of the refuse and location of excavations 23 Fragments of clay with twig impressions suggesting wattle and daub construction . . 24 Temporal distribution and frequency of species of mollusks in levels of G-31, Cuts A, F and H 26 Gravers from Valdivia Phase sites 29 Grinding stone fragments from Valdivia Phase sites 30 Hammerstones of the Valdivia Phase 31 "Jaketown perforators" from Valdivia Phase sites 32 Fishhook reamers from the Valdivia Phase 33 Sandstone saws from Valdivia Phase sites 34 Scrapers from Valdivia Phase sites 35 Valdivia Phase pebble sinkers 36 Temporal distribution of stone artifact types during the Valdivia Phase 37 XIV ILLUSTRATIONS Page Stages in shell fishhook manufacture 3" Temporal distribution of shell artifact types during the Valdivia Phase 40 Rim profiles and reconstructed vessel shapes of Punta Arenas Incised 43 Rim profiles and reconstructed vessel shapes of Punta Arenas Plain Rim profiles and reconstructed vessel shapes of Valdivia Applique Fillet 46 Rim profiles and reconstructed vessel shapes of Valdivia Broad-line Incised, Forms 1-4 . Rim profiles and reconstructed vessel shapes of Valdivia Broad-line Incised, Forms 5-10. 50 Rim profiles and reconstructed vessel shapes of Valdivia Brushed ->z Rim profiles and reconstructed vessel shapes of Valdivia Carved -"j Rim profiles and reconstructed vessel shapes of early Valdivia Phase decorated types . 55 Rim profiles and reconstructed vessel shapes of early Valdivia Phase decorated types . 56 Rim profiles and reconstructed vessel shapes of Valdivia Embossed 58 Rim profiles and reconstructed vessel shapes of Valdivia Excised 5) Rim profiles and reconstructed vessel shapes of Valdivia Fine-line Incised 61 Rim profiles and reconstructed vessel shapes of Valdivia Finger Grooved 62 Rim profiles and reconstructed vessel shapes of Valdivia Fingernail Decorated ... 63 Rim profiles and reconstructed vessel shapes of Valdivia Incised 64 Rim profiles and reconstructed vessel shapes of Valdivia Modeled 66 Rim profiles and reconstructed vessel shapes of early Valdivia Phase decorated types . 67 Rim profiles and reconstructed vessel shapes of Valdivia Nicked Broad-line Incised . 69 Rim profiles and reconstructed vessel shapes of Period C decorated types 71 Rim profiles and reconstructed vessel shapes of Valdivia Pebble Polished 73 Rim profiles and reconstructed vessel shapes of San Pablo Plain and Valdivia Plain . 75 Rim profiles and reconstructed vessel shapes of Valdivia Polished Plain 77 Profiles of Valdivia Polished Red base forms 78 Rim profiles and reconstructed vessel shapes of Valdivia Polished Red 79 Rim profiles and reconstructed vessel shapes of Valdivia Punctate 81 Rim profiles and reconstructed vessel shapes of Valdivia Rocker Stamped 83 Rim profiles and reconstructed vessel shapes of Valdivia Striated Polished Plain . . 85 Rim profiles and reconstructed vessel shapes of Valdivia Zoned Incised 87 Seriation of small cuts at G-31 on the basis of changes in pottery type frequency . . 90 Seriation of G-31, Cut J, Sections D and E on the basis of changes in pottery type frequency 90 51. Seriation of G-31, Cut J, Sections D and E with scale enlarged to show changes in frequency of minor early decorated types 90 52. Seriation on the basis of changes in pottery type frequency of Valdivia Phase sites and levels of relatively short-term occupation 90 53. Seriation of Valdivia Phase sites and levels selected to represent the general trends of change in pottery type frequency 90 Temporal distribution of vessel shapes of Valdivia Phase pottery types 90 Chronological distribution and period of maximum frequency of Valdivia Phase pot- tery types 93 Evolution of Valdivia Phase vessel shapes 94 Pottery types represented at G-115 and their period distribution in the Valdivia Phase sequence, based upon figure 55 95 Typical figurine heads of the San Pablo type 93 Typical figurine heads of the Buena Vista type 99 Valdivia Phase figurines of unclassified types 100 Unique Valdivia Phase stone figurine ^qi Two examples of a rare unclassified type of Valdivia Phase pottery figurine .... 101 Top, side and bottom views of two probable pottery figurine stools from the Valdivia Phase .jQ., 64. Temporal distribution of figurine types and details of body treatment during the Val- divia Phase .. (-,4 Evolutionary changes in figurine style during the Valdivia Phase ...... 106 Sketch map of G-l 10: La Cabuya, a Period C site of the Machalilla Phase '. Ill Gravers from the Machalilla Phase * * ~ Small hammerstones from the Machalilla Phase 113 Fishhook reamers from the Machalilla Phase 114 ILLUSTRATIONS Sandstone saws from Machalilla Phase sites 115 Snub nosed scrapers from the Machalilla Phase 116 Carnivore tooth perforated for suspension; from the Machalilla Phase 117 Rim profiles and reconstructed vessel shapes of Ayangue Incised 118 Rim profiles and reconstructed vessel shapes of rare Machalilla Phase decorated types . 119 Rim profiles and reconstructed vessel shapes of Cabuya Plain 121 Rim profiles and reconstructed vessel shapes of Chorrera Plain 122 Rim profiles and reconstructed vessel shapes of Machalilla Double-line Incised . . . 123 Rim profiles and reconstructed vessel shapes of Machalilla Embellished Shoulder . . 125 Rim profiles and reconstructed vessel shapes of Machalilla Finger Punched 127 Rim profiles and reconstructed vessel shapes of Machalilla Incised 128 Rim profiles and reconstructed vessel shapes of Machalilla Incised and Red Zoned and Machalilla Punctate and Red Zoned 130 Rim profiles and reconstructed vessel shapes of Machalilla Plain 131 Rim profiles and reconstructed vessel shapes of Machalilla Polished Plain 132 Rim profiles and reconstructed vessel shapes of Machalilla Polished Red 133 Rim profiles and reconstructed vessel shapes of Machalilla Red Banded 135 Rim profiles and reconstructed vessel shapes of Machalilla Red Incised 137 Rim profiles and reconstructed vessel shapes of Machalilla Striated Polished Plain, Forms 1-10 138 88. Rim profiles and reconstructed vessel shapes of Machalilla Striated Polished Plain, spouted jar Forms 11 and 12 139 Seriation of Machalilla Phase sites on the basis of changes in pottery type frequency . . 140 Temporal distribution of vessel shapes of Machalilla Phase pottery types 140 Chronological distribution and period of maximum frequency of Machalilla Phase pottery types 143 Fragmentary head of a Machalilla type figurine from M-28 144 Period distribution of pottery types of Machalilla Phase origins in Valdivia Phase sites and of Valdivia Phase origin in Machalilla Phase sites 148 Stratigraphic origin of carbon-14 samples from G-31 and G-54 151 Differences in agreement between period identification of selected sites and obsidian dates derived by three different scales for conversion of hydration layer thickness into elapsed time 154 96. Early Valdivia Phase ceramic features at Early and Middle Jomon sites on Kyushu and Honshu 159 97. Map of Japan, showing location of Jomon sites producing pottery resembling Val- divia Phase types 160 98. Carbon-14 dates for Jomon sites on Hokkaido and Honshu, with selected Valdivia Phase dates for comparison 161 99. Jomon and Valdivia Phase sherds with similar excised decoration 162 100. Similarity between rim profiles of Jomon (solid) and early Valdivia (outline) pottery vessels 163 Jomon and Valdivia Phase sherds with similar castellated rim treatment 164 Late Jomon stone and pottery plaques from Honshu sites with decorative motifs re- sembling those on Valdivia Phase pottery 165 103. The northern Pacific Ocean, showing direction and speed of principal currents, paths of cyclonic storms and the great circle route between Kyushu, Japan and the Guayas coast of Ecuador 168 104. Northwestern South America showing location of pottery complexes carbon-14 dated between 5000-4000 years ago and their possible derivation from the early Valdivia Phase 169 105. Occurrence of selected Valdivia and Machalilla Phase decorative techniques and ves- sel form elements in other Formative complexes of Colombia and Peru 170 106. Late Jomon stone figurine head from Sakaizaki Shell Mound, Kyushu bearing a slight resemblance to some examples of the Buena Vista type of the Valdivia Phase 171 107. Chronological position of certain Mesoamerican and Colombian complexes incorpo- rating decorative techniques and motifs resembling Valdivia and Machalilla Phase (Ayangue Incised) types 174 ILLUSTRATIONS Page 108. Northwestern South America showing location of pottery complexes carbon-14 dated between 4000-3000 years ago and possible routes by which they were spread ... 175 109. Northwestern South America, showing location of selected pottery complexes carbon- 14 dated between 3400-3000 years ago and possible routes of communication be- tween them *?'' Stereograph drawing of skull BV-5, G-54, Burial 8 221 Stereograph drawing of skull BV-6, G-54, Burial 9 222 Stereograph drawing of skull BV-8, G-54, Burials 1-7 223 Stereograph drawing of skull BV-11, G-54, Burial 8 224 Stereograph drawing of skull M-l, G-l 10, Cut 1 228 Sites reflecting the first occurrence of skull deformation on the coasts of Ecuador and Peru 231 PLATES Frontispiece: Valdivia Bay, looking from the Machalilla Phase site of G-110: La Cabuya north- ward toward the Valdivia Valley (upper left) showing typical overcast conditions during garua 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. Typical views of the Guayas coast. Typical views of the Guayas coast. Typical topography and xerophytic vegetation of the coast of Guayas Province. Views of San Pablo Salitre. G-25: Punta Arenas, a Period D site of the Valdivia Phase. Looking north toward the Valdivia Bay from the vicinity of G-31. The environment of the Valdivia area. G-31: Valdivia, the type site for the Valdivia Phase. G-31, Cut J at conclusion of excavation. Stratigraphy of G-31, Cut J, southeast face. G-54: Buena Vista, a Period C site of the Valdivia Phase showing topography and modern vegetation. G-54, Burials 1-7 during excavation. The environment of the Posorja region. Environment of Palmar Salitre. Flake blades or knives from Valdivia Phase sites, with the cutting edge downward. Miscellaneous stone artifacts of the Valdivia Phase. Pebble choppers from the Valdivia Phase. Valdivia Phase stone tools. Miscellaneous objects from Valdivia Phase sites. Fishhook reamers from Valdivia Phase sites showing uniformity in size and form. Miscellaneous shell objects from the Valdivia Phase. Shell artifacts of the Valdivia Phase. Shell ornaments of the Valdivia Phase. Shell fishhooks of the Valdivia Phase. Bone and tooth artifacts of the Valdivia Phase. Type sherds of Punta Arenas Incised. Type sherds of Valdiv Type sherds of Valdiv Type sherds of Valdiv Type sherds of Valdiv Type sherds of Valdiv Type sherds of Valdiv Type sherds of Valdiv Type sherds of Valdiv Type sherds of Valdiv Type sherds of Valdiv Type sherds of Valdiv Type sherds of Valdiv Type sherds of Valdiv Type sherds of Valdiv Type sherds of Valdiv a Applique Fillet, vertical parallel bands on body wall. a Applique Fillet, curvilinear and intersecting patterns on body wall. a Applique Fillet, variations in rim treatment. a Broad-line Incised. a Broad-line Incised. a Broad-line Incised. a Broad-line Incised. a Broad-line Incised. a Broad-line Incised. a Broad-line Incised. a Broad-line Incised. a Broad-line Incised. a Broad-line Incised. a Broad-line Incised. a Broad-line Incised. ILLUSTRATIONS Complete vessels of Valdivia Broad-line Incised. Type sherds of Valdivia Brushed. Type sherds of Valdivia Brushed. Type sherds of Valdivia Brushed. Type sherds of Valdivia Brushed. Type sherds of Valdivia Carved. Type sherds of Valdivia Combed. Type sherds of Valdivia Combed. Type sherds of Valdivia Combed. Type sherds of Valdivia Cord Impressed. Type sherds of Valdivia Corrugated. Type sherds of Valdivia Corrugated. Valdivia Phase decorated types. Type sherds of Valdivia Cut and Beveled Rim. Type sherds of Valdivia Cut and Beveled Rim. Type sherds of Valdivia Embossed. Type sherds of Valdivia Excised. Type sherds of Valdivia Excised. Type sherds of Valdivia Excised. Type sherds of Valdivia Fine-line Incised. Type sherds of Valdivia Fine-line Incised. Type sherds of Valdivia Fine-line Incised. Type sherds of Valdivia Fine-line Incised. Type sherds of Valdivia Finger Grooved. Type sherds of Valdivia Fingernail Decorated. Type sherds of Valdivia Incised. Type sherds of Valdivia Incised. Type sherds of Valdivia Incised. Type sherds of Valdivia Incised. Type sherds of Valdivia Incised. Type sherds of Valdivia Incised. Type sherds of Valdivia Incised. Type sherds of Valdivia Incised. Type sherds of Valdivia Incised. Type sherds of Valdivia Incised. Type sherds of Valdivia Incised. Type sherds of Valdivia Modeled. Type sherds of Valdivia Multiple Drag-and-Jab Punctate. Type sherds of Valdivia Multiple Drag-and-Jab Punctate. Type sherds of Valdivia Nicked Broad-line Incised, single line on exterior adjacent to rim or carination. Type sherds of Valdivia Nicked Broad-line Incised, single line on exterior adjacent to rim. Type sherds of Valdivia Nicked Broad-line Incised. Type sherds of Valdivia Nicked Broad-line Incised. Type sherds of Valdivia Nicked Rib or Nubbin. Type sherds of Valdivia Nicked Rib or Nubbin. Type sherds of Valdivia Nicked Rib or Nubbin. Type sherds of Valdivia Nicked Rib or Nubbin. Type sherds of Valdivia Nicked Rib or Nubbin. Complete vessels of Valdivia Nicked Rib or Nubbin. Type sherds and a complete vessel of Valdivia Pebble Polished, Variant A. Type sherds of Valdivia Pebble Polished, Variant A. Type sherds of Valdivia Pebble Polished, Variant A, with supplementary decoration by broad-line incision, or nicked broad-line incision. Type sherds of Valdivia Pebble Polished, Variant B. Type sherds of Valdivia Phase unpolished plain types. Type sherds of Valdivia Polished Plain. Type sherds of Valdivia Polished Red. Typical bases of Valdivia Polished Red. Complete vessels of Valdivia Phase plain pottery types. 767-841-6E ILLUSTRATIONS Type sherds of Valdivia Punctate. Type sherds of Valdivia Punctate. Type sherds of Valdivia Red Incised. Valdivia Red Incised, castellated rim and tetrapod bowl. Type sherds of Valdivia Red Incised. Type sherds of Valdivia Red Zoned Punctate. Type sherds of Valdivia Red Zoned Punctate. Type sherds of Valdivia Rocker Stamped. Type sherds of Valdivia Rocker Stamped. Type sherds of Valdivia Rocker Stamped. Type sherds of Valdivia Rocker Stamped. Type sherds of Valdivia Rocker Stamped. Type sherds of Valdivia Rocker Stamped. Type sherds of several Valdivia Phase decorated pottery types. Type sherds of Valdivia Zoned Incised. Type sherds of Valdivia Striated Polished Plain. Sherds of Valdivia Phase types from sites of the Machalilla Phase. Stone figurines of the Valdivia Phase. Stone figurines of the Valdivia Phase. Broken pottery figurines of the Valdivia and San Pablo types showing method of construction. Figurines of the Valdivia type, showing variable treatment of the common long bob hairstyle. Figurines of the Valdivia type, showing variable treatment of the long bob hair style. Figurines of the Valdivia type, showing variation in hair style. Figurines of the Valdivia type, showing variation in size, workmanship and stylistic detail. Bodies of Valdivia type figurines showing variation in surface finish and typical leg treatment (shown in profile). Figurines of the San Pablo type showing variation in hair style. Figurines of the Buena Vista type, showing variation in hair treatment. Large grinding stone from G-110: La Cabuya, a Period C site of the Machalilla Phase. Natural pebbles used as polishing stones from Machalilla Phase sites. Worked shell from Machalilla Phase sites. Shell fishhooks from the Machalilla Phase, showing range in size. Type sherds of Ayangue Incised. Type sherds of Ayangue Incised. Type sherds of Ayangue Incised. Type sherds of Ayangue Incised. Minor decorated types of the late Machalilla Phase. Type sherds of minor Machalilla Phase decorated types. Type sherds of Machalilla Double-line Incised. Trade sherds of Machalilla Double-line Incised from the Valdivia Phase site of G-54. Type sherds of Machalilla Embellished Shoulder. Type sherds of Machalilla Embellished Shoulder. Type sherds of Machalilla Embellished Shoulder. Decorated pottery types of the Machalilla Phase. Type sherds of Machalilla Embellished Shoulder, showing variation in size and shape of bosses. Type sherds of Machalilla Phase decorated types. Type sherds of Machalilla Incised and Red Zoned and Machalilla Punctate and Red Zoned. Type sherds of Machalilla Plain. Type sherds of Machalilla Phase decorated types. Type sherds of Machalilla Punctate and Red Zoned. Type sherds of Machalilla Red Banded, narrow variety. Type sherds of Machalilla Red Banded, narrow variety. Type sherds of Machalilla Red Banded, narrow variety. Type sherds of Machalilla Red Banded, wide variety. Machalilla Red Banded sherds from the Valdivia Phase site of G-54: Buena Vista. Type sherds of Machalilla Striated Polished Plain. Stirrup spouts of Machalilla Striated Polished Plain. Machalilla Striated Polished Plain stirrup and cylindrical spout fragments. Unclassified decorated sherds from Machalilla Phase sites. ILLUSTRATIONS Figurines and unclassified decorated sherds from Machalilla Phase sites. Worked sherds from Machalilla Phase sites. Jomon and Valdivia Phase sherds showing similar technique and motif in broad-line incised designs. Jomon and Valdivia Phase sherds with similar broad-line incised motifs. Jomon and Valdivia Phase sherds with broad-line incised designs of similar technique and motif. Jomon and Valdivia Phase sherds with incised decoration in similar technique and motifs. Jomon and Valdivia Phase sherds with similar incised decoration in vertical or horizontal zigzag motif. 165. Jomon and Valdivia Phase sherds with incised decoration combined with one or two rows of punctation at the base of the neck. Jomon and Valdivia Phase sherds showing similar combinations of motifs in incised decoration. Jomon and Valdivia Phase sherds with incised decoration in similar technique and motif. Jomon and Valdivia Phase sherds showing similar execution of zoned punctate decoration. Jomon and Valdivia Phase sherds with similar pseudo-corrugated decoration. Jomon and Valdivia Phase sherds with similar decoration by multiple drag-and-jab punctate. Jomon sherds from Yoshida site with multiple drag-and-jab punctate decoration. Jomon and Valdivia Phase sherds with similar shell combed decoration. Jomon vessels with combed decoration from Nanshu Shrine site. Jomon and Valdivia Phase sherds showing similar overall texturing by brushing or shell scraping. Jomon and Valdivia Phase sherds with patterned overall shell scraping. Jomon and Valdivia Phase sherds with similar decoration by finger grooving. Jomon and Valdivia Phase sherds with excised decoration in similar technique and motif. Jomon and Valdivia Phase sherds with similar types of decoration. Jomon and Valdivia Phase sherds with similar decoration by rocker stamping. Jomon and Valdivia Phase sherds with shell stamped decoration. Jomon and Valdivia Phase sherds showing similar lobed rim treatment. Jomon and Valdivia Phase sherds with folded-over plain or finger-pressed rim treatment. Jomon and Valdivia Phase sherds with similar fine-line incised and drag-and-jab punctate decoration. 184. Jomon and Valdivia Phase pottery with similar relief decoration in the form of a stylized face. 185. Jomon and Valdivia Phase sherds with similar decoration by cord impression, and Jomon vessels resembling Valdivia Phase examples. Jomon vessels of typical Valdivia Phase shapes. Pre-Jomon and Valdivia Phase stone figurines. Decorated sherds from Puerto Hormiga. Jomon and Valdivia Phase sherds with similar decoration by nicked and finger-pressed applique ribs. Jomon and Valdivia Phase sherds with similar nicked broad-line incised decoration. Decorated sherds from Kotosh. Skulls from the Valdivia Phase. Skulls from the Valdivia Phase. Skulls from the Valdivia Phase. Skulls from the Valdivia Phase. Skull of the Machalilla Phase, showing tabular erecta flattening and crowding of canine teeth. The Early Formative Period of Coastal Ecuador The Valdivia and JMachalilla Phages Introduction THEORETICAL APPROACH TO ANALYSIS AND CLASSIFICATION Archeology, unlike other scientific disciplines, has no universally recognized system of classification and nomenclature. Each investigator feels free to invent his own frame of reference, with the result that data are frequently presented in terms that are not easily compared. Since the philosophical approach under- lying a classification has a significant influence on the conclusions, it seems worthwhile to explain the point of view employed in this report and the rationale behind selection of certain kinds of information as significant. Any scientist approaching the material encompassed by his field of specialization must organize it into distinguishable categories whose functions and inter- relations can be observed, described and if possible explained. Depending on the discreteness of the phenomena and their susceptibility to alteration, his task varies from easy to difficult. A physicist, for example, has no trouble distinguishing a neutron from an electron; in structure and behavior they are defin- able with high precision. Nor is there any argument about whether carbon is different from nitrogen, which follows it in the periodic sequence of atoms. Carbon has several subvarieties, of which radioactive carbon-14 is well known to archeologists. No one argues that carbon-12 should not be separated from carbon-14 for purposes of analysis and description, nor is there any disagreement that both belong to the same larger category or type. Their behavior can be predicted; it is known with what other elements they will combine and in what proportions; and it is even known under what circumstances such com- binations are likely to take place. The biologist has a more difficult job. The types with which he must deal differ from those of the chemist or physicist in that, unlike the elementary particles or the chemical elements, they represent populations that vary through time and place. Car- bon-12 is identical in all respects today to what it was ten million years ago, while the modern horse is not only recognizably different from its Eocene predecessor but also variable within the limits of a small inbreed- ing population. Mutation, genetic drift and other factors operate under natural selection to produce a kaleidoscope of results that may make boundaries drawn between varieties and occasionally between species arbitrary (Dobzhansky quoted in Grant, 1963, pp. 317-318; Simpson, 1961, pp. 117-119). Simpson (1961, p. 152) has summarized the situa- tion recently: Among evolutionary species there cannot possibly be a general dichotomy between free interbreeding and no interbreeding. Every intermediate stage occurs, and there is no practically definable point in time when two infraspecific populations suddenly become separate species. Fortunately for the neon- tologists, the majority of living populations have either definitely passed that hypothetical point or are not yet close to it. Nevertheless speciation is actively occurring today, and many populations are in the intermediate stages of some, but reduced, interbreeding. Again, if there are distinct gaps between ranges of characters, it is sufficiendy probable that isolation is at least complete enough to warrant specific separation. There remain numerous doubtful cases where decision depends on the per- sonal judgment of each practitioner of the art of classification. To insist on an absolute objective criterion would be to deny the facts of life, especially the inescapable fact of evolution. The archeologist who attempts to classify potsherds is also confronted with a continually changing class of phenomena (cf. Vogt, 1960, pp. 19-21), varying geographically and temporally as a result not only of cultural differences respecting function, production, and style, but also of accidental inconsistencies in raw materials, differential skill of potters, their suscepti- bility to influence from exposure to other pottery 1 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 styles, and occasional individual deviations from typical forms and decorative patterns. The difficulty of arriving at generally acceptable criteria for sep- arating this continuum into a series of types has re- sulted in lack of uniformity in the approach to pottery classification not only between workers in different areas, but even among people dealing with similar material. A number of commentaries have been published on the theory of pottery classification, and authors differ especially on the precision with which ceramic types should be described. In our view, part of the problem comes from a difference in theoretical position: archeologists who see culture as an evolving con- tinuum will find it difficult to define precise categories; those who take a timeless view tend to look for absolute types, describable in very exact terms. Since we incline toward the first approach, our pottery type classification represents an effort to recognize divisions within the mass of material recovered from tests and excavations that can serve as a basis for reconstructing the culture and its history. In many respects our view coincides with the population approach that has come to prevail in the biological definition of types, and has been well stated recently by Simpson (1961, pp. 183-184): It has already been emphasized often enough that taxa are inherendy variable and that attention to their variability is essential in their description and necessary in their practical definition. That naturally demands taking into account all available specimens and involves the principle that no one specimen referred to the taxon is, for these purposes, any more important or any more typical than any other. Some specimens are of course more nearly average than others as regards particular characters in the sense of being nearer the mean, although this is rarely true of all characters of one organism. The mere fact that a valid average is recognized means that all specimens have been taken into account and none especially weighted. Its application to classification problems in archeology was discussed a decade ago by Ford (1954), to whom the reader is referred for more detailed explanation. When several thousand sherds, representing the contents of a level in one of the sections of Cut J at Site G-31: Valdivia, are spread out on a table, it is possible to distinguish several kinds of differences. Among them are decorated and plain sherds, polished and unpolished surfaces, red-slipped and unslipped surfaces, coarse and fine sand temper. A classifica- tion based on one or more of these differences can accommodate without any difficulty about 75 to 85 percent of the sherds. There will remain, however, a group of borderline examples, where the surfaces are neither unpolished nor well polished, where the temper is intermediate between coarse and fine, where a type of decoration normally applied to an unslipped surface is found on a red-slipped one, etc. These variants are an expression of lack of standardization in production, of differences in sources of raw ma- terials, of the individuality of the potters, of the instability inherent in all evolving systems. At a later point in time, some variants may become sufficiently well defined and frequent to warrant recognition as separate pottery types, in which case the problem of where to draw the line is between parent and offspring rather than between contemporaries. Although each pottery type is a combination of many kinds of traits (temper, color, surface treatment, vessel shape, technique and motif of decoration), these may be independent variables within the total ceramic complex. It is frequently observed that gray or tan surfaces may occur with coarse or fine temper, or that a particular technique of decoration may be applied to polished plain or polished red surfaces, or that polished red surfaces may be plain or decorated. The problem is to decide which combina- tions of traits make convenient descriptive entities and useful categories for chronological and compara- tive analysis. The principal subdivision we have made is between decorated and undecorated surfaces. Although this is an obvious basis for distinction, it is one that is often not made in ceramic classification. Failure to separate decorated from undecorated sherds is usually justified on the ground that it prevents sherds from a vessel decorated on only part of the surface from being divided between two pottery types. This argument can be ignored, however, if the purpose of the classi- fication is to deal with the multitude of sherds recov- ered from habitation refuse and to show change through time. Failure to separate decorated from undecorated sherds has a serious disadvantage in that it makes it difficult or impossible to ascertain the exact proportion of the pottery that was decorated, or the frequency of certain kinds of decoration at any particular point in time, both of which can be used as a basis for inferences about level of competence in ceramic technology, which in turn provides clues to level of sociopolitical organization (Meggers and Evans, 1958). This, in our view, is an advantage outweighing the strictly puristic superiority of a classification that attempts to deal with complete vessels. Looking first at the undecorated sherds, differences can be recognized in temper and surface finish. Preliminary classification makes as many distinctions as possible, first in surface finish (unpolished, striated polished, completely polished or red slipped) and then in temper (fine, medium or coarse beach sand, or crushed rock). Temporal variation in these cate- gories was tested by classifying several levels from different depths in an excavation, and those features WHOLE VOLUME INTRODUCTION that showed the greatest variation were selected as type diagnostics. In the case of Valdivia Phase plain types, surface finish is the principal classificatory criterion, with temper used to make a secondary distinction between two types with unpolished sur- faces (Valdivia Plain and San Pablo Plain). Sep- aration of the polished types into coarse and fine tempered varieties served to multiply the number of types and obscure the trend in surface treatment without adding to chronological information on temper differences illustrated by Valdivia Plain and San Pablo Plain. Decorated types have been viewed as the occasional application of decoration to the surface of plain vessels. Technique and motif of decoration have been conse- quently selected as the primary criteria of differentia- tion. This approach is justified by the fact that decorated polished surfaces, for example, show the same characteristics of color, temper and surface treatment as do undecorated polished surfaces, the only exception being that decoration was often ap- lied to surfaces at the better finished end of the range of variation. Surface finish is usually consistent within a decorated type. Temper is more variable, and reflects in general the trend of the Phase, in which fine sand is more characteristic in the early periods and coarse temper in the later ones. Since more than one decorative technique is sometimes present on a single vessel, a hierarchy of priority was established for classificatory purposes in which the rarer techniques were given preference. By this rule, a sherd bearing rocker stamping was classified as Valdivia Rocker Stamped, even if broad-line incision was incorporated in the design, since broad-line in- cision is common throughout the Valdivia Phase sequence whereas rocker stamping is both rare and temporally restricted. Such "cross-overs" are noted in the type descriptions where they occur, and serve to emphasize the associations between certain tech- niques to form a subcomplex characteristic of a particular time period. The alternative of separating out a third type, in this case one combining rocker stamping and broad-line incision, served to proliferate the number of minor decorated types without adding any new information of chronological or descriptive importance, and therefore was rejected. By the same criterion of decorative priority, a rare example aber- rant in surface finish is considered a member of the decorated type, such as a polished plain surface in Valdivia Red Zoned Punctate or a red-slipped surface in Valdivia Nicked Broad-line Incised. Many types have no such deviant members, and this is in itself an interesting situation, reflecting a higher degree of standardization or perhaps closer agreement between the archeologists and the potters on definition of the type. Differences in standardization can also be detected in vessel shape and rim profile. In Valdivia Phase ceramics, no two rims are alike, although they tend to cluster in groups with similar construction, shape and orientation. Certain forms are unique enough to be readily distinguishable, such as the folded-over rim. The majority, however, comprise a continuum in which subdivision may be arbitrary. Examples are the cambered jar rims of Forms 20 and 21, and bowls of Forms 5 and 7. Distribution about the norm may be relatively tight as in Form 4, or widely scattered as in Form 3, and rim profiles classified into these forms show these differences in range of varia- tion. In ceramic complexes of later phases, stand- ardization may be greater and rim profiles show little variation in form or diameter. The attempt to define vessel shape thus produces information useful for comparative analysis on different levels of interpre- tation. Another assumption implicit in the point of view that the ceramic complex as a whole is a developing continuum is that each type will exhibit change through time. This can be measured most easily by tabulating the frequency of decorative motifs and vessel shapes according to stratigraphic levels of excavated habitation refuse. The pottery type de- scription is thus an ideal average to which a majority of the sherds will conform in a majority of features, but which will not apply completely to all examples classified as belonging to the type. It should be noted that pottery types recognized in the classification bear no necessary resemblance to any subdivision the potter himself might have made. The unpolished and polished variants of Valdivia Rocker Stamped, for example, may have been con- sidered separate by the makers, while types that we have distinguished, such as Valdivia Incised and Valdivia Nicked Rib or Nubbin, may have been considered inseparable by them. In our view this is an inevitable situation. The archeologist can have no way of knowing what the potter's conception may have been, and any attempt to reconstruct it would certainly inject ethnocentric factors into the analysis that would distort the outcome as much, or perhaps more, than a classification developed independently by examination of the sherd remains. The attitude just described applies not only to the analysis of pottery, but to all kinds of artifacts. It underlies the decision to recognize four types of pottery figurines in the Valdivia Phase rather than several dozen, which could have been isolated descrip- tively. From a developmental point of view, variation reflects an evolutionary process resembling that SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 labeled by biologists and linguists as "drift", and the classification attempts to recognize steps at which drift has proceeded far enough to produce a readily observed difference in appearance. A typology of the pottery figurines based on finer descriptive detail, such as specific forms of hair treatment, was attempted but variability was so great and occurrence so erratic that it was abandoned. With a different theoretical orientation, such factors would not have seemed detrimental to recognition of more minute classifica- tory distinctions. Another consequence of the view that has just been expressed is selection of terminology. Since Valdivia Phase pottery is the product of unspecialized artisans and made principally for the producer's domestic use, it is highly variable in all aspects. Firing is poorly controlled, so that a small sherd may show a wide surface and paste color range. Polishing tends to be more complete on the upper wall than the lower. Rim profiles are unstandardized and curva- ture may not be consistent. In recognition of these facts, pottery type descriptions make liberal use of words like "generally", "often", "slightly", etc. Since surface color is the result of uncontrolled firing, it is not described with reference to a color standard. If a type description is to be useful for recording differ- ences within the pottery complex, for indicating the quality of the ware, and for reflecting technological competence, then it can best do so by employing a terminology in accord with the level of competence of the makers. A highly standardized ceramic complex should be described with high precision; an unstand- ardized one is not made more understandable by being so described. Another level of classification is the recognition of phases, or complexes of archeological materials that correspond to extinct societies or cultures. Although it is easy to recognize at certain times and places significant groupings of traits representing architec- ture, arts and crafts, burials and other kinds of re- mains, it is often uncertain whether such a complex equates with a tribal, political or linguistic unit on the ethnographic level, or is something altogether different. For this reason, the term "phase" seems preferable, since it has no connotation. From the standpoint of typology, definition of a phase in archeology might be compared with definition of a genus or even a family in biology. A genus is com- posed of a number of species, which are evolving at different rates. It may happen in the course of time that some species will become sufficiently divergent to be put into a new genus; on the other hand, an environmental change may cause extinction of the whole genus. Similarly, a phase is made up of a number of types of artifacts, some invariable and longlived, others changing rapidly or shortlived. In the course of time, change in all component types may be so marked as to require the recognition of a new phase. The problem is to create categories that will allow recognition of significant disconformities as well as more gradual kinds of transformation. A fine break- down or emphasis on small details may produce an unrealistic impression of lack of continuity. Cate- gories should be suitable for historical reconstruction, for evolutionary investigation, and for comparative analysis. It cannot be decided in advance whether four phases or ten will describe adequately the his- torical sequence in any particular area, nor is it useful to proceed from the postulate that 100 years is too short, or 2000 years too long a duration for a particular phase. Change does not proceed at a uniform rate; what is too long in one context may be too short in another. The important qualification is cultural continuity or unity. Ideally, the early and late por- tions of a phase should show a homogeneity that sets them apart from other contemporary or subsequent phases. Discussion of alternative approaches to establish- ment of the Valdivia Phase will illustrate some of the criteria employed to distinguish the phases that comprise the coastal Ecuadorian archeological se- quence. The Valdivia Phase lasted some 2000 years. As would be expected, a great deal of change can be recognized particularly in pottery decoration, and figurine style. In fact, if decoration were employed as the principal basis of classification, as it commonly is in other areas (e.g., Rouse in Venezuela, and Rowe and students in Peru), the seriated sequence would undoubtedly be separated into three or four phases. However, continuity in plain types, vessel shapes, settlement pattern, and other features is so great that we have chosen to regard the Phase as a unity that can be subdivided into four time periods. This per- mits us to contrast the Valdivia Phase with the Machalilla Phase, from which it differs totally in ceramic complex. The distinction between the Machalilla Phase and the succeeding Chorrera Phase is less absolute. A number of pottery types continue, others die out or make their appearance. In some respects, a Machalilla - Chorrera combined seriation chart is similar to that for the Valdivia Phase. However here a division into two phases is justified by changes in settlement pattern and subsistence dependence as well as in the ceramic complex. Alterations in vessel shape that are important horizon markers elsewhere in the New World include disappearance of stirrup spout jars and the introduction of annular based bowls. In short, a phase corresponds as far as WHOLE VOLUME INTRODUCTION possible to a grouping of traits not only of pottery manufacture, but also settlement pattern, subsistence, sociopolitical organization, burial practice, etc., which forms a unique entity with temporal persistence and geographical range. This entity can be observed in relation to other entities, shedding light on trade relations, acculturative pressures and other interphase relations, and establishing a basis for cross-dating. The technique of quantitative analysis of pottery fragments and seriation of surface collections or levels of stratigraphic excavations has been described in detail by Ford (1962), and it is of interest here to emphasize only two aspects of the process. First, every effort must be made to ensure the reliability of a seriated sequence on which interpretations are to be based. A combination of survey with surface col- lections, small stratigraphic tests and intensive ex- cavation of deep deposits is most likely to permit recognition of disturbing factors of cultural or natural origin that may affect the outcome of analysis in some of the sites. Differences between the picture presented by a ceramic sequence derived from a site inhabited a long time and another derived from seriation of sev- eral sites of short occupation are discussed in the section on the seriated ceramic sequence of the Valdivia Phase (pp. 87-89). Second, it must be understood that although pottery types because of their greater abundance are used as the primary indicators of change, the final seriation represents a compromise that does least violence to all kinds of chronological evidence available. The relative position of levels arranged first in terms of trends in plain types, may be slightly readjusted after analysis of decorated sherds to minimize disconform- ities that may appear. In general, such rearrange- ment alters the position of levels whose plain type frequency was too similar to suggest which should be placed earlier. Less frequently, further minor read- justments may be made after classification of rim shapes, figurines, or shell and stone artifacts. Dis- conformities must be analyzed in terms of the geo- graphical proximity of the sites and of potential local factors that may cause deviation from the norm. A sequence finally arrived at is not readily susceptible to alteration. When several carbon-14 dates are available, it should be possible to reconcile them with the seriated sequence, as is the case in the Valdivia Phase. If it is not, the validity of the dates rather than of the sequence is open to question. THEORETICAL APPROACH TO INTERPRETATION Archeology is the science of reconstructing the development and spread of past cultures from incom- plete and often random bits of direct and inferential evidence. It has frequently been pointed out that the data of archeology are a small and unrepresentative sample of the once functioning culture, the implication perhaps being that if surviving evidence were more complete the job of historical reconstruction would be greatly simplified. Less attention has been given to the fact, well documented by living cultures, that cultural change does not proceed at a uniform rate, either in isolation or in acculturation situations. Some traits are of fleeting popularity, while others endure for centuries; some diffuse rapidly, others diffuse erratically, popping up in widely separated regions; some spread with little modification, others take on drastically altered forms in different portions of the area of distribution. As a result, evaluation of archeological remains is not simply a problem of dealing with incomplete and unrepresentative evi- dence, but also of evidence modified to different degrees and in different ways at different points in time. Much of the lack of agreement between arche- ologists on the significance that should be attached to certain kinds of archeological remains results from absence of a uniform theoretical approach to evalua- tion of the effect of these variables. The seriousness of the situation makes it worth- while to look to biology, the scientific discipline whose subject matter is most comparable to anthropology, for possible clarification. Although biology and culture are two distinct categories of phenomena, the content and behavior of these phenomena are of a similar level of complexity. Biologists are confronted with a vast array of species, differing widely in struc- ture and in capacity to react to the external world; anthropologists are confronted with a vast array of cultures of differing complexity. Biologists must classify living and extinct plants and animals into meaningful categories that shed light on the process by which this diversity arose; anthropologists have attempted to do the same for cultures. Biologists study the relations of fauna and flora with each other and the physical environment in order to understand some of the principles underlying extinction, survival or modification of species; anthropologists have under- SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 taken acculturation and cultural ecological studies for similar reasons. Our problems are thus much alike, but the material under study has one major difference. Anthropologists are unable to free themselves com- pletely from anthropocentrism in dealing with cultural phenomena, with the result that they frequently resort to explanations like "genius" or "free choice" when confronted with an alternative implying cultural determinism. Biologists, on the other hand, do not feel impelled to attribute the development of the horse's hoof or a hawk's eye to anything but the operation of natural forces. With their advantage of greater objectivity, biolo- gists have made considerable progress toward un- ravelling the complicated fabric of evolution, thereby making it possible to suggest some of the techniques by which it was produced. Four basic or primary evolutionary forces are now recognized: mutation, gene flow (or recombination), selection and drift (Grant, 1963, pp. 149-151, 431). The first two pro- duce variation by introducing new elements or by altering the combination of existing elements; in culture their counterparts are invention (discovery) and diffusion (acculturation), which are recognized as fulfilling a similar role in producing cultural varia- tion (cf. Linton, 1955, pp. 661-2). The second two forces "sort out this variability and establish the variant types in new frequencies in a population" (Grant, 1963, p. 150); in other words, lead to the formation of subspecies, species and increasingly divergent forms of life. In cultural anthropology, these processes have received less attention, although drift has been recognized by linguists as an important mechanism leading to change in languages (cf. Vogt, 1960, 1964). Since three of the biological processes have cultural parallels, it is reasonable to suppose that natural selection may also operate in culture in a manner similar to its operation in biology, although anthropologists are concerned with the results of individual variability in capacity to behave in accord with the cultural ideal, while biologists deal with the results of genetic variability expressed in alterna- tive genes and alternative combinations of genes. Although generally thought of as a process of change, natural selection in biology is also a process for maintaining stability, with the result that a species may persist in nearly identical form (i.e., with so little alteration that it continues to constitute a single species) for millions of years. The survival value of this kind of selection is explained by Grant (1963, p. 213): Any population of organisms exists in a certain environment and must be fitted or adapted to live successfully in its particular habitat. If the environment remains stable and if the popula- tion has already arrived at a high state of adaptedness, the main effect of selection will be to eliminate such peripheral variants or off-types as arise by mutation, gene immigration, or recom- bination. A certain range of genotypes of proven adaptive fitness is thereby preserved from generation to generation. This form of selection, known as stabilizing selection, does not bring about evolutionary changes, but rather maintains an existing state of adaptedness. The operation of a similar mechanism in culture would account for the stability of many primitive cultures under special environmental conditions, and the failure of such groups to accept new traits when exposed to opportunities. The failure of agriculture to substitute for shell-fish gathering on the southern coast of Brazil until almost the end of the aboriginal period (Silva and Meggers, 1963, pp. 126-7), and the long survival of nonmaterial traits such as myths among wandering groups (Meggers, 1964, pp. 514-5) may be cultural examples of stabilizing selection. If it is true that maize was introduced around 1400 B.C. on the coast of Peru without causing any important change in cultural pattern (Kidder II, Lumbreras and Smith, 1963, pp. 92-3), stabilizing selection may have been the reason. The uniformity of a biological population depends on maintenance of a constant genetic composition. Because of random fluctuation by chance, the fre- quency of a certain allele at any given time will be greater or less than the statistical average. Such chance fluctuations are termed "genetic drift" and represent a potential source of great variation under certain circumstances. A biological example pro- vided by Grant (1963, p. 278) shows how drift might operate to alter flower color: If the gene A controls flower color in a plant, and if the various alleles determine a series of shades from blue to white, like deep blue, light blue, pale blue, and white, the large and poly- morphic parental population will comprise a variable mixture of individuals having the different shades. As a result of drift a fragment of this population may become homogeneous for one flower color. If drift occurs repeatedly in different segments of the original population, a series of derivative colonies might arise which are characterized by different flower colors in pure form. One daughter colony might be all white, another all deep blue, and still another all light blue. In other words, drift may be responsible for "diver- gence between different contemporaneous colonies derived from a common ancestral population" (Grant, 1963, p. 286). Such divergence may be rapid and striking, even when the colonies inhabit similar environments (Grant, op. cit., pp. 288, 459). The implications of such a theory for culture are intriguing. For example, attempts to relate the Valdivia Phase to other early ceramic complexes on the coasts of Peru, Colombia and Panama has brought out the fact that these differ widely from each other, although each possesses a few traits linking it with the WHOLE VOLUME INTRODUCTION Valdivia Phase (fig. 105). The operation of a mech- anism of cultural drift could account for such variety. As each biological population is made up of a pool of genes, each culture is made up of a pool of traits, and just as each gene has many alleles, each trait has many individual variations in expression. In a large population, individual differences become averaged out and their ability to effect marked alterations in the cultural complex is minimized. A colony, how- ever, would be composed of a small group of indi- viduals not likely to represent the total range of the parent culture. In pottery, this "sampling error" might result in rapid divergence either by selection of some decorative techniques and abandonment of others, or by diminution of the range of variation in surface treatment or vessel shape, or by a combination of changes decreasing the heterogeneity of the offshoot. The differences between the ceramic complexes of early Valdivia Phase, Puerto Hormiga, Monagrillo and Guafiape are what might be expected to result from a process of cultural drift operating in a way similar to genetic drift in biology. The existence of cultural drift not only helps to explain divergences between related but isolated cultural complexes, but the concept offers a guideline for assessing cultural connections by making differ- entiation rather than similarity the expected result. Complexes with common ancestry should share a certain number of general characteristics, but need not duplicate all or even most of their component traits. In other words, we should not expect to find even with complete preservation, reproduction of the parent complex in all its variety. On the contrary, existence of close correspondence can be interpreted as reflecting special circumstances, such as organized population movement, or a conscious effort to main- tain the former cultural norm. Drift cannot be used to postulate cultural connections where there is no evidence, on the assumption that change has pro- gressed to the logical extreme of complete alteration, but a combination of drift and selection can make understandable the "watered down" appearance of many cultural or ceramic complexes that have moved to new environments and become isolated from the parent group. Another interesting and potentially relevant bio- logical concept is that of parallelism, defined by Simpson (1961, p. 103) as "the independent occur- rence of similar changes in groups from common ancestry and because they had a common ancestry." This is distinguished from homology, the sharing of traits derived from a common ancestor, and con- vergence, the independent development of similar traits by unrelated groups (Oschinsky et ah, 1964). The strength of cultural tradition in determining ac- ceptance or rejection of new traits and the necessity for new traits to be compatible with continuing function of the whole culture are propositions long recognized in anthropology. As in biology, the result is a tend- ency to channel changes in a certain direction (cf. Kroeber and Kluckhohn, 1952, p. 189). Since new elements result principally from modifications and combinations of old ones, the chances of duplication will be greater in groups sharing a similar background than in groups without this common heritage. Ac- ceptance of the hypothesis that parallelism operates in culture as in biology consequently seems preferable to explaining resemblances as fortuitous, particularly when they appear in several groups with common ancestry. The appearance of nicked broad-line in- cision in Japan, Ecuador and Colombia subsequent to the presumed date^T of separation between the Jomon, Valdivia and Monagrillo ceramic complexes would fit such a hypothesis. Although it is possible to trace the origin and devel- opment of many archeological complexes, in other cases new and striking cultures appear to spring up suddenly without clear antecedents. A similar situa- tion in biology has led Simpson to the concept of quantum evolution (Grant, 1963, pp. 458-9; see also pp. 555-7): Simpson argued that the absence or rarity of fairly complete fossil series connecting new major groups of organisms witii their ancestral stocks would be difficult to explain if their population size were as large in the period of their origin as it was in their later history, when the fossil representation becomes more adequate. Furthermore, die geological time available for the divergence of the new major group from its parent stock requires a much more rapid evolution during the period of origination than during the subsequent period of expansion. These facts can be accounted for on die genetically plausible hypothesis that the new major groups-genera, families, orders, etc.-originate from small isolated populations undergoing rapid shifts from the ancestral to a new adaptive state, that is, by quantum evolution. Since cultures, like biological populations, must be adapted to the environment to be effective, and since cultural traits seem to be subject like genetic ones to drift, it can be postulated that cultural traits or com- plexes are also subject to quantum evolution. In biology, "quantum evolution is believed to be the normal process by which new major groups come into being" (Grant, 1963, p. 556) and the existence of a similar process in culture would account for events like the sudden rise of the Inca, or the rapid appear- ance of Mochica in Peru, or the quick florescence of the Bahia Phase on coastal Ecuador. Quantum evolu- tion might also explain the absence of antecedents for the Machalilla Phase, although portions of the Pacific coasts of Mesoamerica and South America are too little known as yet to rule out the possibility SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 of future discovery. In such instances, something happened to give one small local group an advantage over others and when equilibrium was reestablished, the cultural result was markedly different. Such rapid shifts are characteristically followed in the paleontological record by long periods of much slower change resulting from normal interaction between the four primary evolutionary forces, and a similar situa- tion is observable in the archeological record. Cultural evolution has often been contrasted with biological evolution by picturing it as a tree of inter- connected and reconnected branches while the bio- logical tree is shown with divergent and redivergent branching (e.g., Kroeber, 1948, p. 260). This dis- tinction is not completely accurate, however. Bio- logical evolution is not a simple process of increasing divergence. Biological lines may diverge only slightly and then run parallel for millions of years, as has the skunk cabbage in Asia and North America (Grant, 1963, p. 443). Or diverging lines may reconverge if they have not progressed too far for hybridization (op. cit., fig. 79). Biological evolution is a vastly complicated process, not because the principles by which it operates are numerous or particularly com- plex, but because few organisms are subject to their unobstructed operation. Constantly changing en- vironmental or ecological conditions are the rule rather than the exception, and change may alterna- tively favor one kind of adaptation and then another. The same evolutionary mechanisms may produce stability in one species, variety in another, and ex- tinction in a third. Since the essential element to biological survival is adaptation, the environment is the most influential "creative" force (Grant, 1963, pp. 117, 546-8). Although variation will arise in seemingly non- adaptive traits through chance fluctuation and drift, important or drastic alterations result from selection of traits made adaptive by changes in the habitat. These may be environmental (such as climatic change or transferral to a new environment) or social (such as competitive relationships with other groups). So important is adaptation to the survival of any orga- nism that some biologists suspect that even traits of no obvious adaptive value may in fact have an undetected adaptive aspect if they persist. Maintenance of adaptation is a complicated process, as Grant (1963, p. 270) has explained: The environment to which an organic unit must adapt is a complex of many different factors, physical, social, and biotic. Each separate factor may carry out its own selective processes separately. The adaptations created by selection for one aspect of the total environment are not necessarily useful, and may even be detrimental, in relation to other facets of the en- vironment. Furthermore selection is opportunistic in that it brings about adaptations to existing environmental conditions. Such adaptations may or may not be valuable to their possessors in future environments. The collective processes of natural selection, while they promote the formation of adaptations of diverse kinds, do not guarantee evolutionary success in the long run under what Darwin termed the complex conditions of existence. Indeed, for every gene allele, genotype, or species that is preserved by natural selection on account of its adaptive properties, many sister alleles, genotypes or species are ex- terminated by the same process. Culture, being the principal instrument of man's adaptation to his physical, social and biotic environ- ments, is subject to similar continued pressures. Cultural change can be seen as the result of selection of more adaptive traits, whether of technology, socio-political organization or other aspects of culture. From this point of view, social disorganization may be a reflection of loss of adaptive value for the cultural configuration, rather than a primary cause of cultural breakdown. "Choice" of one value system over another reflects its superiority in terms of integration with other aspects of the culture or in terms of adjust- ment to the physical world, rather than conscious or unconscious human preference. In fact, change is often in unpreferred directions, as is evident in orga- nized resistance to automation, desegregation and socialized medicine in the United States today. The differential survival of the two early Formative complexes on the coast of Ecuador can be analyzed in terms of their relative ability to adjust to a new environment, since two important changes appear to coincide approximately with the end of this period. One is the replacement of inlets and mangrove swamps by salitres, and the other is the introduction of a new subsistence basis in the form of productive agriculture. Both the Valdivia and Machalilla Phases were organized around a seafood gathering, hunting, and plant gathering subsistence pattern, possibly supplemented by incipient agriculture as it is known to have been practiced by contemporary groups on the coast of Peru. The Valdivia Phase had perfected its adjustment to the particular conditions of the southern Ecuadorian coast over nearly two millenia, and the relatively slight alteration in com- munity size, settlement pattern, or technology that seems to have taken place during this time implies a high state of adaptation to existing resources, main- tained by stabilizing selection. The Machalilla Phase appeared late on the scene, and may have been pre- vented by the time factor alone from developing as intimate an adjustment to local conditions. In addition, location of Machalilla Phase sites on the coast rather than adjacent to salitres suggests lesser dependence on whatever resources the former bays produced. Although such differences seem insignifi- cant, and in terms of the competition between the two WHOLE VOLUME INTRODUCTION Phases appear not to have given one an important advantage over the other, they (and perhaps others not evident in the archeological record) apparently allowed the Machalilla Phase to adjust to a changed environment whereas the less flexible Valdivia Phase became extinct. Valdivia Phase pottery, which might seem equally suitable for culinary purposes, dis- appeared almost completely, while elements of Machalilla Phase vessel shape and decoration can be traced in the subsequent archeological record of the Ecuadorian coast for hundreds of years. Since the people of the Valdivia and Machalilla Phases were apparently in friendly communication, it might be expected, other things being equal, that both would have participated equally in the transition to the new kind of life. Since they did not, the logical conclusion is that other things were not equal, and further that this inequality may have consisted in the differential ability of the two cultures to adapt to changed ecological conditions (cf. Simpson, 1964, pp. 250-251). It should not be necessary to conclude by remarking that these ideas are expressed only as suggestions that appear to offer fruitful leads for archeological inter- pretation. Although facts are the foundation of any hypothesis, facts are meaningless in themselves. Meaning can be given to them only by an under- standing of the processes of which they are the tangible manifestation. It is therefore of fundamental im- portance to attempt to discern the invisible patterns and processes that make facts meaningful. In sug- gesting that some of the hypotheses developed in biology may be applicable to cultural phenomena, we do not mean to imply that attempts to develop hypotheses directly from study of culture are worth- less ; on the contrary, it is only by this kind of analysis that the applicability of any hypotheses to cultural phenomena can be adequately evaluated. The great complexity of culture and our intimate involvement with it, however, makes objective appraisal tremen- dously difficult. It seems to us that objectivity can be increased by borrowing certain concepts developed by biologists, and fitting them to archeological data. To us, the fit looks very good so far, but as long as many things remain unknown, others may be mis- understood. Much work needs to be done before we can be sure that we understand what is happening in the present, much less what has happened in the past. ENVIRONMENTAL CHARACTERISTICS Known sites of the Valdivia and Machalilla Phases are limited to the coasts of Guayas and southern Manabi Provinces of Ecuador, a distribution that might be attributed to more intensive archeological survey were it not for the fact that it corresponds to a well defined environmental zone. Topography, climate, and consequently vegetation set it apart from adjacent zones on the north, south and east, and there is a good probability that the early Forma- tive Phases were adapted to the specific resources limited to this part of the Ecuadorian coast. Its characteristics can best be understood in the context of the larger picture, and since inferences about the spread of the early complexes also derive to some extent from the geographical and environmental characteristics of western Colombia, it will be appro- priate to sketch these briefly as an introduction. The environmental characteristics of the southern Ecuadorian coast result from its geographical location between the excessively wet Pacific coast of Colombia and the arid desert coast of Peru. The transition is accomplished in a series of sharply defined zones created principally by differences in topography and rainfall. The northernmost zone, extending from Panama southward to Cabo Corrientes on the central coast of Colombia (fig. 1) is mountainous and densely forested. The Serrania del Baudo, a coastal range separated from the Andean chain by the valley of the Atrato, rises abruptly close to the shore to a height of 1000 or more meters. The irregularities of the coast resolve into one small and two large bays. The smallest and northernmost is Humboldt Bay, which together with the valley behind provides the "first extensive habitable and cultivable land along the shore" (Murphy, 1939, p. 8). Broad beaches in this bay contrast with the beachless condition of this seg- ment of the coast, where forest typically extends to the water's edge. Continuing to the south, forest gives way to beach again at the larger Golfo de Cupica. In addition to providing the best natural harbor on the Pacific coast of Colombia, this bay is at the point where the Atlantic drainage is closest to the Pacific. According to Murphy (1939, pp. 9-10): In March 1871, Selfridge walked from the beach of Lim6n Bay to the headwaters of the River Napipi in an hour and a half. 10 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Figure 1 .-Northwestern South America showing location of geographical features and selected modern towns. WHOLE VOLUME INTRODUCTION 11 Restrepo reports, however, that the continental watershed between the River Chirichire and the Nemequeda, a tributary of the Napipi, can be crossed on foot in 20 minutes! The Napipi is a tributary of the Rio Atrato, which flows northward through wet lowlands to enter the Golfo de Uraba and the Caribbean coast of Colombia (fig. 1). Rounding another point leads to the Golfo de Tribuga, described as "filled with dangerous shoals, and studded with rocks" (Murphy, 1939, p. 12). Behind the shore, forested hills rise to 1300-1500 meters. Several short streams flowing from the high- lands do not provide access to the Atlantic drainage since the southward flowing Rio Baudo intervenes between the coast and tributaries of the Atrato system. Cabo Corrientes marks the boundary between the northern zone of high land and seasonal rainfall, and the central zone of low land and almost continual rainfall. Murphy (1939, pp. 13-14) describes this sector as follows: South of Cape Corrientes the mountains retreat from the sea and a broad, clayey, completely forested coastal plain, broken inland by transverse spurs, stretches all the way to northwestern Ecuador. Moderately elevated land at die coast is found at only few localities, such as Palmas Island, the adjacent mainland about the mouths of Magdalena and Buenaventura Bays, and the islands with bold red cliffs on each side of the entrance into Tumaco Road. This is the maritime Choc6, a flooded lowland of perpetual rains, of selva and morasses, of hundreds of streams, many or most of which pour into the Pacific through multiple mouths. The line between earth and ocean becomes tenuous, for the greater part of the shore is fringed with a maze of mangrove-covered flats and islands, separated by a network of esteros and grading into shifting bars and shallows, which in many places extend for miles offshore. The largest river is the Rio San Juan, which fraction- ates the Punta Charambira* into a maze of channels and small islands. Navigable for several hundred miles, it follows a southward direction along most of its course, turning sharply westward at the latitude of its mouth. From its headwaters, it is a short over- land trip to the headwaters of the Rio Atrato, the two river systems forming an inland waterway between the Pacific and the Caribbean (fig. 1). In addition to wetness created by the low elevation of much of the land, part of this zone boasts the highest rainfall in South America, in some places exceeding 10,000 mm. per year (Reichel-Dolmatoff, G. and A., 1961, p. 242). Approximately at the Rio Esmeraldas comes a sec- ond and more abrupt change in topography and vegetation, as well as a marked climatic shift back to an annual cycle between rainy and dry season. Land elevation increases sharply, and vegetation takes on progressively xerophytic aspects toward the south. Murphy (1939, pp. 17-21) describes the transition thus: 767-841-65 3 The general altitude of the land rises, and the shore becomes cliffy .... The trees are smaller than in Colombia; arbores- cent ferns, lianas, and orchids and other epiphytes tiiin out; Euphoribiaceae become prominent, and cactus makes its appearance .... As one proceeds past Cape San Francisco the transition from rain forest to aridity is at first gradual and then fitful .... From Cape Pasado to Salango Island the landscape becomes a patchwork of wet and dry areas, the latter either being at low altitudes or lying in the rain shadow of headlands that adiabatically extract the moisture of the sea winds on their southern slopes .... At Salango the verdant land ends, save for a few still smaller specks and gorges choked with rank and impenetrable tropical vegetation. Beyond Point Ayangue there are no more trees near the shore until the traveler has rounded the gaunt peninsula of Santa Elena as far as the mangroves, savanas and corridor forest of the Guayas .... This zone extends to the vicinity of Tumbes, on the north coast of Peru, where there is another sharp transition to barren desert that extends in a strip be- tween the foothills and the shore to about 30 degrees south latitude, near central Chile. In terms of floral and faunal resources, this zone probably differs more from the regions to the north than any of them do from each other. All, however, seem to offer par- ticular advantages or disadvantages for human exploitation. Before leaving this general survey, it is of interest to note that the Caribbean coast of Colombia is in cer- tain respects very similar to the Guayas coast of Ecuador. Not only is the vegetation xerophytic, but the shore is in a similar stage of development, offering active mangrove flats as well as dried-up old bays. Seasonal alteration between rainfall and drought is likewise characteristic. The southern coast of Ecuador, between the modern towns of Machalilla and Posorja, corresponds to the dryest part of the country. Aridity increases from north to south and from east to west until the Santa Elena Peninsula, where semidesert conditions prevail. Between Machalilla and Valdivia, low hills reach the shore, broken periodically by flat open plains at the mouths of rivers. Where subjected to pounding waves, hills are sliced off to produce nearly vertical towering cliffs (pis. 1, 2b) such as that occupied by the Machalilla Phase site of G-110: La Cabuya. Except at points, where fallen rocks cover the tidal zone, the beach at low tide is a broad expanse of firm brown sand with few stones or shells. Toward the south, cliffs are typically reduced in elevation and the beach becomes narrower, reflecting a steeper gradient offshore (pi. 2a). Typical of this region is evidence of rising shoreline in the form of salitres of varying extent (fig. 2), the majority barren of vegetation (pis. 3b, 4, 13, 14). Dead and dying mangrove indicates this process is still under way, enlarging the already extant salitre of 12 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Gulf of Guayaquil ;| lO'3^' \NWfl*JW>w^'; APPROXIMATE LIMITS OF MAJOR SALITRES A VALDIVIA PHASE SITE  MACHALILLA PHASE SITE ? MODERN CITY La Plata Island Figure 2.-Location of sites of the Valdivia and Machalilla Phases. WHOLE VOLUME INTRODUCTION 13 El Morro near Posorja. All of the known Valdivia Phase sites are on the margins of these salitres, sug- gesting that they at one time provided optimum conditions for a food collecting way of life. At present, they are firm surfaced during the dry season, becoming mucky or partly inundated in the rainy season as a result of their low elevation. Within historic time, increasing aridity has been noted throughout this area of the Guayas Province, but it is not certain whether this is a progressive or cyclical phenomenon. Around Play as there were large cattle ranches before the turn of the century; now cattle find little to drink and less to eat during the dry season. Vegetation classifications agree in designating the southern Ecuadorian coast at the present time as xerophytic, although differing slightly in terminology. Eyre (1963, map 6), whose large scale analysis allows only two categories for the coast as a whole, shows a strip extending from Cabo San Francisco to the Peruvian border as thorn forest, noting in the text (op. cit. p. 231) that it represents the more arid type known as semidesert scrub. In a more detailed study, Miller (1959, fig. 1) divides the Ecuadorian coast into 8 vegetation belts. Two of these, coastal xerophytes and tropical subxerophytic bushwoods, extend from Machalilla to Posorja. From Machalilla northward to almost the Colombian border, the vegetation is classified as tropical sup- pressed forest. Among the constituents of the xero- phytic vegetation are two species of particular interest. One is ceiba, or silk cotton, which produces pods filled with silky fibers; the other is a tall bush cotton with small bolls, which may have been collected for the manufacture of cord. Rainfall is variable regionally and annually. For example, the station at Ancon near the Santa Elena Peninsula (fig. 1) recorded 1,144 mm. rainfall during 1934 and 153.7 mm. in 1940 (Ferdon, 1950, p. 31). On the average, however, variation from the wettest to dryest parts of this xerophytic coastal Ecuadorian zone during a single year is 125 to 875 mm. (Miller, 1959, p. 184). The dry season lasts 7-11 months, with duration decreasing generally from south to north. From the Santa Elena Peninsula northward, the rainy season terminates in 1-2 months of fine, misty precipitation known as garua (ibid.). Average annual temperature range is 5.3 degrees Centigrade or less (Ferdon, 1950, p. 39). Subsistence and commercial fishing are today the principal activities of coastal residents. Although farming is possible a few kilometers inland in southern Manabi Province and about as far southward as the Rio Valdivia, most of the area is too dry for cultivation without irrigation. Since most of the rivers cease to flow above ground in the dry season, irrigation is not yet feasible. Subsistence fishing is done close to shore in shallow water with nets played out to enclose a segment of water and gradually drawn toward the beach. This is supplemented by collecting a variety of mollusks at low tide (Appendix 1, table 1). Com- mercial fishermen go to deeper water to secure shrimp, white sea bass, sail and sword fish, robalo, Spanish mackerel, tuna and bonito. Crabs, lobsters and tortoises are additional potential sea food resources. The most abundant variety of land fauna is birds, including doves and ducks which concentrate around water holes during the dry season. Pelicans, gulls, and white herons are common among larger varieties. These and other species must have been more abun- dant in the past, before the habitat was disturbed by inroads of civilization, and it is consequently note- worthy that bird bones are poorly represented in Valdivia Phase refuse (Appendix 1, table 4). Deer can still be hunted and were undoubtedly more abundant before the introduction of cattie. Large iguana are excellent eating, as are their eggs If the hunting area extended inland to the tropical forests of the Guayas Basin and interior Manabi Province, a larger variety of small mammals would have been available. However, absence of such faunal remains from the refuse suggests that it was not neces- sary to go this far afield for food, at least until the latter part of Valdivia Phase Period C (see p. 25; Appendix 1, table 4). The environment of the Guayas and southern Manabi coast was probably somewhat different 4000- 5000 years ago than it is today, but there have been no geological or paleobotanical studies on which to base reconstruction of its characteristics. It would be of interest to know whether the increasing aridity observed during the past century is a recent trend, a cyclical and seasonal change, or one that can be pro- jected farther back in time. In the latter case, it would be necessary to picture the climate during the Valdivia Phase as wetter, and the vegetation conse- quently more luxuriant. Of greater significance is the past condition of the salitres. Today, each rainy season adds a layer of fine silt, since the salitres offer the lowest place for the accumulation of runoff. Although they may be partially inundated at such times, the elevation is now out of reach of the ocean water. The absence of an escarpment along the shore, the gradual slope of the beach including the tidal shelf, and the slight elevation of the salitres above present sea level suggest that they were prob- ably never under deep water. It would appear that they were in the past marine inlets that remained full at all times, or were inundated at high tide and exposed at low. Another alternative is that they were overgrown wholly or partly by mangrove as is 14 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 the case in many parts of the Guayas estuary today. The fact that shell refuse in archeological sites of the Valdivia culture lacks any species of mangrove mol- lusks but does contain many species of shell identified as intertidal in habitat tends to favor the reconstruc- tion of the salitres as former shallow bays partly or entirely exposed by tidal action. Aerial photography of the region and detailed geological studies would help clarify many of these problems, which today remain incompletely solved. Figure 3. -Sketch map of G-25: Punta Arenas, a Period D site of the Valdivia Phase, showing extent of the refuse and location of excavations. The Valdivia Phase DESCRIPTION OF SITES AND EXCAVATIONS Ten sites along the coast of Guayas Province have been identified with the Valdivia Phase (fig. 2). Eight of these have been examined by the present writers; two are described by other investigators. All are located close to the shore, typically at the margin of salt flats or salitres whose formation indi- cates an alteration in shoreline subsequent to the close of the Valdivia Phase. Depth of refuse was greatest at G-31: Valdivia, and least in the southernmost sites, where condition of the cultural remains was also poorest. Several sites produced skeletal remains, attesting to their use as burial grounds as well as habitation areas. The deposit typically consisted of powdery soil containing large amounts of shell, sherd and other kinds of natural and cultural refuse. No evidence was found of walls, floors, or other kinds of structures, and no significant natural stratigraphy could be identified at any of the sites. Two of the sites, G-31 and G-84, were partly buried beneath refuse of later occupations, both dating from the Regional Developmental Period. The re- mainder produced only refuse of the Valdivia Phase. G-25: Punta Arenas G-25 is located on the peninsula formed by the intersection of the north shore of the Gulf of Guayaquil with the Pacific coast, reached by a trail off the main road between Play as and Posorja. The site is now separated from the beach by rough terrain, dissected by gullies and salt flats muddied from inter- mittent flooding during high tide in the dry season and under water during the rainy season. Where salt content is not too high, the surface is overgrown with a tangled mass of scrubby vegetation including cactus (pi. 5). Habitation refuse occupies a small natural rise connected by a narrow neck to a larger elevated area to the south (fig. 3), surrounded on the remaining sides by low mud flats. At the time of excavation in January, traces of rainy-season flooding were already evident along the north edge of the site, where elevation is about 25 cm. lower than on the east side (pi. 5a). On the opposite side of the flats, the terrain rises to higher elevation than that adjacent to the site. The surface of the site is extensively eroded, pro- ducing irregular gullying of the sides and leaving the surface liberally sprinkled with small fragments of badly broken and bleached marine shells. Exposure is most marked at the north edge, where the bank is nearly vertical and rises 1.5 meters above the flats (pi. 5a). The summit, an area of about 50 by 80 meters, is relatively level except for a small rise to the west. Vegetation cover was limited to small scattered clumps of low scrub. Habitation refuse is distributed around the margins of the knoll, leaving a sterile zone at the center (fig. 3). Preliminary tests in 1956 had identified the occupation as belonging to the Valdivia Phase, and four stratigraphic excavations, each 2 by 2 meters, were undertaken in 1957 to determine the depth and composition of the deposit in more detail. Cut A was placed in the north corner of the site, sufficiently inward from the bank to avoid the possi- bility of disturbance of the deposit by erosion or land- slide and redeposition. Level 0-20 cm. contained hard gray clay that broke into compact lumps under the pick. Shell fragments were abundant but sherds were sparse and badly eroded. Level 20-40 cm. continued to produce shell and sparse sherds, but was damper and slightly more sticky in composition. Sterile gray clay was encountered at 45 cm., becoming yellowish at 60 cm. 15 16 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Cut B was excavated 28 meters southeast of Cut A, in a section partly separated from the major portion of the site by erosional gullies. The soil was again heavy, compact, gray clay containing shell fragments. Sherds were infrequent and very badly eroded. The deposit became sterile at 35 cm. Cut C was begun 43 meters southwest of Cut B. Refuse and soil conditions were similar to those in Cut B, but the deposit was shallower and became sterile at 30 cm. Cut D was placed on the summit of the north bank, where conditions of refuse on the eroded slope sug- gested a greater concentration than in other parts of the site already tested. However, refuse proved to be equally sparse and sterile clay appeared at a depth of 40 cm. The poor drainage at G-25 and the acid condition of the soil caused extreme erosion of the surfaces of most sherds, making it impossible to identify the plain types whose principal distinguishing feature is surface finish. Seriation of the site in the Valdivia Phase sequence (fig. 52) is based consequently on relative frequency of decorated types, supplemented by evidence from vessel shape. No figurine frag- ments were encountered. G-31: Valdivia G-31 occupies a low spur tapering from the west end of the range of hills forming the southern bound- ary of the Valdivia Valley. Flat sandy beach ex- tends from the foot of the site to the shore, 125 meters away. The modern village of San Pedro lies immedi- ately to the south of the site (pi. 7b), while Valdivia is adjacent to the north (pi. 6). The coast forms a gentle curve, but the sweep is so gradual that it cannot be termed a real bay at the present time. The Valdivia River, a small, meandering stream 100 meters to the north, provided the closest source of fresh water. At present it flows into a narrow lagoon paralleling the beach on the northeast side of Valdivia village, and cut off from the shore by a low sand bar across which water passes in either direction (pi. 6). The valley is about 2.5 kilometers wide at the shore, but constricts rapidly so that 2.0 kilometers inland it is little more than a narrow pass (pi. 7a). Above this, width is erratic, including a series of wide level areas that to judge from the frequency of sites were attractive agricultural terrain during the Re- gional Developmental Period. The spur occupied by the site rises to a height of 12 meters above the adjacent level surface (fig. 4; pi. 8a). Except for a few small huts on the lower flank, the area is free of modern occupation. Aside from scattered brush, prickly pear cactus and scrub trees, the principal vegetation is agave planted in rows originally along property or field boundaries, but now run wild and completely covering some parts of the surface. Bare spots abounded with surface sherds, and it was possible to collect a large sample. This showed two occupations of the site, an earlier and more extensive one corresponding to the Valdivia Phase, and a later and smaller one representing the Guangala Phase (fig. 4). Valdivia Phase refuse ex- tended from the foot of the hill southeastward over a 12 meter high knoll and continued eastward along the ridge and up the slope behind to an elevation of 15 meters. Overall dimensions of the deposit were approximately 160 by 150 meters, narrowing to 70 meters at the east end. Depth of refuse was greatest at the north edge, where it extended below the present surface of the adjacent level ground. The Valdivia portion of the refuse was 3.80 meters thick here. At the upper edge of the site, thickness was only 45-65 cm. In the intervening area, distribution was un- even, and apparently correlated to some extent with the contour of the underlying surface. In the area occupied by Cut E, for example, no Valdivia Phase refuse appeared in the 80 cm. deposit. Testing to the east of Cut E also produced only Guangala Phase refuse, to a depth of 20-30 cm. In the vicinity of Cut I, Valdivia Phase refuse reached a depth of 1.30 meters. In 1956-57, nine small stratigraphic excavations were made: Cuts A, B, C, and D by Estrada in the northwestern part of the site, and Cuts E, F, G (pi. 6b), H and I, by Evans and Meggers in the eastern and southwestern portions (fig. 4). The results of Cuts A, B, F, and H formed the basis for the first ceramic sequence for the Valdivia Phase. Cuts C and D were in areas of shallow deposits, predominantly Guangala Phase in origin, and Cut E produced only Guangala Phase remains. Material obtained from Cuts G and I was deposited in the Museo Arqueo- logico "Victor Emilio Estrada", where it was subse- quently classified by Estrada (1958, Cuadro 2). A test pit near the summit, south of Cut I, and an ex- tension of the south side of Cut F produced a lame number of decorated sherds, which were added to the general site sample. None of the excavations showed any construction features or any evidence of disturbance with one exception. An intrusive Guan- gala burial was encountered at a depth of 60 cm. in Cut A, above the unmixed Valdivia Phase refuse. Although the soil over the site varied slightly, at the time of excavation during the dry season, 'it was typically gray sandy clay of fine, powdery, flourlike consistency throughout the refuse deposit. Sterile soil at the bottom of the cuts was characteristically limey, very compact, hard clay containing 30 percent calcium carbonate by volume, which upon exposure WHOLE VOLUME THE VALDIVIA PHASE 17 Figure 4.-Sketch map of G-31: Valdivia, occupied during Periods A-C of the Valdivia Phase, showing extent of die refuse and location of excavations. The dotted line delimits extent of refuse deposited by a later Guangala Phase reoccupation of the site. dried into an indurated clay of the hardness of low- grade mudstone. Color varies from whitish to yellow or light brown (this deposit also underlays the refuse at G-54, and is easily recognized in the photograph of the bank profile, pi. 11 a). All of the cuts produced large amounts of shell, fireburnt and unburned stones, sherds, fish and animal bones, crab claws and carapace fragments. Everything recovered from Cuts A, F, and H was saved for analysis and samples of shell and animal remains were kept from other excavations. During January 1961, a large trench was excavated at the northwest edge of the site, where previous work had shown the refuse to be deepest and to represent the earliest part of the occupation. The excavation as a whole was designated as Cut J (fig. 4; pis. 8b, 9). It consisted of five squares 5 by 5 meters, designated as Sections A-E, excavated in arbitrary levels of 30 cm., making a trench 5 by 25 meters long. A sixth 5 meter square (Section E) was added at the north- west side of Section F in the hope of recovering addi- tional fragments of a castellated rim vessel (pi. 103a, b) found near the northwest side of Section E. Each section was taken down to and into sterile dirt, and the completed trench clearly shows the sloping nature of the original surface (pis. 9a, 10). Refuse continued to a depth of 2.10 meters in Section A, 2.40 meters in Section B, 3.10 meters in Section C, 4.20 meters in Section D, and 4.40 meters in Sections E and F. The upper levels, containing Guangala Phase ceramics, were examined and only Valdivia Phase items or unusual objects were saved. The remainder of the deposit was sifted (pi. 9b), producing a tremendous quantity of sherds as well as shell, stone and other kinds of refuse. All sherds from Sections D and E were classified; from the remaining sections only decorated sherds were available for study. The field work was under the supervision of Julio Viteri who did not notice any unusual features in the deposit, except for rare small clusters of large rounded cobbles (pi. 9c, d), the significance of which is undetermined. 18 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 G-54: Buena Vista Survey of the Valdivia Valley during the 1957 season revealed a site of the Valdivia Phase along the south edge, about one kilometer inland from G-31 (fig. 5). Brief testing suggested that the deposit was superficial, and classification of a surface collection of 480 sherds led to correlation of the occupation with the latter part of the seriated sequence (Evans, Meg- gers, and Estrada, 1959, Tabla B). In December of 1960, during large scale excavations at G-31, the site was revisited by Julio Viteri, who discovered and excavated a group of burials adjacent to the road (fig. 5; pi. 12). His investigations showed that the deposit was in fact rather deep, but masked by a 25-30 cm. thick sterile deposit, apparently washed down as talus from the adjacent hillside during flash flooding. In order to obtain more information on the characteristics of the site and its chronological position in the Valdivia Phase seriated sequence, several stratigraphic excavations and test pits were excavated during the summer of 1961, by Evans and Meggers. G-54 is located at a bend in the river, where the 6 meter vertical bank is divided into two 3 meter high steps by a flat terrace 40 meters wide (fig. 5). The refuse extends from the edge of the upper terrace backward and up the flank of the hill for a distance of about 70 meters. Maximum lateral extent is about 50 meters. Recent distubance includes a road that runs from east to west across the widest part of the site, a house near the bank, and some agave hedges. A second house was constructed just outside the northwest limits. The surface is exposed on a large portion of the southern half, and cut by erosion where not protected by sparse xerophytic vegetation. Sur- face sherds occur only here. On the other side of the road, weeds and scrub growth obscure the ground, and surface materials were recovered only from the bank. At the end of the dry season in January, the Valdivia River was reduced to a broken series of i i i i i i 0 5 10 15 20 25 M Figure 5.-Sketch map of G-54: Buena Vista, a Period C site of the Valdivia Phase showing extent of the refuse and location of excavations WHOLE VOLUME THE VALDIVIA PHASE 19 stagnant pools, and the local water supply was derived from wells dug in the riverbed. Cut 1, 3 by 3 meters, was placed near the edge of the bank, approximately equidistant between the east and west limits of the site (fig. 5). Excavation was in 20 cm. levels. The dirt was sifted and only large, fire-cracked rocks and unworked shells were discarded in the field. Coarse, lumpy clay occupied the upper 30-34 cm. over most of the cut, dropping to 50 cm. at the east corner. This layer was largely the sterile talus wash (fig. 6; pi. 11a), but a few sherds and shells were discovered. Below, the soil became powdery in consistency, light tan to light gray in color, and contained abundant sherds. Shells were less common than at G-31, animal bones more frequent (Appendix 1, table 4). Granular medium gray clay appeared at a depth of 85 cm. at the north side, along a sloping surface that was 100 cm. below the original surface at the south side. The upper 20 cm. contained a few sherds. No unusual features were recognized in the refuse deposit. After comple- tion of the stratigraphic excavation, the north and west walls were cut back to enlarge the sample of figurines, decorated sherds, and artifacts of shell and stone. Material from the northeast extension of Cut 1 was retained in two levels, the upper 0-85 cm., the lower 85-120 cm. The north side of this abutted the south side of the test trench (fig. 5). Cut 2, also 3 by 3 meters, was located 18 meters south of the road (pi. lib), behind the area where a group of burials had been removed earlier in the season. The upper 25-30 cm. were sterile, hard, medium-gray clay. Below 30 cm., excavation was controlled in arbitrary levels 10 cm. thick. The refuse deposit, extending to a depth of 1.20 meters on the east and 1.30 meters on the west, consisted of light tan to gray, powdery soil. A thin lens containing burnt rock extending laterally for a meter at the center of the south wall was the only feature en- countered. Shells were less numerous than at G-31 and were identified as predominantly Anomal- ocardia, with a few oyster. Cut 2 produced evidence of burial pattern at two levels. An inverted Valdivia Incised jar, encountered in Level 50-60 cm., contained badly deteriorated fragments of bone, apparently representing an infant. The remains of an adult appeared at a depth of one meter in the northwest corner. Details are provided below, designated as Burial 9. Subsequent to com- pletion of the stratigraphic excavation, the west half of the north wall of Cut 2 was cut back to increase the sample of decorated sherds and other artifacts. Additional excavations included two trenches and a test pit. One trench was located between the north side of Cut 1 and the edge of the bank and another 767-841-65 4 at the east side of Cut 2 (fig. 5). A 1.5 by 1.5 meter test at the upper (south) edge of the site established the depth of the deposit here as about 75 cm. Only rims, decorated sherds and artifacts were saved from these excavations as the digging was not controlled by levels nor the refuse sifted. During the excava- tions, local residents and their children devoted con- siderable energy to digging into the face of the bank, and all decorated sherds and figurines from this activity were also saved. Subsequently, the bank was straightened and cleaned to reveal the distinct natural stratigraphy, corresponding to the sterile upper and lower layers and the refuse deposit (fig. 6; pi. 11a). A vertical column of sterile dirt 56 cm. in width, joining the upper and lower sterile layers near the center of the cleared zone, seems best accounted for as a disturbance subsequent to laying down of the Valdivia Phase deposit. Seven burials were excavated in February 1961 by Estrada's field assistant, Julio Viteri, adjacent to the south side of the road. Details of position of the individuals and information on association is limited to what can be seen on the photographs (pi. 12). The bones were in poor condition and five of the skeletons were placed so close together that they are difficult to distinguish. Only three skulls were reconstructible, two adults (BV-1, 8) and one child (BV-14). Two (Burials 1 and 4) show the leg bones in sufficient detail to observe that they are tighdy flexed. Both in- dividuals are lying on the right side and facing toward the west. Arm position is less easy to recognize but in one case the right arm is extended, so that the hand is on or next to the thigh. A polished stone ax was adjacent to the lower arm (pi. 12b). Detailed description of the cranial characteristics is provided in Appendix 2 (pp. 219-224; pi. 192 a-c). Burial 8 was encountered 1 meter east of the edge of the area occupied by Burials 1-7, 25 cm. below the present surface. Skeletal fragments were distributed over an area 1.00 (east-west) by 0.30 meters, and portions of the cranium were widely scattered. Care- ful examination showed them to represent four indi- viduals (BV-5, 7, 9, 11). Original position could not be determined. No artifacts were associated, and the vicinity was free from habitation refuse. Burial 9 (BV-6; pi. 193 d-f), lying on sterile clay in the northwest corner of Cut 2, occupied an area 50 by 46 cm. The individual was again lying on the right side with the legs tightly flexed and drawn up toward the chest, and the arms extended toward the hips. The skull was at the west end. Sherds, shells and fireburnt stones were intermingled with the bones, apparently having filtered downward from the refuse above rather than associated as offerings. 20 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 NOIlVdOOOO ? 3SVHd tflAIOIVA 3 ^ c s -3 3* &.S -a .> CO -I ^3 > u V J-h 4-¯ n , ?>. a 0 5, O S sl ?X) m u C ^3 O a '& 5 ®e S 6 31VDS "IVJ.N0ZI80H QNV TV0IñH3A WHOLE VOLUME THE VALDIVTA PHASE 21 Burial 10 (BV-4; pi. 193 a-c) appeared 50 cm. northeast of Burial 8 at a depth of 40 cm. below the present surface, resting on sterile clay. Remains were badly scattered and deteriorated, and position of the body could not be ascertained. Sherds and rocks were intermingled with the bones, but appear to repre- sent accidental rather than intentional association. Burial 11 was encountered beneath a lens of brown coarse sandy soil during probing of the bank by local residents. Its position was not identifiable. A figurine found nearby was not clearly associated. Burials 12 (BV-10) and 13 (BV-3; pi. 194 a-c) were adjacent at a depth of 1.00 to 1.10 meters at the west side of the west extension of Cut 1, again lying on sterile clay. Condition of the bones and proximity of the skeletons were such that a clear separation could not be made. Burial 12 appeared to be lying on the right side, legs tightly flexed, facing toward the west. Burial 13 appeared to be on the left side, also with the legs tightly flexed, and with the arms extended at the side. The body was facing toward the north. Sherds were again wedged around the bones, but no in- tentional offering appeared to be associated. G-84: Posorja Most of the buildings comprising the modern town of Posorja are on low ground, along the water's edge. Inland, the elevation increases rather abruptly (pi. 13). Valdivia Phase remains were concentrated on the summit of the hill, superimposed by refuse belong- ing to the Jambeli Phase of the Regional Develop- mental Period (Estrada, Meggers and Evans, 1964, p. 489) and by existing houses. A stratigraphic excavation was undertaken here by Estrada in Janu- ary 1960. Arbitrary levels of 10 cm. showed Valdivia Phase refuse to extend from a depth of 0.40-1.30 meters. No further details are available on the site area, nature of the deposit, or size of the cut. G-88: Palmar Norte The Palmar salitre is a broad low area bounded by steep sided hills, resembling in general aspect the lower Valdivia Valley (pi. 14). G-88 occupies a flank at the north side, with refuse extending from the margin of the salitre up the steep slope and over the top of a small ridge projecting from the higher hill (fig. 7). Only a very small proportion of the site is level, and the steepness characteristic of most of the area suggests that houses must have been raised on posts. The site is shielded from the shore by a pro- jection of the hill. The salitre surface is largely barren, salt content of the soil being too high to permit growth of vegetation. Beyond its limits, xerophytic growth including thorny brush and cactus is charac- teristic and often dense. Habitation refuse was dis- tributed over an approximately circular area about 100 meters in diameter. Three stratigraphic excava- tions were made in July 1961 by Estrada, Evans and Meggers to determine the nature and depth of the deposit and to obtain a sample of the pottery and other artifacts. Cut 1, 2.5 by 2.5 meters, was located on a terrace- like portion of the east slope, at an elevation 8 meters above the salitre floor (fig. 7). Sterile soil, reminiscent of the talus wash situation at G-54, occupied the upper 30 cm. of the deposit. Level 30-50 cm. contained very powdery, soft, light gray dirt with few shells. Shells became slightly more abundant farther down, but in other respects no difference was evident. Sterile medium-gray clay appeared at 90 cm. on the east side over an irregular surface that dropped to 115 cm. in a small pocket at the north edge of the excavation. Cut 2, 3 by 3 meters, was placed at the sloping edge of the summit, about 6 meters above and northwest of Cut 1 (fig. 7), and was excavated in 30 cm. levels. Refuse was present from the surface, and also included fire-cracked stones and a variety of molluscs. Anomalo- cardia formed a lower proportion of the total shells than at G-54, and columella and conch shells were far more abundant. Soil was powdery and light gray, as in other Valdivia Phase deposits. Compact sterile clay appeared at 50 cm. below the surface on the downhill (southeast) side and at 85 cm. on the opposite side. Fragments of human bones were recog- nized just above the sterile zone on the southwest side, but nothing could be determined about condi- tions of burial. Cut 3, on the top of the ridge near the point, was begun as a 2 by 2 meter square. Soil was lumpy, medium-gray clay containing sherds, but only to a depth of 20 cm., with sterile clay beneath. Since the sample included types characteristic of the earliest part of the Valdivia Phase, the excavation was en- larged to cover a 4 by 5 meter area to obtain a more adequate sample of pottery. G-L-2 A low bank at the edge of Lagarto salitre (fig. 2; pi. 3b) produced badly eroded shell fragments and a few sherds, apparently of Jambeli Phase origin. In addition, half of a Valdivia Broad-line Incised bowl was discovered, of a style associated with Period D (pi. 42e). G-L-3 Shells scattered unevenly over a rounded hill de- fined a habitation site about 18 meters in diameter, and 2 meters in maximum elevation above the salitre 22 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 floor. The presence of mangrove oyster and charac- teristics of the paste of most of the sherds identified the principal occupation as Jambeli Phase in origin. However, one area on the east side produced a few sherds of Valdivia Phase types. Shells in the vicinity were predominantly clam rather than oyster. G-L-27 A few Valdivia Phase sherds came from a badly eroded knoll at the south edge of Lagarto salitre (fig. 2; pi. 3b), now reduced to 10 meters in diameter. Nothing could be determined regarding the original extent of the site. Data from Other Investigations Two Valdivia Phase sites have been reported by other investigators, both on the Guayas coast between the Valdivia Valley and the Santa Elena Peninsula. For convenience of reference, they have been incor- porated into the site numbering system for Guayas Province. G-115: San Pablo A large Valdivia Phase site was discovered at the margin of the San Pablo salitre (pi. 4) in 1956 by a group of Ecuadorian archeologists including Francisco Huerta Renddn, Carlos Zevallos Menendez and Olaf Holm. Extensive excavation during 1959 provided detailed information on the composition of the site and produced a tremendous quantity of pottery, figurines and other cultural remains. Final analysis is still in process, but a preliminary report gives infor- mation relevant to evaluation of the temporal position of the site in terms of the seriated sequence. The site is located about 3 kilometers from the present shore, on the inner margin of the salitre. At present, fresh water is available only from wells except at the height of the rainy season. Refuse covers a relatively large area but nowhere exceeds 1.30 meters in depth. Its general composition duplicates that of other Valdivia Phase sites, including presence of marine shell. A considerable number of burials were encountered during excavation, a situation most parallel to that at G-54. Metric data and indices of the skulls are included in Appendix 2 (p. 226). Condition of the skeletons is reported to be poor, but those from lower levels were sufficiently intact to show a flexed position. There were no burial goods (Zevallos and Holm, 1960, p. 11 and pi. 6, top). Fragments of human bones were also found scattered in the refuse closer to the surface, some of which are believed to show breakage prior to deposition (ibid.). A feature not reported for other Valdivia Phase sites is isolated finds of inverted complete vessels (Zevallos and Holm, 1960, pi. 4, 5, and 7, bot.). These are said to contain a yellowish substance with high organic content and occasionally shell spoons or unworked shells of species used for manufacture of fishhooks. Groups of shell spoons or unworked pearl oyster shells were repeatedly encountered superimposed (op. cit., p. 10). Tabulation and description of the large number of artifacts eventually will provide important additional information on the cultural inventory of the Valdivia Phase. Among objects mentioned in the preliminary report are shell fishhooks (Zevallos and Holm, 1960, pi. 25), spoons (op. cit., pi. 26), and disks (op. cit., p. 10), the latter possibly fishhook blanks. Stone artifacts include saws and reamers (op. cit., pi. 27); scrapers, blades and "Jaketown perforators" (op. cit., pi. 28); hammerstones, polishing stones, and grinding stones, some of the latter associated with manos (op. cit., pp. 7-9). More than 2000 pottery figurine fragments were recovered, representing the Valdivia, San Pablo and Buena Vista types. These show variations not repre- sented in the samples from G-31 and G-54, filling gaps in the evolutionary continuum on which the typology is based (pp. 104-106). There are also a few examples of stone figurines (op. cit., pi. 20), but the Period A Palmar types are absent. Decorated pottery types illustrated or mentioned in the text are: Valdivia Applique Fillet, Valdivia Brushed, Valdivia Excised, Valdivia Incised, Valdivia Broad-line Incised, Val- divia Modeled, Valdivia Nicked Broad-line Incised, Valdivia Nicked Rib or Nubbin, Valdivia Pebble Polished, Valdivia Punctate, Valdivia Red Zoned Punctate and Valdivia Rocker Stamped. A few rims represent jars of Form 14 with a folded-over, finger- pressed rim (op. cit., pis. 14-6, 7), and lobed bowls (op. cit., pi. 8-2, -5; 9-5, -8). G-117: LaLibertad In 1951, Bushnell reported a small group of pottery, shell and stone artifacts from La Libertad, which he attributed to a post-conquest date on the basis of European objects presumed to be associated. Iron- ically, this material belongs to the Valdivia Phase, making it not the latest but the earliest on the Ecua- dorian coast. The site is about 100 meters inland from the Engoroy cemetery, which produced remains of the Machalilla Phase (fig. 2). Details of the original refuse composition have been largely obliter- ated by later reuse, but the loose, dusty composition of the soil and inclusion of marine shells is in accord with the situation at other sites of the Valdivia Phase (Bushnell, 1951, pp. 123-4). The pottery includes Valdivia Applique Fillet, Valdivia Red Zoned Punctate, Valdivia Pebble Pol- ished and Valdivia Broad-line Incised (Bushnell, 1951, WHOLE VOLUME THE VALDIVIA PHASE 23 fig. 51-52). One rim bears a motif characteristic of Punta Arenas Incised (op. cit., fig. 52c). The illus- trated example of Valdivia Pebble Polished has broad- line incision superimposed on the pebble-polished upper wall, a rare treatment limited to the late part of the Phase. Rim profiles correspond to cambered jars of Forms 21 and 22 and a bowl of Form 8 (op. cit., fig. 54). The single pottery figurine (op. cit., fig. 52d) is atypical in style, but similar to some examples of the Buena Vista type in hair treatment and body form. Other artifacts believed by Bushnell to be associated with the pottery are shell scoops or spoons and a pitted hammerstone (Bushnell, 1951, fig. 52 k-1), both of which are represented at other sites of the Valdivia Phase. The Site Sequence and Its Implications The only two sites so far reported for Period A are G-31 and G-88, at the northern limit of the Valdivia Phase area of distribution (fig. 2). G-88 shows a short occupation dating from the first part of the Phase, perhaps representing temporary removal of a portion of the populace residing at G-31. The pattern of refuse at G-88 differs from that at G-31 in depth as well as continuity, indicating that it was either occu- pied by a smaller group that moved around on the site, or that it was intermittentiy abandoned. Period B is also best represented at G-31 and G-88, with a suggestion that inception of settlement at G-115 Figure 7.-Sketch map of G-88: Palmer Norte, a Period A-B site of the Valdivia Phase, showing extent of the refuse and location of excavations. 24 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Figure 8. -Fragments of clay with twig impressions suggesting wattle and daub construction. corresponds to the latter part of this Period. The bulk of the refuse at G-115, and all of that at G-54 belong to Period C, which is also represented by a brief reoccupation of G-31. G-117, on the Santa Elena Peninsula some distance to the south, appears to date from late Period G or early Period D. G-25 and the sites in the Lagarto salitre belong to Period D. The area, depth of refuse and geographical distri- bution of known sites indicates that the initial appear- ance of the Valdivia Phase was in the vicinity of the Rio Valdivia Valley, possibly at G-31. During Period A, knowledge of pottery making did not spread more than a few kilometers, an expansion that may reflect enlargement and subdivision of the popu- lation of the original settlement. Period C, by con- trast, seems to have brought a major alteration in both settlement pattern and artifact inventory. Sites are more extensive and slightly farther from the present shore, although not sufficiently distant to permit an inference that land food was becoming more important than sea food. Occupation during Period G expands southward and at the end of the period embraces the Santa Elena Peninsula. This represents the widest expansion of the Phase, sug- gesting further enlargement of the population. Dur- ing Period C, Valdivia Phase isolation was broken by the arrival of people of the Machalilla Phase, with whom peaceful relations appear to have been main- tained to judge from the closeness of the sites and the amount of pottery interchanged. Upset of this bal- ance marks the end of Period C. Valdivia occupation shifts to the Playas-Posorja region of the southern Guayas coast, where it is represented by shallow sites with sparse refuse, and a ceramic complex markedly inferior to that of Period G. The physical location is the same as during Period C, that is, along the margin of a salitre several kilometers from the present beach. Little indication of house type exists. Although the temperature is mild, there is a rainy season several months long, followed by a period of "garua" or light mist, which would make some kind of shelter desirable. Small lumps of clay with twig, stick, and grass impressions (fig. 8) were identified in the refuse at G-31, G-54 and G-88, suggesting constructions with wattle-and-daub walls. The steepness of the slope at G-88 makes it seem probable that houses were raised on piles. Although large waterworn cobbles were occasionally encountered, they were too few and scattered to indicate use of subterranean structures like those reported from early sites on the coast of Peru (Bird, 1948, p. 23). At best, they may have been used around hearths, since fireburnt and broken cobbles are common refuse components. The only other evidence of hearths was in the form of WHOLE VOLUME THE VALDIVIA PHASE 25 rare, small, thin lenses of finely powdered charcoal, appearing in the lower levels of G-31, Cut J, Sections D and E. Since there is abundant ethnographic evidence of dwellers of pile houses using a separate dirtfloored shelter for cooking, these two bits of evidence do not conflict. No conclusive picture can be drawn of changes in subsistence emphasis because of incompleteness of the evidence. All shell, bone and other organic remains from G-31, Cuts A, F and H were saved for classifi- cation and identification (Evans, Meggers and Es- trada, 1959, Tabla C). Samples of shell were retained from other G-31 excavations. All bone from G-54, Cut 1, and G-88, Cut 2 was preserved and identified. Unfortunately, no sample had been saved from the excavation at G-84, the only levels representing Period D, but it is probably reasonably safe to project most of the nonceramic characteristics of late Period C upward into Period D. Shellfish remains are the principal organic com- ponent of refuse at all Valdivia Phase sites. Classi- fication and analysis of species distribution by levels in cuts where the complete sample had been saved (Appendix 1, table 1) suggest little alteration in the pattern of exploitation of this food source (fig. 9). Shells fall into two classes: Pelecypodea and Gastro- podea, the former including clams, oysters, arks, scallops and similar forms, and the latter including conchs, snails, cowries and other univalves. Of the 38 marine species represented, all but 8 inhabit mudflats or intertidal zones. The remainder can be found just below low tide or in tidal pools. Shellfish could thus have been collected by wading among rocks and exploring flats at low tide, a practice still observable along the Guayas coast today (Ferdon. 1950, p. 19). Efforts to discern a change in emphasis on the species of mollusks collected are frustrating because of the erratic trends shown by many species (fig. 9). Most common throughout the sequence is the Venus clam {Anomalocardia subrugosa Sby), which constitutes more than 50 percent of the shells from all but the upper two levels. Second in frequency is the Horn shell {Cerithidea valida), which fluctuates wildly from level to level, but generally occurs in a frequency of about 14-18 percent. A few of the minor species show marked changes in popularity. One of the arks {Anadara tuberculosa) is more than twice as frequent during Period A as it is in Periods B and C, and even more marked decline characterizes one of the scallops {Aequipectan circularis). Spondylus {Spondylus princeps), on the other hand, is rare to absent in Period A, but relatively common during Periods B and C. Several other species appear not to have been ex- ploited during Period A, including two kinds of clams {Panamicorbula inflata, Mactra velata), two arks {Area pacifica, Glycymeris inaequalis), a chama (Chama echinata) and three gastropods {Cerithium adjustum, Natica chem- nitzil unifasciata, Oliva peruviana). Two gastropods {Bulla aspera, Trivia radians) are represented only in Period C levels. The land snail {Porphyrobaphe iostoma) appears in the upper level identified with Period A and shows a generally consistent increasing popularity to about 5 percent in Period C, but whether this can be interpreted as indicating any alteration either in environment or collecting habits is not clear. The habitat of the marine species covers the same range from intertidal to tidal pools or below low tide level throughout the seriated sequence, implying sta- ble collecting conditions and suggesting that rise of the shoreline was probably subsequent to the end of the Valdivia Phase. Particularly interesting in this respect is absence of any mangrove oysters {Ostrea columbiensis), a species abundantly represented in later coastal sites. Bone (including crab carapace) remains are less abundant, and consequently even more susceptible than mollusks to misinterpretation when subsistence trends are sought. Here, the time distribution is longer, covering all of Periods B and C (Appendix 1, table 4). The lower levels of G-31, Cut A were reduced in area because of excavation problems, and the sample from Period A is consequendy very sparse. Fish and crab remains are abundant throughout, a situation that undoubtedly can be projected backward into Period A and forward into Period D. Sea turtle and reptilian remains occur in small amounts in levels corresponding to Period B and Period C, and their absence earlier may be accountable to reduced size of the sample rather than lack of exploitation. Deer bones, on the other hand, show such a marked increase in frequency during Period C as to suggest that more emphasis may have been placed on terrestrial hunting at this time. Restriction of small land mammal and cameloid remains to the same portion of the sequence lends support to this interpretation, although inade- quacy of the sample from earlier levels cannot be ruled out of consideration. Bird remains were identi- fied only from late Period C. A few bones tentatively identified as dog are also restricted to this part of the sequence. Again, the rarity of these kinds of fauna casts doubt on the accuracy of attaching significance to the chronological position of the occurrence. Against this reservation is the coincidence of these distributions with an apparent change in subsistence orientation reflected in other faunal remains. Data now available permit only highly tentative interpre- tations about what may have occurred, but they provide suggestions for examination in the light of other kinds of evidence. 26 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY DESCRIPTION OF ARTIFACTS VOLUME 1 Stone Artifacts The natural soil components in the region where sites of the Valdivia Phase are located are indurated clay (sometimes called mudstone), low grade shale, micaceous schist and sandstone conglomerate. All other kinds of rocks found intermixed in the various layers of refuse had to be brought from nearby rivers and streams, the beds of which are strewn with well rounded cobbles and pebbles representing a wide variety of quartzite, chalcedony, quartz, jasper, opal, diorite, chert, basalt, and porphyry. At G-31: Valdivia, a source material for the peb- bles and cobbles is the Valdivia River only 300-500 meters to the north, while at G-54: Buena Vista the same river is even closer. With such easy access, it is not hard to explain the wide variety of waterworn pebbles encountered in excavation, some to be used intact, others to be chipped and flaked into tools, and others to be discarded without evidence of use. Al- though cores, chips, and flakes occur in considerable quantity in the various refuse layers (Appendix 1, table 3), there are only a limited number of intention- ally shaped artifacts. These have been classified into types and descriptions and are arranged in alphabetical order. Abraders Pieces of fine to coarse-grained sandstone, coquina and pumice have been roughly shaped and used as abraders. They are not naturally shaped, waterworn pebbles like the pebble polishers, but irregular frag- ments used for abrading or rubbing something that required a coarser and faster cutting action. In part, use has given the fragments shape; a few are roughly shaped by percussion. Form is variable and may be discoidal, rectanguloid, or irregular. Often several facets have been worn by extended use, and one side is usually slightly concave from serving as a whetstone. For comparison of cutting effectiveness, texture of the abraders can be rated in terms of size of particles in sandpaper. By these standards, range is from no. 0 to no. 4, with most of the abraders equivalent to no. 2 sandpaper. Size is highly variable, representative examples measuring as follows: 0.6-1.3 by 2.9 by 4.0 cm., 0.8 by 5.5 by 6.0 cm., 1.5 by 5.0 by 6.0 cm., 1.8 by 4.5 by 5.5 cm., 2.0 by 5.5 by 6.0 cm. The abraders were used to shape some of the worked shell and bone tools. Blades or Knives Plate 15 Irregularly shaped flakes struck off from cobbles or pebbles of chalcedony, quartzite, opal, and chert* were used as knives or cutting blades and apparently discarded after use so that they show no clearcut evidence of wear (pi. 15). All except those of quartz- ite are naturally very sharp due to the concoidal fracture of the rock, a fact that was apparentiy taken advantage of, since it is easier to knock off a few flakes from a cobble, cut a piece of meat, fish or hide and discard the flake rather than shape a knife or blade for repeated use. A flake blade or knife is distinguished from a flake scraper by absence of fine pressure chipping on the cutting edge. There is no observable wear on the cutting edge of the flake knives or blades. Flakes often preserve the natural waterworn area of the cobble or pebble on one surface. Shape is very irregular although many are clearly classifiable as long prismatic flake blades (pi. 15 q, r, u), struck off by percussion blows onto a striking plane on a core after a few percussion blows had cleaned off a face of the cobble or pebble for easy striking and fracture. Size ranges from 1.5 by 2.0 by 0.5-1.8 cm. to 4.5 by 7.8 by 1.5 cm. Bowls Plate 16 a-b Two well shaped andesite bowls, with flattened bottom, rounded sides and slightly incurving rim, were encountered during excavation of G-31, Cut J. One (pi. 16b), from Section B, Level 120-150 cm. equating with late Period B, is about 16 cm. in diameter and 11 cm. deep. The other (pi. 16a), from Section D, Level 270-300 cm. and representing middle Period A, is more symmetrical and has a more evenly ground and pecked surface. Diameter is about 23 cm., and depth is 7.8 cm. Choppers Plates 16c, 17 Some quartzite and chalcedony pebbles or cobbles have had a few large flakes knocked off by percussion blows producing an edge that is irregular and uneven but would cut effectively with repeated pounding. The waterworn butt end fits neatly into the cupped palm. The cutting edge appears to have been used repeatedly until dulled from battering. Size varies according to how much of the pebble or cobble has *Two obsidian flake blades mentioned in the first Valdivia report (Evans, Meggers, Estrada, 1959, p. 22) proved to be accidental intrusion from the superimposed Guangala Phase refuse when hydra- tion rim thickness was measured (Friedman, Smith, Evans and Meggers, I960, p. 508). h'tLtUYHUUb IHIVALVtb] SAbIKUTUUb SCALLOPS 1 0 10 20 30 40 50 % ^\ ?* a S a. Q. CUT H, 0-20 CM 20-40 40-60 60-80 80-100 100-120 CUT F , 0-20 20-40 CUT ri, 120-130 CUT F, 40-60 _CUT H, 13 0-140 CUT A , 160-180 180-200 200-220 220-240 i / i i i i i i i : i i ) i \i i i ) ) i i i i ) i -i-r \i i i i i i i i i i~r\ v I I I I i i i i i n ? IZZZU rr~n U.,31 i / / / I INTERTIDAL BELOW LOW TIDE OR TIDAL POOLS DEEP WATER Figure 9.-Temporal distribution and frequency of species of mollusks in levels of G-31, Cuts A, F, and H. Seriated sequence is based on changes in pottery type frequency (fig. 49). WHOLE VOLUME THE VALDIVIA PHASE 27 been struck off to produce an edge. Diameter ranges from 3.0-7.0 cm., length 5.1-12.0 cm. Cores Cores can be divided into two varieties: 1) portions of waterworn pebbles and cobbles from which flakes have been struck off leaving part of the natural surface intact, and 2) prepared cores with a striking platform from which small prismatic flakes were removed to make scrapers, blades and other cutting tools. Stone materials in the first category include quartzite, fine-grained andesite, chalcedony, opal, jasper; but the second category is restricted to chal- cedony and opal. Gores range in length from 3-10 cm. Flake scars along the faces measure 2-5 cm. long and 0.8-2.0 cm. wide, with most of the scars under 1.5 cm. wide. The stone working technique was not sufficiently standardized so that the striking platform was made at any regular angle. Rather, it resulted from knocking off several flakes to rid the pebble of its waterworn smooth edges and give a clean face and platform from which to strike flakes. The num- ber of clearly identifiable cores was small compared to the total quantity of stone in the refuse (see Appen- dix 1, table 3). Gravers Figure 10; Plate 18 a-j Small flakes of chalcedony, opal and fine-grained quartz of nondescript shape have a small point on one edge, usually opposite the bulb of percussion, suggesting use as an engraving tool. Examination of the points under high magnification binocular micro- scope shows wear along one side in all cases, and in half the cases wear along both sides of the small point. Some specimens also show wear on a longer edge suggesting additional use as a cutting or scraping tool. None of the flakes are large and none show regular shaping. The smallest are 2.0 by 2.8 by 0.3 cm., 2.0 by 2.5 by 0.5 cm., 3.0 by 3.0 by 1.0 cm.; the largest 1.0 by 5.2 by 1.5 cm., 4.8 by 3.3 by 1.3 cm., 3.8 by 5.5 by 1.1 cm., 6.0 by 2.5 by 1.3 cm. All of them have one or two short nibs or points, measuring 1.0-3.0 mm. long, with the length reaching 5.0 mm. on three specimens. Grinding Stones Figure 11; Plate 16 d-f Since all but one specimen are incomplete, and worn surfaces are either too small or too variable in con- tour (slightly concave to flat to convex) to permit reliable classification, fragments are all grouped to- gether in a single category of grinding stones. Amount of wear suggests grinding was a method of food preparation of the Valdivia Phase. The mano and metate fragments are of sandstone, quartzite, sandstone conglomerate, indurated clay, and coquina. The edges or back of several fragments have been shaped by percussion blows giving the artifact an ovoid or rectanguloid outline with rounded corners. One complete mano is a waterworn sand- stone conglomerate pebble slightly shaped by batter- ing around the edges and on upper surface so that it is oval. One side has a well-smoothed, convex sur- face from use as a mano. It is 11.0 by 9.0 cm. and 3.8 cm. thick at the center, tapering to 3.0 cm. thick at the edges. The remaining specimens are too frag- mentary for measurement, but contours suggest that neither the grinding slab (metate) nor the hand-held stone (mano) were large and both were generally rectanguloid. Hammerstones Figure 12 Hammerstones can be divided into two types: 1) unworked waterworn cobbles, and 2) cores. In the first category are natural pebbles and cobbles of quartzite, andesite, diorite, chalcedony, diabase, shale, sandstone and indurated clay (mudstone) used for battering or pounding. Mudstone and sandstone pebbles, so soft that a few blows broke them, probably represent hearthstones used because of accessibility and discarded after use. Repeated pounding some- times knocked off large spalls, creating a blunt cutting edge, and such tools have been classified as choppers although their first use may have been as hammerstones. Some hammerstones were bat- tered on one end and then turned over or around so that several parts of the surface show pitting from use (fig. 12f). The pebbles vary considerably in natural form, from flattened to spherical to egg-shaped to irregular, but all fit conveniently in the hand or between the fingertips. Diameter ranges between 1.7-5.0 cm., length between 4.0-15.0 cm., with the majority from 5-8 cm. long. Core hammerstones are pebbles that have had most or all of the natural surface removed by percussion flaking before being used for battering. The smallest in this group measures 1.3 by 2.2 by 3.2 cm. and is battered on both ends; the largest is 3.0 by 7.5 by 11.5 cm. Form is generally rectanguloid to tri- anguloid, and the absence of consistent shape or size suggests that rather than a delibertately shaped tool, this variety is a secondary result of the manufacture of flakes for knives or scrapers, in which the residual core was sometimes used for hammering. 28 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 "Jaketown Perforators" Figure 13 Several microflint specimens that looked like drills were so classified in the original Valdivia report (Evans, Meggers, Estrada, 1959, p. 22; fig. 13 a-d, 14 a-d). However, additional excavation at Val- divia Phase sites produced enough artifacts for reanal- ysis of the type. In the light of experimental work done by Ford and Webb (1956, pp. 76-79, fig. 25) on the Poverty Point materials, all the Valdivia artifacts originally classified as drills {perforadores) were examined under a high-magnification binocular microscope (fig. 13 c, f, i, 1, m). Only two showed circular polish striations on the tapering sides and point indicating use as a drill; on all others shape was the end product of use in cutting or working a material like bone or antler. The edges of the larger flake were chipped off from use creating a specimen shaped like an awl or perforator. The description of the experiment is so vividly expressed by Ford and Webb (ibid.) that it does not need repetition here except to indicate that the Valdivia Phase specimens are so much like the used microflint blades of Jake- town and Poverty Point in southeastern United States that the term "Jaketown perforators" has been ap- plied to the Valdivia specimens. The "perforators" are all flakes of chalcedony or opal, of colors ranging from mottled brown and white, to white, to creamy white, to gray, to black, to rusty rose. Length is 2.2-4.6 cm. The butt end, unmodified from use, is 0.8-1.8 cm. wide and 4-8 mm. thick; and length of the modified section, creating what looks like a tapered point, ranges from 1.0- 1.8 cm. The two specimens that show polish striations re- sulting from circular motions that might reflect use as a drill deserve additional comment. In both cases the form of the tool suggests shaping from use as a "Jaketown perforator". However, after the tool had acquired a long tapering point, it was secondarily employed as a drill. If all of the specimens had been shaped consciously for use as drills, the majority should show striations resulting from circular motion of drilling. Paint Stones Irregular, angular fragments of hematite are prob- ably the source of the red slip color. All the fragments are a brick red, rub off easily on the fingers, and crush easily. Pieces recovered from the refuse (see Appen- dix 1, table 3) measure 1.0 by 1.2 by 3.0 cm.; 1.0 by 1.0 by 2.4 cm.; 0.6 by 0.8 by 1.8 cm.; 1.0 by 1.0 by 1.5 cm.; and 4.0 by 5.0 by 3.0 cm. Pebble Polishing Stones Plate 19 a-j Well-rounded, smooth, waterworn pebbles of a variety of rock materials were probably used for polishing pottery. Shape may be hemispherical, spherical, egg-shaped, rectanguloid or irregular. Those clearly showing use were classified as polishing stones; others that could have been used a few times but not sufficiently to produce polishing planes were classified as natural waterworn pebbles (Appendix 1, table 3). Quartz, chalcedony, jasper, diorite, shale, mudstone, and even a petrified bone, are represented, suggesting selection for smoothness and convenience of shape. Several specimens have been used so much that a highly polished beveled edge has formed. The smallest polishing stones are egg-shaped to spherical quartz pebbles 1.5-1.8 cm. in diameter; this group usually measures 2.5-3.0 cm. in diameter, with some as large as 5.3-4.0 cm. The largest examples are flat waterworn pebbles measuring at the large end of the range 1.5 by 3.0 by 6.5 cm.; 0.9 by 4.0 by 7.5 cm.; and at the lower end of the range 1.0 by 1.5 by 3.0 cm.; 1.0 by 2.5 by 2.5 cm.; and 0.8 by 1.5 by 2.0 cm. Polished Axes Plate 19 o-r Three complete specimens were found in strati- graphic excavations at G-31 (two from Cut J, Section E, Level 3.00-3.30 meters, one from Cut J, Section B, Level 2.70-3.00 meters) and one in association with Burial 1 at G-54 (pi. 12b). Although no additional fragments or complete specimens appeared in any of the other Valdivia Phase sites excavated by Estrada Evans and Meggers, several are reported from G-115: San Pablo (Zevallos and Holm, 1960, p. 8, and per- sonal communication). All the examples are pecked and then polished from gray-black diorite. The two smaller axes are wedge shaped with a slightiy flattened, oval outline, battered poll and a fairly sharp, straight bit. One measures 5.5 cm. long, 4.8 cm. wide at bit, and 3.0 cm. wide at the poll (pi. 19p); the other is 6.0 cm. long, 3.5 cm. in maximum width, 3.0 cm. wide at bit and 2.5 cm. wide at poll (pi. 19o). One specimen shows incipient T-shaped form produced by slightly flaring nubbins at the poll, which would have provided grips to affix the ax to a handle. The bit is convex. This specimen (pi. 19q) measures 6.8 cm. to 7.5 cm. long, 4.5 cm. wide at bit, and 5.5 cm. wide at poll, decreasing to 5.0 cm. at a point 1.5 cm. below the expanded flattened poll. The fourth speci- men (pi. 19r), from G-54: Buena Vista, is a clearcut T-shaped ax with tapered sides and a flattened poll WHOLE VOLUME THE VALDIVIA PHASE 29 0 I 2 CM Figure 10.-Gravers from Valdivia Phase sites. similar to the previously described specimen but a less strongly curved bit. It measures 5.5-6.0 cm. long, 8.0 cm. wide at poll, 4.8 cm. wide at bit, and 5.5 cm. wide at a point 2.5 cm. below the poll where the sides of the T are distinctly formed. Reamers Figure 14; Plate 20 A series of artifacts identifiable as reamers for the manufacture of shell fishhooks, all show the same type of concentric wear from a circular motion. The specimens are of two materials, fine-grained sandstone and coquina, the latter composed of coarse, com- pacted waterworn shell particles cemented together. The majority resemble a plumb bob in shape, but a few are egg-shaped. Complete specimens range in length from 4.0-6.5 cm., with the majority between 5.0-5.5 cm. A cross section through the area of largest diameter of the specimens presents two dis- tinct shapes: circular or oval with flattened sides. The end is always circular. Diameters of circular bodies range from 2.5-3.5 cm., oval ones measure 1.5 by 2.5 cm., 1.2 by 2.0 cm., 1.2 by 2.7 cm., 1.2 by 1.7 cm., and 1.3 by 2.0 cm. Saws Figure 15 Thin slabs of fine-grained sandstone have been ground and polished to produce flat surfaces and one or more tapered edges that could have been used for sawing, especially such soft materials as shell. The slabs are the result of natural cleavage of sandstone, 30 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 a ? i i i-i 0 I 2 3 4 CM --??f--jir""':jiti'i&i'¯iiif ittaksisiaaB&tt/ iatafe v-^-;- afifa bm&mB&Wt&ilatm**' Figure 11.-Grinding stone fragments from Valdivia Phase sites, a-c, Mano fragments, d, Metate fragment. WHOLE VOLUME THE VALDIVIA PHASE 31 I I I ¯ 0 I 2 3 CM Figure 12.-Hammerstones of the Valdivia Phase. 32 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 which divides in flat parallel planes. In some cases a minimum amount of grinding was done to smooth the surfaces, in other cases cleavage marks are totally obliterated by abrasion. Thickness of the slabs ranges from 3-11 mm. with the majority 5-6 mm. thick. Many of the saws are fragmentary but they all appear to have been rectanguloid, with the upper edge or back sometimes battered with percussion blows used to work the slab into a more convenient shape. Of the more complete artifacts, the smallest and thinnest is 3.0 cm. long, 2.2 cm. wide at one end and 2.0 cm. wide at the other, and 3 mm. thick (fig. 15e). The largest is 6.5 cm. long, 7.0 cm. wide and 6 mm. thick. The best proportioned saw is 5.4 cm. long, 1.0 cm. wide at the one end, 1.4 cm. wide at the other, and 5 mm. thick on the back. The upper edge is roughly battered and pecked into shape. Cutting edges are typically straight, with bevels ranging from a sharp long taper to a blunt wide angle (cf. cross sections fig. 15a and 15b). One specimen has a slightly curved edge (fig. 15c). Many of the fragments of cut shell have grooves that could have been made by these saws and experi- ments with some of the saw fragments showed that they will cut shell as rapidly as a modern steel file or slightly used hacksaw blade. i i i 0 I CM c,f,ij-m CM Figure 13.-"Jaketown perforators" from Valdivia Phase sites. Worn end has been enlarged to show chipping resulting solely from WHOLE VOLUME THE VALDIVIA PHASE 33 Scrapers Figure 16; Plate 18 k-w Scrapers are usually made from cores or large flakes of quartz, quartzite, chalcedony, opal or chert. They are irregular in shape and show no consistent form. Core scrapers are made by reshaping a wedge- shaped core, usually ranging in size from 3.3 cm. long by 3.0 cm. wide by 1.2 cm. thick to 3.5 cm. long by 6.6 cm. wide by 2.2 cm. thick, a series of percussion blows along one side producing a rough concave or slighUy curved cutting or scraping edge. Viewed in cross section, most core scrapers may be called plano- convex, humpbacked, or snubnosed (fig. 16 a, b, e, g). In many examples, the curved face is the result of part of the natural surface of the pebble remaining after large flakes were struck off. The cutting edge varies in width from 1.8 to 5.5 cm. Flake scrapers represent unretouched flakes chipped by use so that the scraping side of the flake has an indentation or rounded notch (fig. 16 c, d, f, h). The classificatory distinction between the flake scraper and the flake blade or knife is the presence of fine chipping along one edge resulting from its use as a scraper, a feature clearly observable under a microscope. Flakes may show scraping wear on a single edge or several edges over a length of 2.5-5.5 cm. Notches are 8-12 mm. wide and 2-5 mm. deep. Flakes show considerable range in size, from 1.5 by 2.0 by 0.5-1.5 cm. to 4.0 by 7.5 by 1.5 cm. Sinkers Figure 17 Mudstone cobbles of similar size and shape have notches or grooves in the sides and/or ends suggesting that they were tied to fish nets or fish lines as sinkers. All are ovoid, flattened and thickest at the center, tapering toward the edges. Percussion-made notches typically occur near the middle on two or three of the four sides. On two examples, a narrow groove con- nects the end notches. In one case the end of the cobble shows battering as if it had been used as a hammerstone before it was notched and used as a sinker (fig. 17c). A groove 1 mm. wide has been worn beside the notch on one example from friction of the cord into the soft stone (fig. 17a). Sinkers range from 4.5-8.0 cm. in length, 3.2 to 6.0 cm. in width, and 1.0-2.5 cm. in maximum thickness. Fireburnt Rocks Scattered throughout the refuse of all Valdivia Phase sites is a large number of stone fragments burned or cracked from exposure to fire. These fire- burnt and firecracked stones are of three materials: sandstone and sandstone conglomerates, indurated clay (mudstone) and coquina. In excavations where all materials were saved and tabulated, sandstone and sandstone conglomerate was from two and a half to three times more common than mudstone, with only an occasional occurrence of coquina (Ap- pendix 1, table 3). Sandstone and sandstone con- glomerate fragments are all very friable and burntout, bright orange to red-orange in color, with angular irregular cleavage. Size ranges from large pieces a 0 I 2 CM Figure 14.-Fishhook reamers from the Valdivia Phase. 34 SMITHSONIAN CONTRD3UTIONS TO ANTHROPOLOGY VOLUME 1 I I I CM Figure 15.-Sandstone saws from Valdivia Phase sites. 15-20 cm. in diameter to smaller pieces 1-2 cm. of diameter. This situation suggests that hunks in sandstone and sandstone conglomerate were used around a hearth, where they broke as a result of repeated heating and cooling. Further crumbling probably occurred from trampling during daily living activity at the site. The indurated clay fragments are either broken- down waterworn cobbles brought from the riverbed or angular rocks from talus slopes nearby. Fire has turned the light gray-tan natural color to a light orange or orange-tan. The high calcium carbonate content of the mudstone (30 percent by volume) has produced an effect of slackening, creating a brittle and crumbly residue. In contrast to the irregular fracture of the sandstone, mudstone cobbles and fragments tended to spall off conchoidally as the heat penetrated inward. Fragments range from large hunks 20-25 cm. in diameter to smaller pieces 1-3 cm. in diameter. The coquina fragments are all from larger water- worn pieces with well rounded edges, but none show grinding abrasion. One side is fireburnt changing the natural white to cream color into orange, red or tan. Texture is very open and porous with particles of fine gravel and shells up to 1-3 cm. in diameter. Fragments vary in size from 1-15 cm. Chronological Distribution of Stone Artifact Types The attempt to analyze the chronological distribution of stone artifact types is hampered by the relatively small size of the classified sample and its limitation primarily to Periods B and C of the Valdivia Phase (Appendix 1, table 3). There are several reasons for this situation. Most important is the failure of nonceramic refuse material to be saved from excava- tions at G-84, which represents Period D, and from G-31, Cut J, Sections D and E, which provided the best pottery samples for Period A. The lower levels of G-31, Cut A, which extend into Period A, pro- vided a very small sample because of constriction in the size of the excavation with increasing depth as the walls were sloped to prevent collapse. Interdigita- tion of levels from G-31, Cut B added some infor- mation, but principally to Period B. Because of the small size and consequently somewhat sporadic distribution of the various types by levels, attempts at graphic presentation by percentage occurrence were not illuminating. An idea of the temporal duration of different types can be obtained, however, by joining in a vertical bar the earliest and latest occurrences of each artifact type as indicated by the seriated sequence of levels derived from pottery type analysis (fig. 18). WHOLE VOLUME THE VALDrVIA PHASE 35 b 0 I 2 3 4 CM Figure 16.-Scrapers from Valdivia Phase sites. 36 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 I I I CM Figure 17.-Valdivia Phase pebble sinkers. In spite of the small representation of most stone artifact types, the graph produces a picture of nearly total stability throughout the duration of the Valdivia Phase. The apparent inception of many artifacts near the middle of Period A rather than at the be- ginning can be explained as the result of inadequate samples from the earlier levels, and there is no reason to doubt that all extending this far back were present at the beginning of Period A. Sinkers, paint stones, saws, and choppers may be more recent additions, in view of their somewhat later appearance, but there is an alternative possibility that this more limited distribution may derive from greater rarity of these objects in the refuse. Since fishing was important from the beginning of the Phase, it seems improbable that saws, which are used in fishhook manufacture, and sinkers, which are associated with nets or lines, would have been absent from the early inventory. There is no change in the appearance of fishhooks in Period B that might correlate with alteration in their method of manufacture by introduction of saws at this time. The widespread occurrence of saws in preceramic sites of adjacent Peru and Chile also argues for their existence at the beginning of the Valdivia Phase. The restriction of paint stones to Period B may be also accidental. However, since this is the time of greatest use of red slip on both figurines and pottery vessels, it is possible that their occurrence reflects more intensive collecting of this raw material with resulting increased likelihood of some being lost in the refuse. It is also possible that paint stones had some other use during Period B, but at present this is only a matter of speculation. Choppers appear toward the middle of Period B and occur throughout Period C, probably continuing in Period D. Again, this distribution may be an accidental product of the small sample from early levels. On the other hand, it may be correlated with the increase in deer and other large mammal bones in the refuse during Period C, since choppers might have been a useful addition to blades for dismembering the carcass of a large animal. The chronological distribution of polished axes is not definitive. The sample is limited to only 4 speci- mens-three from Period A levels of Cut J at G-31 and one T-shaped ax associated with a skeleton at G-54, a Period C occupation-hence the bar on the graph (fig. 18) extends from Period A through C. This distribution suggests the possibility of evolution from a parallel-sided and slightly expanded form to a T-shape. The extreme rarity of polished axes in Valdivia Phase refuse in comparison with unshaped or slightly shaped stone tools may indicate they are of trade origin. On the other hand, polished stone WHOLE VOLUME THE VALDIVIA PHASE 37 Figure 18.-Temporal distribution of stone artifact types during the Valdivia Phase. Bars connect earliest and latest occurrence of each type when levels are arranged in the seriated order indicated by changes in pottery type frequency (Appendix 1, table 3). tools are often less abundant than cruder kinds in refuse deposits simply because they were more care- fully made and less readily discarded. Some of the alternative possibilities of interpretation that cannot be evaluated on the basis of evidence from G-31 and G-54 alone may be resolved when the larger sample of polished stone axes from G-115: San Pablo has been described. In summary, the Valdivia Phase stone artifact in- ventory can be characterized as an assemblage of crude and undistinctive implements persisting with rare exceptions throughout the seriated sequence without modification either in the form of individual tools or in the types represented. Since stone imple- ments had specific practical applications, this con- sistency implies that the forms were well adapted to the functions they served, and that these functions did not change in any major respect throughout the Valdivia Phase. Shell Artifacts Shell was rarely used for the manufacture of arti- facts, although abundantly available as a raw material. Of 17,612 shells collected during excavation of G-31, Cuts A, F and H, only 23 were worked. The majority of these are utilitarian objects such as fishhooks; beads, pendants and other kinds of ornaments are rare. For description, shell artifacts have been classified into types. Frequency and provenience are given in Appendix 1, table 2. Abraders and Polishers Plates 21 m-p, 22 a-b This category can be divided into two groups: A) artifacts cut from a thick piece of Spondylus princeps Broderip or conch {Strombus granulatus Wood or Strombus galeatus Swainson) and then abraded and polished into ovoid or rectanguloid form; and B) fragmentary or whole waterworn shells picked up from the beach and used for polishing or rubbing. Group A.-The consciously cut and abraded speci- mens show striations on all sides left by shaping and finer polish on several facets or edges resulting from use as a polishing tool. Several of the pieces are rectanguloid or ovoid (pi. 21 m, n), measuring in their maximum dimensions 3.0 by 2.8 cm., 2.5 by 2.5 cm., and 4.5 by 3.3 cm. Thickness is 1.6, 1.0 and 0.8-1.2 cm. One irregular fragment from the hinged end 38 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 of a Spondylus from G-31, Cut B, Level 3.00-3.20 m., shows clearcut evidence of cutting by a sandstone saw (pi. 21o). One edge is abraded, suggesting that shaping had not been completed. This specimen measures 4.7 by 4.0 cm., with the thickness varying according to the natural thickness of the shell from the hinge area to the back. A similar fragment, also of Spondylus, from the same Cut, Level 3.40-3.60 m. shows distinct saw marks on two sides (pi. 21p). It measures 1.0 by 1.2 by 4.0 cm. Group B.-Some of the waterworn shells cannot be identified as to genus and species, while others retain the distinctive colored lip of Spondylus. Except for one complete Olivella (pi. 22a), examples appear to represent bivalves. Outline is oval (pi. 22b) and measurements vary from 5.8 by 2.5 by 0.8-1.7 cm. thick to 4.0 by 1.5 by 0.7 cm. Use for polishing is suggested by the existence of higher polish on the tips and edges rather than on the rest of the waterworn surface. Beads Plates 23 m-r Eleven perforated disks probably represent beads. Eight are cut from flat portions of large shells, two from pearl oyster {Pinctada mazatlanica Hanley), one from Spondylus, and the remaining six insufficiently diagnostic for identification of species. Diameter is 1.1 to 2.8 cm. and thickness 1.5-3.0 cm. Perforations are biconical, decreasing from 3-4 mm. in diameter at the surface to 1-2 mm. in diameter in the interior. Three examples were made by cutting off the upper, spiral end of the Olive shell {Oliva peruviana Lamarck) and grinding the inside until some of the spiral channels were removed. The apex of the spire was rubbed until a perforation was produced. Diameter of this variety is 1.5-3.0 cm., thickness 4-6 mm. Bowl or Cup Plate 22 f A large waterworn fragment of a pearl oyster {Pinc- tada mazatlanica Hanley) was suitable for use as a shallow bowl or cup without further reworking. It measures 10 cm. in diameter and 1.5 cm. deep. The shell is 1.0 cm. thick. Disks Several disks of pearl oyster {Pinctada mazatlanica Hanley) appear to be too evenly smoothed and well shaped on all edges to be classified as shell fishhook blanks. They may be blanks for pendants or beads. Thickness ranges from 1.5-2.5 mm., diameter from 1.9 to 3.5 cm. Drilled Clam Shell Pendants Plate 23 a-b, d-h Ten clam shells, nine belonging to the species Dosinia dunkeri Philippi and one to Mactrella clisia Dall, have a perforation 1-4 mm. in diameter at the hinge end produced by grinding the beak, or umbo, sufficiently to break through to the interior. This process flattened the beak area and on several of the shells grinding extended to the adjacent surfaces, removing some of the natural ridges. Only one frag- ment does not have the beak completely worn through. However, this specimen and four others have perforations 1-3 mm. in diameter on the body, drilled conically from the exterior. On the specimen with no hole in the beak, the perforation is 1.0 cm. below the beak. On the others perforations are near the lower rim of the shell, 0.8-1.2 cm. inward from the edge. One specimen has two such perfora- tions instead of one. Although for purposes of classi- fication these objects have been called pendants, the extra perforations on the body of the shell in 5 of the 11 specimens suggest they may have been fastened to something. Overall size is controlled by the natural size of the clam shell; complete specimens measure 4.5 by 5.0 cm., and several fragments suggest a maxi- mum of about 5.5 by 6.5 cm. Drilled and Shaped Pendants Plate 23 c, i-k Pendants shaped from shell blanks or waterworn fragments and drilled at the upper end for suspension occur scattered throughout various levels of all periods of Valdivia Phase refuse (see Appendix 1, table 2). They are of three general shapes: rectanguloid, ovoid and trianguloid. The rectanguloid group is represented by two specimens. One, made of a fragment of Spondylus shell from G-54, Cut 1, Level 60-80 cm. measures 3.0-4.0 mm. thick, 4.0 cm. long, 2.1 cm. wide at the upper end and 1.2 cm. wide at the lower edge. The biconical perforation, 3 mm. in exterior diameter and 2 mm. in diameter at the center, was drilled 9 mm. from the upper edge. Another rectanguloid specimen from G-54, Cut 1, Level 80-100 cm. is of conch shell {Strombus) with the edges well rubbed and rounded. The natural contour of the shell creates a pendant that curves outward and inward sharply. It is 3.5 cm. long, 1.5 cm. wide, and 3 mm. thick. The hole, drilled from the exterior 5 mm. from the top edge measures 4 mm. in diameter at the surface, tapering to 3 mm. on the interior. Two ovoid pendants from G-31, Cut J, Section D, Level 3.00-3.30 meters, are waterworn fragments of conch shell {Strombus) that have been further shaped by rounding the edges. They are 2.5 cm. long, 2.0 WHOLE VOLUME THE VALDIVIA PHASE 39 cm. wide, and 3 mm. thick. A biconical perforation 3 mm. in maximum diameter is 7 mm. from the upper edge. Several other ovoid pendants from Cut J, Section E, Level 0.90-1.20 meters are waterworn fragments of shell without further working except for perforations 5-8 mm. in diameter at the upper end. Dimensions are 2.0 by 1.5 cm. and 2.5 by 1.5 cm. A unique pendant (pi. 23i) is a cutoff beak of a conch shell {Strombus) worked into a triangular shape that takes advantage of the natural contour of the shell to produce a pendant with a beak or nose pro- jecting from the exterior surface. The object is 5.3 cm. long, 4.6 cm. wide at the top, tapering to a rounded lower edge 1 cm. wide. The beak protrudes 3.7 cm. A single hole, drilled from the exterior on the lower part of the beak, measures 2 mm. in diameter at the surface, decreasing to 1.5 mm. at the opening. Layers of shell flaked off on the inside around the hole as a result of drilling. All edges are well rounded and vary in thickness according to the natural con- tours and thickness of the shell. Drilled Pecten Pendant Plate 23 l The flat valve of a small scallop {Pecten) is drilled from both sides near the beak, creating a perforation 2.5 mm. in diameter. No shaping or polishing was applied to surface or edges. Fishhooks Figure 19 e-g; Plate 24, h-q, Seventeen fishhooks or fishhook fragments came from Valdivia Phase excavations, eleven from various levels in the seriated sequence (Appendix 1, table 2) and three from the Northeast extension of G-54, Cut 1. All are of pearl oyster {Pinctada mazatlanica Hanley) and similar in general size, proportions, and method of construction. The few complete hooks and almost complete fragments give details of shape. Outline is not perfectly circular but slightly oval with a differ- ence of 2-5 mm. between the diameter from shank to point as compared with the diameter from point to back (see fig. 19g; pi. 24 h-j). The contour on the interior is more truly circular. The point is well smoothed and sharp, while the shank is flattened slightly to offer a better grip for the string. Thickness of the hooks ranges from 1.5-3.5 mm., with the ma- jority 2 mm. Maximum width of the back is 5-7 mm., tapering gradually in one direction toward the sharp- ened point and in the other toward the shank 3-4 mm. in width. Exterior diameters of complete hooks are 1.8 by 2.0, 1.8 by 2.2, 2.0 by 2.5, 2.3 by 2.5 and 2.5 by 2.8 cm. Fishhook Blanks Figure 19 a-d; Plate 24 a-g Worked pieces of pearl oyster {Pinctada mazatlanica Hanley) range in shape and size from roughly squared pieces showing the beginnings of careful shaping to circular discs with the center either unworked or 2 CM Figure 19.-Stages in shell fishhook manufacture, a, Roughed out blank, b, Initial perforation, c-d, Enlarged perforation. e-fy Nearly completed hook, g, Completed hook. 40 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Figure 20.-Temporal distribution of shell artifact types during the Valdivia Phase. Bars connect earliest and latest occurrence of each type when levels are arranged in seriated order according to changes in pottery type frequency (Appendix 1, table 2). SHELL ARTIFACT TYPES ay *œ z > DS HOOKS HOOK ANKS PED NDANT ADERS LISHE ERWOR OOPS M SHE NDANT _i TIN NDAN1 HAPED SCOOPS UJ w "'co XQ. 1 co a. 1 en 1 z 0 ? 1 o 1 (/) 1 BDIV I H ? en B ? " OD I 1 or 1 UJ I o_ 1 A I 1 m m partially drilled or reamed out. The blanks show beveled edges from cutting with sandstone saws. Sometimes cuts overlap each other, other times they are separated by jagged broken edges indicating that the shell was broken after a few cuts were made. The roughed out blanks measure 1.7 by 2.5 cm., 2.5 by 4.0 cm., 2.5 by 2.5 cm., 3.3 by 4.0 cm., up to 3.0 by 6.0 cm. Thickness is 1-4 mm. Circular blanks with- out perforations in the center or with perforations not completely reamed out range in thickness from 2-4 mm. and in diameter from 2.5-4.5 cm. The perfora- tions are 6-15 mm. in diameter. Scoops, Spoons or Spatulas Plate 21 q_-t Waterworn fragments of shells of various types were selected for their natural shape and brought to the habitation sites (Appendix 1, table 2). One specimen shows chipping along the edges to produce a more spatula-like form, but others are not worked and derive their scoop or spoonlike appearance from the natural curvature of the shell. Complete specimens measure 7.2 cm. long, 2.0 cm. wide at spoon end and 1.0 cm. wide at upper end, and 7.3 cm. long, 3.0 cm. wide at spoon end, and 1.3 cm. at upper end. The widest fragment is 5.0 cm. wide but only 6.8 cm. long. Although use is inferential, these objects have been singled out for description because they appear to be restricted to Period B refuse in spite of availability on nearby beaches at all times. Shaped scoops or spoons, cut from the shell of a large gastropod, are reported by Zevallos and Holm (1960, lam. 26) from extensive excavations at G-115. Similar objects are reported (Holm, pers. commun.) to have come from G-31 excavations, but none were included in the shell samples subjected to systematic analysis. Unclassified Worked Shell Plate 22 g-e Most of the unidentified objects of shell are pieces of the waterworn central spiral core of conchs (pi. 22 c-d) brought to the site but showing no evidence of working or use. An eccentrically-shaped piece of pearl oyster is part of a larger object of unknown shape, possibly a stylized figurine (pi. 22e). Most of the surface and edges show cutting and abrasion, although waterworn spots remain from the original shell surface. Length is 12.5 cm., width 5.5 cm., thickness 1.0-2.0 cm. The "leg" end is 6.8 cm. long, tapering from the "thigh" 2.3 cm. wide to the "foot", 1.5 cm. wide. It comes from G-31, Cut J, Section E, Level 2.70-3.00 meters. WHOLE VOLUME THE VALDIVIA PHASE 41 Chronological Distribution of Shell Artifact Types Description of shell artifacts from the Valdivia Phase must be considered preliminary because unfortunately little of the material from the large excavation at G-31, Cut J, dating from Periods A and B was pre- served for classification and analysis. Although refuse from earlier excavations at G-31, and from G-54, Cut 1 and G-88, Cut 2 was screened and searched for artifacts, the extreme rarity of worked shell ruled out the possibility of obtaining a good sample from these small cuts. When levels of the various seriated sequences pro- ducing shell artifacts are interdigitated according to the trends in pottery types (Appendix 1, table 2), and period subdivisions are drawn, it is possible to gain an idea of the temporal distribution of the different kinds of shell artifacts by connecting earliest and latest occurrences in a vertical bar (fig. 20). Although the result suggests somewhat more chronological differentiation than was evident in stone artifacts (fig. 18), this may be a reflection of the rarity of some of the types rather than restricted occurrence. As would be expected, fishhooks and blanks for fishhook manufacture are present from the beginning of Period A, a distribution paralleled by two kinds of ornaments: beads, and drilled and shaped pend- ants. Disks seem to begin only near the middle of Period A, but since they probably represent incom- pleted beads, absence earlier can be interpreted as accidental. Bowls, represented by a single example from the end of Period B, and drilled pecten pendants, represented by a single example from the middle of Period C, can be viewed as unique objects without cultural importance. The absence of abraders and polishers from the first half of Period A is of interest because of the similar temporal distribution shown by stone abraders (fig. 18). In both cases, this may be the result of unusually small samples from early Period A; on the other hand, the coincidence of distribution may re- flect introduction of a new item in the perishable material culture inventory, or a new technique of manufacturing. Restriction of clam shell pendants to Periods B and C may represent adoption at the end of Period A of this type of ornament, requiring less work for production than shaped pendants. Although distributional evidence is again incon- clusive, it is possible that shaped shell scoops are a refinement of the unworked waterworn shells of similar contour present in refuse of Period B. While no shaped scoops were included in our classified shell sample, they are reported to occur frequently at G-115 (Zevallos and Holm, 1960, p. 8, hun. 4, 7), which can be dated by pottery type and figurine style as principally belonging to Period C. Bone and Teeth Artifacts Bone fragments were rare in the refuse of all Val- divia Phase sites, and pieces showing working or use even rarer. A few of antler or fish were identified, suggesting that with better conditions of preservation a large number of implements of this relatively perish- able material might have been found. Deer Antler Awls Plate 25 a-d Eight worked antlers come from stratigraphic levels and miscellaneous tests at G-54. The short antler has been abraded to remove the rough natural surface and to form a flat, chisel-like blade. Degree of abra- sion and polish varies considerably on fragments scat- tered in the refuse; only those with clearcut evidence of shaping of the point were classified as awls. The base of the awl, corresponding to the root of the antler, was not reshaped; several specimens show hack marks near the base where the ander was cut out of the deer skull. Complete specimens measure 10.5 and 10.8 cm. long, and one fragment 11.2 cm. long would reconstruct to about 13.0 cm. long. The worked area of these three specimens is 8.0, 6.5 and 6.0 cm. long respectively. The chisel-like tips show a high polish, some of which must be the result of use. Antler Tip Projectile Point Plate 25 e One antler was cut 2.5 cm. from the tip and then drilled conically from the cut end to a depth of 1.2 cm., producing a hollowed cone. Diameter at the cut end is 1 cm. Shape and size suggest this may have been the tip of a projectile. It is from G-54, Northeast extension of Cut 1, Level 85-125 cm. Fish Bone Awls Five fish bones have polishing striations on the tips and sides suggesting use as awls (called punches or drills in original report; Evans, Meggers, Estrada, 1959, p. 17). They show no working to change the natural shape. Length is 2.5, 3.2, 3.8, 4.0, and 4.8 cm. Fish Vertebra A large fish vertebra has had all the spinous proc- esses cut off, the edge nicked in several places, and the center hole enlarged. The resulting disk is 3.0 cm. in diameter and 1.7 cm. thick, with a central perforation 6 mm. in diameter. The use of this worked bone is uncertain, but it might have been either a spindle whorl on an eag plug. Awls of Teeth from Sawfish Plate 25 f-j Five objects were identified as the teeth from the sawfish, Pristes. Although the points of two are 42 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 broken off, the others have been slightly abraded and shaped into awl points. One fragment clearly shows dressing down on the shaft area. The one complete specimen measures 8.2 cm. long, 1.2 cm. wide and 0.6 cm. thick. Chronological Position of Bone and Teeth Artifacts The scarcity of worked bone and teeth in seriated levels corresponding to Periods A-C, and the absence of any sample from Period D (Appendix 1, table 5) makes it difficult to recognize significant chronological differences. Awls are the principal artifact type. Fishbone awls are present in Period B, and probably missing from Period A only because this part of the sequence is inadequately represented in excavations where the nonceramic refuse was subjected to careful analysis. The addition of antler and sawfish tooth awls during Period C can be correlated with other evidence of greater emphasis on terrestrial mammal hunting during this period, and suggests increased use of skins. The only object identifiable as a possible projectile point is the hollowed-out antler tip from Period C at G-54. It is impossible to judge whether this re- flects the introduction of this artifact in Period C, or the accidental find of a type of object which, because of its relatively perishable nature, rarely survives in the refuse. The use of antler projectile tips would account for failure to identify stone points among the chips and flakes. Pottery Artifacts Pottery Type Descriptions The classification and description of Valdivia Phase pottery types is the result of intermittent analysis over a period of six years of hundreds of thousands of sherds. The basic typology was established from materials excavated at G-31 (Cuts A through I) during the first season, supplemented by a badly eroded sample from G-25 (Evans, Meggers and Estrada, 1959). More extensive excavations at G-31 (Cut J) brought to light several rare early decorated types not represented in the original small sample. Unfortunately, the unpolished plain types from Cut J were classified in Ecuador by an assistant, who did not follow the criteria previously established for separating Valdivia Plain and San Pablo Plain, and none of the rim forms were recorded and tabulated. The final season of fieldwork, including tests at G-88 and more extensive excavations at G-54, suggested the usefulness of making some new distinctions in decorative technique. Since only type samples of sherds from excavations classified in Ecuador were shipped to Washington, it was not possible to recheck the occurrence of these new types in certain cuts and levels. The situation is most unfortunate in regard to G-31, Cut J, which provides the most extensive sample for Periods A and B. Some of the problems resulting from this incomplete information will be discussed in the section dealing with interpretation of the seriated sequences and the reconstruction of Valdivia Phase ceramic evolution (pp. 87-95). All sherds from G-31, Cuts A, F, and H had been re- tained and this material was reclassified. Pottery type descriptions have been arranged in alphabetical order. Since many have appeared in publication in Spanish, the Spanish equivalent of the type name is provided to facilitate identification and future reference. The scale of hardness is Mohs', but the word is not repeated in each description. Colors of paste and surfaces are given in generalized terms because it can be shown that over a surface area 5 to 10 cm. square as many as 10-15 Munsell or a similar type of color designations can be identified, all the result of poorly controlled firing. Appendix 1, table 6 gives the number and pro- venience of sherds subjected to detailed classification, providing the basis for the seriated sequences; in some cases additional selected samples from test trenches or pits are also tabulated. The unselected sample used to establish the sequence of Periods A, B, C and D of the Valdivia Phase (see pp. 89-90 for breakdown by periods) totals 219,488 sherds, of which 40,829 or 18.5 percent represent decorated pottery types and 178,659 or 81.5 percent plain pottery types. The pottery type descriptions that follow are based on sherds recovered from all Valdivia Phase sites. Vessel shape data and decorative motifs are less com- plete, lacking the tabulations of sherds left in Ecuador from G-31, Cut J, G-88, Cut 1 and G-54, Cut 2. Although the ceramic sample for the Valdivia Phase runs to over a quarter of a million classified sherds, the number of complete vessels is less than two dozen. Vessel shapes have been reconstructed from rim profiles and body sherds of diagnostic curvature, and their general validity can be checked to a slight extent by reference to illustrations of complete vessels. In each reconstructed vessel shape, the drawing was made in the middle of the size range to offer a common unit for comparison. In spite of considerable variation through time, it is possible to describe certain general characteristics of Valdivia Phase pottery. Technique of manufacture is by coiling, with coil-junction fractures common in the early part of the Phase. The use of short segments rather than long coils seems indicated by a tendency toward a steplike break. The pottery is not fragile or friable. Surfaces are generally well finished, and WHOLE VOLUME THE VALDIVIA PHASE 43 ' ' ' ' ' ? ? 0 4 8 12 CM VESSEL SCALE 0 I 2 3 CM RIM SCALE Figure 21.-Rim profiles and reconstructed vessel shapes of Punta Arenas Incised. in the polished types equal in excellence to the best products of later periods. Regularity of vessel shape varies with the pottery type, lips of plain types being markedly unlevel and noncircular whereas those of polished or decorated types are often level and circu- lar. Wall thickness is not uniform. Decoration ijs by incision, excision, applique, finger grooving and punching, pebble polishing, punctation, and rarely, stamping. Painted decoration is totally absent, but red slipping on one or both surfaces is common. The proportion of decorated vessels is very high, with decoration typically being limited to the neck on jars and a band between rim and shoulder on bowls, but sometimes covering the bottom of bowls. The orna- mentation generally has an attractive appearance, although close inspection shows unequal spacing and other evidence of lack of care or control in execu- tion. In general, Valdivia pottery is competently made and pleasing in form and decoration-far removed from what might be expected of people just beginning to learn the pottery art. Punta Arenas Incised {Punta Arenas Inciso) Paste: Similar to Punta Arenas Plain; see that type description for details. Surface: Color: Medium to dark gray; rarely, orange or brown. Treatment: Polished; surface, on the whole even but some flaws present. Temper grains not visible. Hardness: 3.5-4. Form: Rim: Direct, cambered, slightly everted or expanded, with tapered, rounded or flattened lip. Body wall thickness: 3-6 mm. Base: Probably rounded or slightly flattened. Principal vessel shapes reconstructed from sherds: 1. Small bowl, rounded bottom, slightly incurving upper wall, expanding rim, rounded lip. Rim diame- ter 6-10 cm. (fig. 21-1). Decoration on exterior. !7 617-841-65 5 Carinated bowl with rounded bottom, low shoulder, slightly concave upper wall, everted rim, flattened or tapered lip. Rim diameter 10-12 cm. (fig. 21-2). Decoration on exterior above carination. Cambered jar with rounded body, constricted neck, angular cambered rim, rounded lip. Rim diameter 8-12 cm. (fig. 21-3). Decoration on rim exterior. Decoration (pi. 26): Technique: Shallow, rounded incisions, 1-2 mm. wide, on polished surface. Depth varies from faint to well de- fined, but not exceeding 0.5 mm., and the bed of incision is typically unpolished. Edges are not dragged or thrown up. Execution is moderately good, with lines generally parallel but not equally spaced, and with junctions sometimes overlapping or falling short. Termination of incision may taper to a point. Red color added to the incision after firing is retained on one example, and may once have been more common. Punctates of irregular form occasionally occur. Out- line is elongated, narrow to ovoid. Some were appar- ently made with the fingernail; others with a pointed object applied at a strong slant. In two examples, some of the punctates are V-shaped (with curved arms) as though the fingernail had been rocked once (pi. 26 v, x). Motif: Straight parallel lines, 2-6 mm. apart, running horizontally, vertically or diagonally, filling bands or zones of irregular form. Occasionally, gashlike puncta- tion is substituted in some of the bands or zones (pi. 26 v-y), or a small rectangle or stepped element is incorporated (pi. 26i). Temporal Differences Within the Type: None dis- cernible. Chronological Position of the Type: Restricted to Period D of the Valdivia Phase sequence (figs. 52, 53). Punta Arenas Plain {Punta Arenas Ordinario) Paste: Method of manufacture: Coiling. Temper: Medium waterworn sand, with scattered grains up to 3 mm. in diameter; majority around 1 mm. Abundant, producing a more sandy paste than charac- teristic of Valdivia Plain. 44 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Texture: Sandy, more granular than Valdivia Plain. Breaks are jagged but not friable. Color: Dark gray throughout the cross section, or fired orange at the surface to 1 mm. inward, leaving a gray core. Occasionally, orange throughout the cross sec- tion. Firing: Typically incompletely oxidized. Surface : Color: Typically, light orange or light tan; less commonly light gray; rarely, medium or dark gray. Both surfaces typically of similar shade; rarely, one may be orange and the other gray. Treatment: Poorly smoothed leaving unevenness, pits and other defects. Smoothing tracks are usually visible. Interior of bowls is smoother than the exterior. Hardness: 3.5-4. Form: Rim: Direct, expanded, exteriorly thickened or cam- bered, with rounded, tapered or flattened lip. Lips are very unlevel. Body wall thickness: 3-6 mm. Base: Probably rounded or slightly flattened. Common vessel shapes reconstructed from sherds: Jar with rounded body, slightly constricted concave neck, everted rim with rounded lip. Wall thickness may expand slightly on the everted portion of the rim. Rim diameter 12-16 cm. (fig. 22-1). Rounded jar with constricted mouth and everted expanded rim with tapered lip. Rim diameter 10-16 cm. (fig. 22-2). Rounded jar with constricted mouth and everted, (((( "/
  • ®t /#/// 1111111 O 4 8 12 CM VESSEL SCALE Figure 41.-Rim profiles and reconstructed vessel shapes of San Pablo Plain and Valdivia Plain. -767-841-65 7 76 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Valdivia Polished Red {Valdivia Rojo Pulido) Paste: Method of manufacture: Coiling. Temper: Wide variation from fine sand, possibly a natural inclusion of the clay, to coarse, whitish inclusions of feldspar and waterworn shell of the type used in Valdivia Brushed. Texture: Typically fine, compact, sandy but not friable. Color: Solid orange, solid gray, or orange along both surfaces leaving a thin gray core. Firing: Incompletely to completely oxidized. Surface: Color: Slipped surface: Exterior and rim interior of jars and both surfaces of open bowls are covered with a paper thin, rich, dark red slip. Variation in firing produces shades ranging from bright red to nearly black. Unslipped surface: Tan, light orange, brown, or gray. Treatment: Slipped surface: Typically, polished to produce a glossy, smooth, even finish (pi. 97 c, j, 1, n). In a minority, the slip is not evenly applied and forms streaks that allow the underlying orange surface to show between. Polishing is sometimes striated rather than complete, leaving the surface slightly uneven (pi. 97 b, d). Slip typically begins just below the lip on the interior and covers the exterior. Open bowls are also slipped on the interior. Unslipped surface: Less well smoothed, sometimes leaving imperfections unobliterated, but typically polished or striated polished (pi. 97 k, m, o). Hardness: 3.5-4. Form: Rim: Direct, expanded, exteriorly thickened, interiorly thickened, cambered or carinated, with rounded, flattened or tapered lip. Rim may be level, undulating or lobed (pi. 181 1-m). Body wall thickness: 5-11 mm. Base: Bases may be rounded, flattened, thickened, tetra- pod or concave. Details are as follows: 1. Rounded, continuing the curvature and thickness of the body wall, so that the junction between wall and base is undefinable. Flattened, so that the vessel will rest without tipping, but with the curvature blending into that of the body wall so that the diameter cannot be measured with accuracy. Flattening typically produces an unlevel surface and, therefore, often the rim is not horizontal. Thickened to between one quarter and one third greater than the thickness of the body wall, and typi- cally flattened slightly on the exterior. Tetrapod, the four small, stubby, truncated conical feet flattened at the end and frequently showing ex- treme wear. Height of unworn specimens varies from 0.4-2.5 cm., with the majority 0.4-1.2 cm. Diameter increases with increased height, and at the point of junction with the body wall ranges from 1.0-3.5 cm. On three sherds with more than one foot present, distance apart measured from center to center is 3-5 cm., with arrangement approximately symmetrical (fig. 43a; pi. 97 e-i, 98 a-b). 5. Concave, with a well defined, circular depression on the exterior and a corresponding convexity on the interior. The wall is rarely thickened either at the center or edges of the depression. Diameter ranges from 4-10 cm., with the majority 6-7 cm.; depth is 2-10 mm., with increased depth not consistently as- sociated with increased diameter (fig. 43b; pis. 98c, H6q). Common vessel shapes reconstructed from sherds: Bowl with flattened bottom, walls upcurving to nearly vertical rim with rounded or flattened lip. Between 1.3 and 3.0 cm. below the lip, a jog is pro- duced by cutting the surface back 1-2 mm., decreasing the body wall thickness by that amount between this point and the lip. Occasionally the inset zone may lack the red slip. Lip is rarely undulating. Rim diameter 12-32 cm., majority 18-22 cm. (fig. 44-1). Deep bowl with flattened bottom, rounded walls, slightly outcurving to slightly incurving rim with rounded or tapered lip. The wall is exteriorly thick- ened from 2-3 mm. for a distance of 2.5-4.5 cm. below the lip producing a broad-collar-like rim. The lip is occasionally lobed. Rim diameter 14-20 cm. (fig. 44-2). Rounded bowl with slightly flattened bottom, walls upcurving to vertical or slightly constricted rim with rounded lip. Beginning about 1.5 cm. below the lip, body wall thickness expands and may attain at the lip a thickness more than double that of the lower body wall. Rim diameter 14-28 cm. (fig. 44-3). Shallow bowl with flattened bottom, walls outslop- ing to interiorly thickened rim with tapered or rounded lip. Rim diameter 14-24 cm. (fig. 44-4). Rounded bowl with flattened or tetrapod base, con- stricted direct (rarely, expanded) rim with rounded lip. Rim diameter 12-28 cm. (fig. 44-5). Shallow bowl with flattened bottom, walls outcurving to direct rim with rounded or flattened, occasionally undulating or lobed lip. Rim diameter 12-28 cm. (fig. 44-6; pi. 99b). Deep bowl with rounded or slightly flattened bottom, walls curving upward to nearly vertical rim with rounded or tapered lip. On the exterior 1.5-3.0 cm. below the lip, the wall is thickened to produce an angular contour. Occasionally, the thickening is accompanied by a slightly inward turn of the wall. Rim diameter 16-22 cm. (fig. 44-7). Carinated bowl with flattened bottom, wall curving outward to join the upper wall at a rounded angle 2.5-4.5 cm. below the rounded or flattened lip. Wall thickness may increase slightly at the carination. Rim diameter 16-32 cm., majority 18-24 cm. (fig. 44-8; pi. 99a). Carinated bowl with flattened bottom, wall curving upward to form an angular junction with the upper wall, accompanied by thickening to produce a smooth curve on the interior. Flattened or rounded lip. Rim diameter 14-20 cm. (fig. 44-9). WHOLE VOLUME THE VALDIVIA PHASE 77 '/Wilt >)1WV> 0 1 2 3 CM RIM SCALE 1 1 1 I i i 1 0 4 8 12 CM VESSEL SCALE Figure 42.-Rim profiles and reconstructed vessel shapes of Valdivia Polished Plain. 78 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Bowl with rounded body separated from the rim by a constriction sometimes accompanied by a change in body wall curvature, creating a broad channel on the exterior. Lip is tapered or rounded and occasion- ally lobed or undulating. Rim diameter 14-26 cm. (fig. 44-10). Jar with concave or flattened base, rounded body, insloping to short nearly vertical neck, direct rim, flattened or rounded lip. Rim diameter 8-22 cm. (fig. 44-11; pi. 99c). Jar with rounded body, slightly constricted neck and cambered rim. Rim diameter 12-16 cm. (fig. 44-12). Occasional Decoration : Technique: In rare examples, the red slip is applied in- completely, producing an ornamental effect. Edges of zones are straight and sharply defined. Motif: Unslipped areas are most common on Form 1, where the inset zone adjacent to the rim exterior is frequently left plain. Only four examples were found in which the red slip was employed to produce a more complicated pattern, three bowl interiors of Form 3 and one exterior of Form 7 (Evans, Meggers and Estrada, 1959, fig. 32b). Associated techniques: Except for occasional lobed or un- dulating rim treatment, none of the sherds classified as Valdivia Polished Red has any kind of decoration. Lobes are rectanguloid or curved and may be plain or decorated with nicks spaced 3-10 mm. apart. Temporal Differences Within the Type: The majority of the crudely-finished, poorly-smoothed or incompletely- polished sherds come from the early part of Period A. By the latter part of Period A, workmanship is typically of good quality, although variation in degree of evenness ? i ? i l 2 3 CM Figure 43.-Profiles of Valdivia Polished Red base forms, a, Tetrapod. b, Concave. WHOLE VOLUME THE VALDIVIA PHASE 79 U))\\\)\ ^lt)\ 'j/n ?M))\ 0 I 2 3 CM RIM SCALE Figure 44.-Rim profiles and reconstructed vessel shapes of Valdivia Polished Red. 80 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME I and polish occurs throughout the sequence. A general deterioration in quality is notable in Period D. Zoned red for occasional decoration is restricted to Period A. The only vessel shapes represented in Period D are Forms 6 and 11. Forms 8 and 12 are restricted to Period C. Forms 9 and 10 appear to be restricted to Periods B and C, but are relatively rare and may not be represented in the small samples analyzed from Period A. Lobed or undulating rims are restricted to Periods A and B. Tetrapod bases are absent after the first part of Period B; concave bases continue into Period C. Chronological Position of the Type : Valdivia Polished Red is common in Periods A and B, attaining a frequency of 30-40 percent of the total sherds in some levels of the seriated sequence in Period A. It declines markedly in popularity in Period C, and in Period D is reduced to a very minor proportion of all the pottery types (figs. 49- 50, 52-53). Valdivia Pseudo-Corrugated {Valdivia Corrugado Falso) Paste and Surface : On Valdivia Plain (p. 72); see that type description for details. Form: Rim: The single rim sherd is slightly thickened on the exterior, tapering to a rounded lip. Body wall thickness: A-l mm. Base: Probably slightly flattened or rounded. Vessel shape reconstructed from sherds: 1. Jar with rounded body, insloping neck, slightly ex- teriorly thickened rim and rounded lip. Decoration begins at the lower edge of the rim thickening. Several body sherds show decoration extending from the base of the neck to a little below the maximum body diameter. Small size of sherds makes more complete description impossible. Decoration (pis. 54 b-i, 116 o, 169 g-i): Technique: Pseudo-corrugation produced by pressing downward with tip of finger on the wet, soft and very plastic surface in vertical rows. On some sherds these rows seem to follow coil lines but on others they do not. "Corrugations" are typically very pronounced, strongly overlapping, with deep depressions and prominent curved ridges. Motif: Applied to exterior of rim and neck or body below neck junction. Associated techniques: Additional ornament of the type classified as Valdivia Modeled is sometimes associated. On several body sherds, it is above the band of pseudo- corrugation. Temporal Differences Within the Type : None. Chronological Position of the Type : Most characteris- tic of Period A, continuing sporadically during Period B (figs. 50, 51). Valdivia Punctate {Valdivia Punteado) Paste : Method of manufacture: Coiling. Temper: Fine to medium sand, with grains typically under 1 mm.; occasional sherds have coarse temper, but this is unusual. Texture: Fine grained, compact, sandy but not friable; occasional long thin, lenticular air pockets. Color: About 60 percent are bright orange through the cross section; the remainder have a medium gray or gray-brown cross section. Surface: Color: Orange, light tan, brown, gray-brown or black, the latter the result of fire clouding. Exterior and interior generally have similar hues except where one has a fire cloud. A red slip is occasionally applied to die rim and neck interior of jars, or rarely to the exterior. Treatment: Great variation within the type from polished to smooth to gritty and uneven, but little variation on an individual sherd. Gashed punctates are associated with the poorest surface finish. Red slipped areas are well smoothed and polished. Hardness: 3.5-4. Form: Rim: Direct or exteriorly thickened, with rounded lip. Body wall thickness: 6-11 mm. Base: Probably slightly flattened or rounded. Vessel shapes reconstructed from sherds: Jar with rounded body, constricted concave-walled neck, slightly everted exteriorly thickened rim with rounded or tapered lip. Rim diameter 16-32 cm. (fig. 45-1). Jar with rounded body, slightly concave-walled neck, direct rim with flattened or rounded lip. Rim diameter 10-22 cm. (fig. 45-2). Rounded bowl with constricted mouth, direct rim and rounded lip. Mouth diameter 6-24 cm. (fig. 45-3). Decoration (pis. 100, 101, 116p): Technique: Punctations vary greatly in form and depth, reflecting a tendency to employ a wide variety of tools. Although shape and depth are similar on a single sherd, differences in the pressure and angle of the tool may produce slight to marked variation. Diagonal pressure may push up the clay to form a ridge at one edge. The majority of punctates can be classified into one of the following categories which tend to grade into one another: Circular depressions made with flat ended or pointed tool 2-5 mm. in diameter, 1-2 mm. deep. In thin walled examples, pressure may create a corresponding boss on the interior. Spacing is from adjacent to 8 mm. apart, and adjacent holes may be deformed by executing of succeeding ones (pi. 100 a-n). Rarely, a hollow tool, like a cane, produces a ring rather than a hole. Trianguloid depressions made with a blunt or pointed tool (pi. 101 a, c-f). Rectanguloid marks (pi. 101 b, 1). Ovoid depressions (pi. 101 g-k, m-n). Long thin cuts (pi. 101 o-q). Tiny to pinpoint size punctates of irregular form but typically under 1 mm. in diameter (pi. 100 o-r). WHOLE VOLUME THE VALDIVIA PHASE 81 Motif: All-over application on the jar neck is most typical, the punctates arranged in horizontal rows or hap- hazardly filling the space. Occasionally, arrangement is in independent rows (pi. 100 m-n) or zones (pis. 100 b, d, r, 101 g, j, k). Rarely, punctate areas are subdivided by incised lines (pi. 101 a, b). Red slipped exterior surfaces are associated with techniques 1 and 6. Temporal Differences Within the Type: None; all techniques occur throughout the duration of the type. Chronological Position of the Type : Present throughout the sequence in minor frequency (figs. 49-53). Valdivia Red Incised {Valdivia Rojo Inciso) Paste and Surface: Similar to Valdivia Fine-Line In- cised (p. 60); see that type description for details. Form: Rim: Direct with flat (occasionally rounded) lip; rarely shouldered. Body wall thickness: 0.6-1.5 cm. Base: Slightly flattened or tetrapod. Vessel shapes reconstructed from sherds: Bowl with wall curving upward or slightly inward. Between 1.0 and 1.5 cm. below the lip, body wall thickness is reduced 1-2 mm. producing a well marked shoulder on the exterior. Rim diameter about 28 cm. (fig. 29-1 top). Bowl with wall curving upward and slightly inward to direct rim, flat (rarely rounded) lip. Rim diameter 24-34 cm. (fig. 29-2 top). Jar with rounded body, wall sloping inward and upward to direct rim, flattened lip. Rim diameter 28-34 cm. (fig. 29-3 top). One example has a castellated rim (pi. 103 a, b). Decoration (pis. 102-104, 177i, 178b): Technique: Incised lines 2-3 mm. deep and 2-5 mm. wide made when clay was fairly dry, giving the effect of careful workmanship although lines are not always evenly spaced or exactly parallel. Intersecting lines may have rounded corners. In about one-third of the examples, there is some excision at corners or in small areas of the design. Some examples include circular punctates. Motif: Geometric patterns, predominantly rectilinear, in a small number of variants. Most common are straight vertical and horizontal lines producing a hachure (pi. 102f) sometimes with punctates at the center of the squares (pi. 104 a, f); interlocking frets (pis. 102 a-b, g, 103c), sometimes with one element excised (pi. 102 c-d); and a band of circles with a dot in the center (pi. 104 b-e, g-h). Rarely, horizontal rows of slanting parallel lines occur (pi. 103 a-b). Decoration is ap- plied to the exterior from rim to below the shoulder. Temporal Differences Within the Type: None; ex- amples with excision occur in minor frequency through- out the duration of the type. Chronological Position of the Type: Most frequent during Period A, declining during Period B, and absent thereafter (figs. 49-53). Valdivia Red Zoned Punctate {Valdivia Rojo Punteado en Jonas') Paste: Method of manufacture: Coiling, with tendency to break along coil junctions, particularly on the neck. L_L ? ' ' ' 0 1 2 3 CM RIM SCALE i i i i i 0 4 8 12 CM VESSEL SCALE Figure 45.-Rim profiles and reconstructed vessel shapes of Valdivia Punctate. 82 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Temper: Abundant fine to medium sand, with occasional grains up to 4 mm. White particles abundant in some vessels. Texture: Compact, fine grained; temper sometimes ap- pears unevenly distributed because of differential size of particles. Color: Variation from dark gray to orange through the cross section, with some examples fired orange 1-2 mm. from each surface leaving the remainder of the core gray. Firing: Completely to incompletely oxidized. Occasional small fire clouds. Surface : Color: Exterior: Red slipped, varying from rich red to red- orange. Rare examples lack a red slip, and are light grayish tan to light brown. Interior: Light tan to light gray-brown most typical; rarely medium gray. Treatment: Exterior: Red slip usually applied rather evenly in a thin coat covering the entire surface; occasionally applied in streaks leaving the orange-tan surface visible between. Decorated area and neck are gen- erally even, smooth and polished to a low luster. Polishing streaks may run vertically on the neck producing burnished lines of darker hue. Red slip was typically applied subsequent to decoration, color- ing the incisions and punctations as well as the sur- rounding surface, and occasionally omitted from zones between the incisions. Interior: Smoothed to a degree ranging from superficial, leaving ridges and unevenness, to good, producing an even, smooth, striated polished to polished finish. Hardness: 3.5-4. Form: Rim: Typically cambered; rarely, direct or everted; rounded, flattened or tapered lip. Body wall thickness: Range 4-9 mm., usually 5-8 mm. Base: Probably rounded or flattened; no complete examples. Vessel shapes reconstructed from sherds: Jar with rounded body, constricted neck and cam- bered rim forming an insloping collar 2.0-3.5 cm. wide. Rim diameter 10-18 cm. (fig. 39-1 top). Rounded jar with constricted mouth, direct or slightiy expanded rim, rounded or tapered lip. Mouth diameter 10-18 cm. (fig. 39-2 top). Rounded jar with constricted neck and short everted rim with tapered lip. Rim diameter 14-20 cm. (fig. 39-3 top). Decoration (pis. 105, 106): Technique: Broad incisions defining zones that are filled with punctation. Incisions are 2-5 mm. wide, the width of a single line varying slightly due to varying pressure of the tool. Depth is 0.5 mm. or less, with a shallow U-shaped cross section. Incisions may be straight and evenly parallel or waving, but typically are not precisely drawn. Punctates are circular (diam- eter 2-4 mm.) or elongated (2 by 4, 3 by 5 mm.) and ovoid or rectanguloid in outline. Circular punctates are made either with a pointed or flat ended tool pro- ducing a conical or cylindrical imprint. Punctates vary from adjacent to 5 mm. apart, and spacing tends to be similar on a single vessel. Red slip was applied after decoration, filling incisions and shallow punctates, but leaving deeper ones uncolored. Polishing subsequent to slipping may cause clay to overlap the mouth of the punctates. Motif: A band of decoration occupies the area above the camber on jars of Form 1. Most commonly, a lobed, undulating or zigzag incision divides the area hori- zontally into two zones, the lower one filled with punctation (pi. 105 a, c-g, j-k). Occasionally, a second incision runs just below the rim, and the area between the two incisions may be left unslipped (pi. 105b). Less common motifs include independent square or rectanguloid areas defined by incision and alternately filled with punctation and left plain (pis. 105i, 106a), and a band of punctation on the upper portion of the rim rather than the lower edge (pi. 105h). On jars of Form 2, a double row of circular punctates may be applied on the exterior just below the lip (pi. 106 d-e), or punctates may occupy a broad band bounded at the rim by a broad incision paralleling the lip. Jars of Form 3 may have punctation on the ex- terior of the constricted neck, or in zones bordered by broad-line incision. Associated techniques: One jar of Form 3 has Valdivia Incised decoration on the neck; one jar of Form 1 has a band 8-9 mm. wide on the center of the punctated zone executed in the multiple drag-and-jab technique. Temporal Differences Within the Type : None. Chronological Position of the Type: Occurring rarely and sporadically in Periods A and B, becoming character- istic in Period C (figs. 49, 52-53). Valdivia Rocker Stamped {Valdivia Estampado en ^ig-zag) Paste and Surface : On bowls of Valdivia Polished Plain (p. 74) or rarely Valdivia Polished Red (p. 76), the interior surface being less well finished than the exterior and typically striated polished, or on jars of San Pablo Plain (p. 45) or less frequently Valdivia Plain (p. 72), the rim interior generally polished plain or polished red. See descriptions of the above types for details. Form: Rim: Direct, expanded or carinated with rounded or flattened lip on bowls; everted with expanded or direct lip or cambered on jars. Body wall thickness: Range 4-10 mm.; Majority 5-8 mm. Base: Probably rounded or slightly flattened. Vessel shapes reconstructed from sherds: Shallow bowl with flattened bottom, walls curving upward to expanded rim with a broad outsloping top. Thickness at the lip is 2 to 4 times that of the body wall. Rim diameter 16-32 cm. (fig. 46-1). Deep rounded bowl with slightly constricted mouth, WHOLE VOLUME THE VALDTVIA PHASE 83 direct rim and rounded, tapered or flattened lip. Mouth diameter 14-26 cm. (fig. 46-2). 3. Carinated bowl with flattened bottom, wall sloping upward to angular shoulder 1-3 cm. below the direct rim, rounded or flattened lip. Carination is thick- ened markedly to produce a smooth curve on the interior. Rim diameter 24-34 cm. (fig. 46-3). Rounded jar with slightly constricted mouth, slightly everted rim sometimes expanded in thickness just below the rounded or tapered lip. Rim diameter 14-18 cm. (fig. 46-4). Jar with rounded body, constricted neck and V"~ I I I 0 I 2 3 CM RIM SCALE I I I I I II 0 4 8 12 CM VESSEL SCALE re 46,-Rim profiles and reconstructed vessel shapes of Valdivia Rocker Stamped. (Depressions on exteriors result from decoration.) 767-841-65 8 84 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 cambered rim. Camber is angular and thickened on the interior. Lip is rounded or flattened. Rim di- ameter 12-20 cm. (fig. 46-5). Decoration (pis. 107-112, 179 i-n): Technique: Rocker stamping executed by a combination of rocking and dragging of a broad ended tool, pro- ducing a series of interlocking trianguloid depressions separated by a narrow ridge. The effect is a negative version of the usual type of rocker stamping. The technique is applied so as to create a band 0.5-1.3 cm. wide and 1-2 mm. deep. In some examples the margin is straight suggesting a broad incision may have been drawn first; in others the margin is jagged and appears to be formed by the rocker stamping itself. Typically, the rocking is evenly executed, with strokes 4-5 mm. apart, but instances occur of "crowding" particularly toward the end of a line where the strokes are so closely spaced that they obscure each other. The trianguloid facets often show parallel horizontal striations. The left edge is typically deepest, with the surface sloping upward to the right, suggesting execution from left to right. Motif: On bowls, rocker stamping is applied in one or more independent rows, either continuously around the circumference or interrupted by blank areas 0.5-2.0 cm. in width (pis. 107-109). On Form 1, the row is just below the angular outer edge of the rim. On Form 2, it is 0.5-1.0 cm. below the lip. On Form 3, a row of rocker stamping typically occupies the center of the upper wall, between two broad-line incisions, one just below the lip and the other just above the carination (pi. 110 a-b); rarely, two rocker stamped rows run parallel (pi. lllg). Occasionally a second row of rocker stamping occurs just below the carination (pi. 110 d, g). One example of Form 2 has additional vertical rows (pi. 179k). On jars of Form 4, the exterior of the neck is typically covered with adjacent horizontal rows of rocker stamping producing an overall textural effect (pi. 112 g-q, s, u). Jars of Form 5 are occasionally similarly decorated, but more commonly rocker stamp- ing is limited to a single row adjacent to the rim or just above the angle of the camber (pi. 112 b-c). Associated techniques: Pebble polishing (pi. 109k), incision or punctation are rarely applied to bowls of Form 2. Bowls of Forms 1 and 3 may have broad-line incision on the body (pis. 110, 111). Jars of Form 5 and rarely of Form 4 may have Valdivia Incised decoration on the neck or body. Nicked ribs sometimes occur on rims of Form 5. Temporal Differences Within the Type : None evident. Chronological Position of the Type: Limited to Period C, particularly the latter half of the Period as exemplified at G-54 and G-115 (figs. 52-53). Valdivia Shell Stamped {Valdivia Estampado con Concha) Paste and Surface: Similar to San Pablo Plain (p. 45) and Valdivia Plain (p. 72); see those type descriptions for details. Form: Rim: Folded-over, exteriorly thickened or everted with flattened lip. Body wall thickness: 5-9 mm. Base: Probably rounded or slightly flattened. Vessel shapes reconstructed from sherds: Jar with rounded body, insloping neck, slightly everted exteriorly thickened rim with flattened lip. Exterior thickening not smoothed onto underlying surface, giving "folded-over" appearance. Rim di- ameter about 14 cm. Lip nicked or finger pressed. Associated with decoration of Motif 1 (fig. 37-1 bottom). Jar with rounded body, insloping neck, everted exteriorly thickened rim, flattened lip sometimes decorated with notches. Associated with decoration of Motif 1 and 2 (fig. 37-2 bottom). Jar with rounded body, constricted neck, strongly everted rim and flattened lip. Exterior rim diameter 14 cm. Associated with decoration of Motif 3 (fig. 37-3 bottom). Jar with everted rim turned inward at almost a 90ø angle 1.5 cm. below the tapered lip. Rim diameter 18 cm. Associated with decoration of Motif 3 (fig. 37-4 bottom). Decoration (pi. 113 a-k): Technique: Series of marks made with a piece of fluted shell giving a slightly curved dentate impression 1.0- 1.5 cm. long, with 3-4 "teeth." Marks are deep, clear and well defined, produced by pressing the shell straight into the clay. Spacing is 1-5 mm. apart. Motif: Three distinct decorative motifs occur: Rows of horizontal marks running vertically down neck to shoulder; rows are typically continuous, 0.3- 1.5 cm. apart, but may be in zones separated by undecorated areas (pi. 113 a-e). One or more rows of vertical marks running hori- zontally along the rim or neck exterior (pi. 113 f-i). A continuous line formed by joining marks end to end (pi. 113 j-k). Temporal Differences Within the Type: Motif 1 is confined to Period A; the other motifs occur in Period B, possibly continuing into Period C. Chronological Postion of the Type: Characteristic of Period A (figs. 50-51). Valdivia Striated Polished Plain {Valdivia Pulido en Lineas) Paste: Similar to Valdivia Plain (p. 72); see that type description for details. Surface : Color: Orange-tan, red-orange, brown-orange, brown or gray-brown, with some range of variation on each sherd. Few fire clouds. Jar exteriors are sometimes blackened, possibly from use in cooking. Treatment: Exterior: After preliminary smoothing, the surface was allowed to dry slightly before rubbing with a smooth, WHOLE VOLUME THE VALDIVIA PHASE 85 uvvs, I I I ' ' ' I 0 4 8 12 CM l i i ? i . I 0 I 2 3 CM VESSEL SCALE RIM SCALE ^m win Figure 47.-Rim profiles and reconstructed vessel shapes of Valdivia Striated Polished Plain. 86 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 waterworn pebble to produce polishing striations 0.5- 1.5 mm. wide (pi. 115). Striations usually vary from overlapping to 5-15 mm. apart, although some sur- faces show only a few widely spaced lines. Direction is generally horizontal. Finish ranges from distinct striations to incomplete polish leaving small Iusterless areas. Interior: On bowls, treatment is similar to that on the exterior. On jars, striated polish is usually limited to the interior of the rim and neck, the remainder of the interior being superficially smoothed. Hardness: 4-4.5. Form: Rim: Folded-over, exteriorly thickened, expanded, in- teriorly thickened, direct or carinated with rounded, flattened or tapered lip. Lip is occasionally finger pressed or lobed (undulating). Body wall thickness: 4-9 mm. Base: Rounded or slightly flattened; sometimes thickened to 10-13 cm., rarely, concave. Common vessel shapes reconstructed from sherds: Jar with flattened bottom, slightly constricted neck, exteriorly thickened folded-over rim with tapered or flattened lip, sometimes finger pressed. Rim diameter 18-22 cm. (fig. 47-1). Jar with flattened or thickened bottom, rounded body, slightly constricted neck, everted folded-over rim with tapered or flattened lip. The lower edge of the exterior thickening is ornamented by a contin- uous row of fingertip impressions; occasionally the lip is similarly treated. Rim diameter 18-36 cm. (fig. 47-2). Jar with flattened bottom, rounded body, slightly constricted neck, insloping or slightly everted, ex- teriorly thickened rim with rounded, flattened or tapered lip. Rim diameter 14-22 cm. (fig. 47-3). Jar with rounded body, short insloping neck, direct rim, flattened or rounded lip. Rim diameter 10-20 cm. (fig. 47-4; pi. 99d). Rounded bowl with slightly constricted mouth, direct rim, rounded lip. Rim diameter 12-26 cm. (fig. 47-5). Bowl with flattened bottom, rounded shoulder and slightly insloping rim with rounded lip. Between 1.5 and 2.0 cm. below the lip, the exterior wall is cut back sharply producing a 1-2 mm. decrease in thickness. Rim diameter 20-26 cm. (fig. 47-6). Open bowl with flattened bottom, upcurving or slightly incurving wall, expanded rim, flattened, rounded or tapered lip. Rim diameter 20-34 cm. (fig. 47-7). Open bowl with flattened bottom, walls outsloping to interiorly thickened rim with rounded or tapered, occasionally lobed lip. Rim diameter 16-32 cm. (fig. 47-8). Open bowl with flattened bottom, walls curving out- ward or upward to direct rim with rounded or ta- pered lip. Rim diameter 10-28 cm. (fig. 47-9). Carinated bowl with flattened bottom, lower wall joining insloping upper wall at angular to rounded junction, typically unthickened on the interior. Direct rim, rounded or tapered lip. Rim diameter 18-24 cm. (fig. 47-10). Carinated bowl with flattened bottom, lower walls curving upward to join insloping upper wall at a typically pronounced angle created by exterior thick- ening 1.0-2.5 cm. below the lip, which is rounded. Interior wall forms a smooth curve. Rim diameter 14-28 cm. (fig. 47-11). Temporal Differences Within the Type: Forms 1 and 6 are absent after Period B; Forms 2 and 5 are absent in Period D. Chronological Position of the Type: Although present throughout the seriated sequence, Valdivia Striated Polished Plain is most popular during Periods B and C (figs. 49-50, 52-53). Valdivia Zoned Incised {Valdivia Inciso en œonas) Paste and Surface : On Valdivia Polished Plain (p. 74); see that type description for details. Form: Rim: Direct or cambered, with rounded lip. Body wall thickness: 3-7 mm. Base: Probably flattened. Vessel shapes reconstructed from sherds: Bowl with rounded body and slightly constricted direct rim with rounded lip. Rim diameter 8-12 cm. (fig. 48-1). Carinated bowl with vertical or slightly insloping upper wall, direct rim and rounded lip. Rim diam- eter 12-20 cm. (fig. 48-2). Jar with rounded body, constricted neck and angular cambered rim with rounded lip. Rim diameter about 12 cm. (fig. 48-3). Decoration (pis. 113 1-s, 114): Technique: Designs are typically composed of broad, rounded incisions or grooves, 2-8 mm. wide and about 0.5 mm. deep delimiting areas filled with ornamenta- tion of 3 principal and 1 rare technique: A single row of short diagonal lines of the same width and depth as those forming the rest of the design (pi. 113 l-o). Two or more rows of short gashes of rectanguloid or ovoid outline (pi. 113 p-r). Diagonal or cross hatch drawn with shallow fine to scratch-like incisions, typically combined with very broad grooves as boundary lines (pi. 114 a-p). A rare variant is composed of incisions 1-2 mm. wide, which are used for both outline and zoned cross hatch (pi. 114 q-u). The broad-line incisions are generally straight and rather evenly spaced but intersecting ends may not meet. Fine-line incisions are unevenly spaced and may stop short of or extend over the edge of the bordering broad line. WHOLE VOLUME THE VALDP7IA PHASE 87 Motif: Bands or areas of hachure bounded by incised lines. In Techniques 1, 2, and 4, the area between hatched zones is typically filled with parallel incisions; in Technique 3, it is typically left plain. Temporal Differences Within the Type: Technique 4 is the only one found in Period C of the seriated sequence. Chronological Position of the Type: Beginning in the latter part of Period C, continuing through Period D in slightly increasing frequency (figs. 52-53). Unclassified Decorated The majority of the decorated sherds not classified into pottery types bear some kind of punctation, but technique is too variable and arrangement to incon- sistent for grouping into a type. Valdivia pottery is remarkably standardized, and unclassifiable sherds are rare considering the quantity of the total sample. Shell stamped.-Imprinting or stamping with the end of a narrow piece of fluted shell to produce a mark with three to four "teeth" is a common method of ornamenting the exterior of cambered rims on Valdivia Brushed and Valdivia Applique Fillet. When associated with these techniques, rims have been classified in the respective types. Three sherds from Period D have a similar horizontal row on the exterior of the thickened rim, the remaining surface apparently left plain. Punctate ridge.-Four sherds from Period B have diagonal ridges 1.3-1.7 cm. apart, 1 mm. in elevation, smoothed onto the adjacent surface with a row of punctates running end to end, 2-4 mm. apart, along the summit. Decoration is applied to the exterior body of small jars. Trade Pottery of Machalilla Phase Types Sherds of Machalilla Phase types, both plain and decorated, occur with considerable frequency in Valdivia Phase sites of Period C, particularly G-54. While plain sherds are not always readily distinguish- able because of similarity in temper and surface treatment to types of the Valdivia Phase, there are several stirrup spouts and rim sherds of diagnostic Machalilla Phase vessel shapes. Identified types in- clude the following (see Machalilla Phase type de- scriptions for details): Ayangue Incised. Machalilla Double-line Incised (pi. 138). Machalilla Embellished Shoulder. Machalilla Incised and Red Zoned (pi. 145j). Machalilla Punctate and Red Zoned (pi. 148 t-u). Machalilla Red Banded (both varieties) (pi. 153). Machalilla Striated Polished Plain. The Seriated Ceramic Sequence and Its Implications The seriated sequence of the Valdivia Phase is based on stratigraphic excavations at five habitation sites on the central Ecuadorian coast. Thickness of the de- posit ranges from 30-45 cm. at G-25 to 3.8 meters at G-31. Conditions for preservation of pottery surfaces are excellent at G-31, G-54 and G-88, moderately good at G-84, and exceedingly unfavorable at G-25, where most of the surfaces are badly eroded. To these variations must be added disturbances created by cultural factors, such as long continued occupation of a site and reoccupation by later groups, and analytic inconsistencies resulting from classification by differ- ent individuals over a period of years, and the lack of opportunity to recheck some of these classifications after the seriated sequences had been completed. In order to recognize and minimize the effect of these variables on interpretation of the ceramic sequence, three separate seriation charts have been prepared (figs. 49-50, 52). A fourth chart (fig. 53) combines portions of the other three into what appears to i i i i i i i 0 4 8 12 CM VESSEL SCALE i i i i 0 I 2 3 CM RIM SCALE Figure 48.-Rim profiles and reconstructed vessel shapes of Valdivia Zoned Incised. 88 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 represent the best reconstruction of trends in pottery type frequency during the Valdivia Phase. The first seriation includes results of three small stratigraphic pits excavated in different parts of G-31 during the first season of work in 1956-7 (fig. 49). The remaining stratigraphic cuts have not been in- cluded because either they duplicate the trends or the preliminary classification was not followed by analysis of rims and decorative motifs. The result is a revision of the chart originally published (Evans, Meggers and Estrada, 1959, fig. 78), incorporating several minor decorated types formerly left unclassified, and a re- analysis of the unpolished plain types in which San Pablo Plain was confined to coarse temper of the kind characteristic of Period C, eliminating many of the sherds originally placed in that type. Clearcut distinctions in popularity are evident in a number of the types, either as a marked alteration in frequency or as presence or absence. Of the plain types, Valdivia Plain is dominant in the lower por- tion of the sequence, where it comprises more than 60 percent of the total sherds. At the top of the sequence, it has declined to less than 20 percent. While Valdivia Plain loses popularity, Valdivia Polished Plain and Valdivia Polished Red gain, the latter reaching its maximum earlier than the former and showing a more pronounced decline. Punta Arenas Plain appears as a trace in the upper half of the sequence, and its seemingly erratic appearance is partly, if not entirely, the result of failure to recog- nize it at the time the principal classification was made. Of the decorated types, Valdivia Fine-line Incised, Valdivia Red Incised, Valdivia Combed and Valdivia Fingernail Decorated are confined to the lower levels. Valdivia Modeled appears in its maxi- mum frequency at this time. Valdivia Finger Grooved is confined to the lower half of the sequence. During the period the above types are popular. Valdivia Multiple Drag-and-Jab Punctate, Valdivia Red Zoned Punctate, Valdivia Nicked Broad-line Incised, Valdivia Excised, Valdivia Pebble Polished, Valdivia Brushed, Valdivia Nicked Rib or Nubbin, and Valdivia Applique Fillet are typically absent, while Valdivia Incised and Valdivia Broad-line Incised are present in minor frequency compared to later por- tions of the sequence. Among these later decorated types, clearcut differences in popularity cannot be identified with certainty, except that Valdivia Nicked Broad-line Incised and Valdivia Carved begin con- siderably later than the others. Valdivia Broad-line Incised shows a slight but consistent tendency to in- crease. Valdivia Incised reaches a climax of about 23 percent about one-third of the distance from the top of the sequence, followed by a climax in Valdivia Brushed. A few trade sherds of Machalilla Phase types were identified from levels near the top. This picture presents an interesting contrast to that resulting from interdigitation of sections D and E of the large Cut J at G-31, excavated in 1961 (fig. 50). Although similar tendencies can be recognized in a few of the types, the majority occur throughout the seriated sequence. There are several possible ex- planations for this situation. In the normal accumu- lation of refuse during habitation, a certain amount of mixing will inevitably occur, not only from the move- ment of people but as a consequence of digging of pits, postholes or other physical alterations of a con- scious nature. This happened not only during Valdi- via times, but also later, when the portion of the site surrounding Cut J was reoccupied by people of the Guangala Phase. Such stirring up of the refuse would have brought early sherds upward and to a lesser extent moved later ones downward in the manner indicated by the seriation chart.* This situation was not brought out in the first season of work because of the relatively small sample per level in the small stratigraphic cuts. In Cut J, samples as large as 24,000 sherds per level showed only a tenth or a hundredth of a percent frequency for some of the types, but this is sufficient to obscure the trends. In spite of such distortion, Valdivia Shell Stamped is confined to the early levels as it is on figure 50, and Valdivia Multiple Drag-and-Jab Punctate, Valdivia Pebble Polished, Valdivia Applique Fillet and Valdi- via Cord Impressed show a similar restriction to the latter part of the sequence. It should be noted that frequencies for Valdivia Plain and San Pablo Plain are absent from this chart because the types were misclassified by Estrada's laboratory assistants in Ecuador, and the error was not caught until too late to be corrected by reclassification. The large sample from Cut J includes many minor decorated types, most of which occur in a frequency of less than 2 percent so that trends in popularity can only be brought out by magnifying the scale 100 times (fig. 51). Magnification reveals a sharp decline in frequency of Valdivia Fine-line Incised, Valdivia Red Incised, and Valdivia Combed, and less marked drops in Valdivia Cut-and-Beveled Rim, Valdivia Fingernail Decorated, Valdivia Finger Grooved, Valdivia Corrugated and Valdivia Pseudo-Corru- gated. A slight decline seems to have occurred in Valdivia Embossed, although this is obscured by the erratic occurrence of the type. Valdivia Excised, by contrast with the others, shows a tendency to increase in popularity. *A similar conclusion was reached by Lathrap (1962, p. 89) as a result of the stratigraphic distribution of Early and Late Tutish- cainyo sherds in his excavations at Yarina Cocha, Peru. WHOLE VOLUME THE VALDIVIA PHASE 89 The third seriated sequence (fig. 52) is composed of levels of stratigraphic excavations at four additional sites of the Valdivia Phase: G-25, G-54, G-88 and G-84. The sites were occupied successively and each is of relatively short duration compared to G-31. The pottery type occurrences are consequently sub- ject to a minimum of distortion by mechanical mix- ture or disturbance of the refuse and their presence or absence at a given part of the sequence is more readily observed. Trends are generally similar to those rec- ognized at G-31, the greatest discrepancy being in the relative frequencies of Valdivia Incised and Valdivia Brushed. The site of G-84, occupying the upper part of the sequence shows the continuation of many of the earlier trends but deviates markedly in the plain types. In evaluating this disconformity, it should be taken into consideration that surfaces are often eroded, making it difficult to identify polished plain types. Of the typical decorated types, only Valdivia Broad- line Incised, Valdivia Nicked Broad-line Incised, Valdivia Incised, Valdivia Pebble Polished, and Valdivia Brushed remain. Four new types, weakly represented earlier, are diagnostic: Punta Arenas Incised, Valdivia Applique Fillet, Valdivia Carved and Valdivia Zoned Incised. In the middle of the sequence, represented by G-54, trade sherds of Machalilla Phase origin are frequent. For the purpose of more detailed analysis of changes in pottery types, levels for which information was available on motif and vessel shape were selected and interdigitated to reproduce the trends just described. This chart (fig. 53) includes the lower levels of Cut A and all levels of Cuts F and H from G-31, Cut 1 from G-54, Cuts 1, 2, and 3 of G-88 and Cut 1 of G-84. In the case of G-84, the 10 centimeter levels have been combined into 20 centimeter levels to avoid unrealistic prolongation of the late portion of the sequence. Whereas each of the previous charts showed a consistent trend, the juxtaposition of G-31 levels with those of other sites of presumably similar age introduces some problems. The principal diffi- culty is in the tendency toward a double curve of frequency exhibited by Valdivia Brushed and Valdivia Applique Fillet. Valdivia Incised has a less marked tendency in the same direction. However, the plain types and most of the other decorated types produce more consistent trends when G-31, Cut H is placed below G-54, Cut 1. As will be discussed below, this relative position also concides best with trends in vessel shapes and figurine types (figs. 54, 64). The disconformity can be explained by evidence suggest- ing a brief late reoccupation of G-31, with admixture of the later and earlier refuse to such an extent that it cannot be separated by levels. Particularly sugges- tive of this is the high frequency of Valdivia Applique Fillet, which is comparable to that at G-84 and much greater than at G-54. Such an admixture of late types would also distort the percentages of Valdivia Incised and Valdivia Brushed in the manner repre- sented. This conclusion is supported by the presence of a number of late decorated sherds in the surface collection of G-31, as well as of a late vessel shape (Form 22), rare to absent at G-54. The upper levels of Sections D and E of G-31, Cut J show similar evidence of late reoccupation, notably the high fre- quency of Valdivia Applique Fillet and Valdivia Pebble Polished and the presence of Valdivia Zoned Incised, which in the sites of short duration are not associated with the early decorated types. By ruling out differences in pottery type frequency apparently resulting from factors extraneous to the ceramic evolution (mechanical mixture, reoccupation, inconsistencies in classification, etc.), it is possible to make a fourfold division of the Valdivia Phase se- quence that applies to all four of the seriation charts and equates them with one another chronologically. Since ceramic change is continuous, a situation ex- hibited in all of the seriation charts, the point selected for period division is arbitrary to the extent that moving it up or down one or two levels could be de- bated. Our division represents our best judgment after evaluating all lines of evidence available not only from ceramic differences but from differences in other kinds of artifacts. Distinctions between one period and the next are more a matter of relative frequency of types than of presence or absence. Em- phasis has been placed on contrasts in characterizing the period differences (fig. 55), but the continuity implied by the curves of popularity for each type throughout the entire Valdivia Phase should be kept in mind. Except for a small sample from Cut 3 at G-88, Period A is represented only at G-31 in levels below 2.20 meters in Cut A, below 1.80 meters in Section D of Cut J and below 2.70 meters in Section E of Cut J. Valdivia Shell Stamped is the only decorated type confined to Period A. The following types are at maximum frequency during Period A: Valdivia Plain, Valdivia Polished Red, Valdivia Fine-line Incised, Valdivia Combed, Valdivia Cut-and-Beveled Rim, Valdivia Red Incised, Valdivia Modeled, Valdivia Embossed, Valdivia Pseudo-Corrugated, Valdivia Corrugated, Valdivia Finger Grooved, and Valdivia Fingernail Decorated (fig. 55). Valdivia Excised, Valdivia Incised, Valdivia Broad-line Incised, Val- divia Brushed, San Pablo Plain, Valdivia Polished Plain and Valdivia Striated Polished Plain are present. During Period B, represented at G-88 and G-31, decorated types characteristic of Period A continue in diminishing frequency. At maximum frequency 90 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 are Valdivia Polished Plain, Valdivia Striated Polished Plain, Valdivia Excised, Valdivia Incised, Valdivia Multiple Drag-and-Jab Punctate, and Valdivia Cord Impressed. Three new decorated types: Valdivia Red Zoned Punctate, Valdivia Nicked Rib and Nubbin and Valdivia Pebble Polished (the latter possibly beginning in Period A) appear, and Valdivia Plain, Valdivia Polished Red, San Pablo Plain, Valdivia Broad-line Incised and Valdivia Brushed continue. Period C is characterized by increase to maximum popularity of Valdivia Brushed, Valdivia Pebble Polished, Valdivia Red Zoned Punctate, Valdivia Nicked Rib or Nubbin, and San Pablo Plain. Two new decorated types appear and achieve maximum frequency: Valdivia Nicked Broad-line Incised and Valdivia Rocker Stamped. Three others occur errati- cally: Valdivia Applique Fillet, Valdivia Zoned Incised and Valdivia Carved. Still surviving sporadi- cally and in minor amounts are the early types: Valdivia Finger Grooved, Valdivia Fingernail Dec- orated, Valdivia Multiple Drag-and-Jab Punctate, and Valdivia Cord Impressed. Valdivia Excised, Valdivia Incised, Valdivia Corrugated and Valdivia Broad-line Incised also continue, as do all of the plain types. Period D reflects a general decline in both quality and variety of the ceramic complex. Valdivia Broad- line Incised, Valdivia Applique Fillet, Valdivia Zoned Incised, Valdivia Carved, and Punta Arenas Incised are the dominant decorated types. Of those present earlier, only Valdivia Incised, Valdivia Brushed, Valdivia Pebble Polished and Valdivia Nicked Broad- line Incised survive. Punta Arenas Plain is the domi- nant plain type, the others continuing in minor frequency. In order to find out whether changes through time could also be discerned in vessel shape popularity, 23 general shapes were discriminated by the combina- tion of rims of similar form scattered among the vari- ous pottery types as listed on table A. When the percentage frequency of each form had been com- puted and plotted, the resulting graph was arranged so that the sequence of levels of the various sites du- plicated their position in the pottery type seriation (fig. 53), and period divisions were made in the same places. Except for levels below 2.60 meters in Cut A at G-31, the sample was large enough to be plotted. In these early levels, a line continues downward to indicate presence of the form. The graphic presentation of this composite vessel form study (fig. 54) shows several interesting features, not all of which can be attributed to cultural change. For example, the difficulty of orienting correctly small sherds with unlevel lips is probably responsible for the seeming sudden decline of Form 5, and the sudden reexpansion in popularity of Form 7 in Period D. These two forms represent arbitrary subdivisions of a continuous series of rims, in which the orientation ranges from incurving to outcurving. On the other hand, the marked increase of Forms 9 and 17 during the latter part of Period C probably reflects a real situation, since these forms are rather clear cut. The expanded frequency of Forms 21 and 23 during Period D, and the restriction of Form 10 to this late period also seem reliable. The frequency curves for Forms 14 and 22 show abnormal trends. In the case of Form 22, its absence at G-54 is puzzling. A few examples were identified in the surface collection, indicating that the form was present although in insufficient frequency to appear in the sample from Cut 1 at G-54. Form 14 shows two peculiarities. The first is the low frequency in various levels of Cut F at G-31, Cut 2 at G-88, and Level 120-130 of Cut H at G-31. The second is the abrupt disappearance of the form at G-54. Several other forms show less pronounced but still abrupt altera- tions in popularity between the upper level of Cut H at G-31 and the lower level of Cut 1 at G-54, which follows it in the seriated sequence. It is possible that the explanation lies partly, at least, in the fact that the refuse deposit represented by Cut H at G-31 is subject to contamination from earlier periods, since the site exhibits a long-term occupation during Periods A and B. G-54, on the other hand, has no refuse earlier than Period C, and it would seem reasonable to infer that the vessel shape frequencies represented there reflect more realistically the actual situation during Period C. Although many vessel shapes occur throughout the duration of the Valdivia Phase, some show limited periods of maximum frequency. A few are restricted in time. These factors permit characterization of the four periods in terms of vessel shape. In Period A, jars with folded-over rims represented by Forms 13 and 14 are dominant. Form 18 is the only other jar common enough to be considered characteristic. Thick-walled, constricted bowls of Form 4, often with a tetrapod base, are also typical of Period A. Bowls of Forms 5 and 7 frequently have lobed or undulating rims, and irregular lips occur occasionally on jars of Forms 14 and 15. A rare form, whose trend at the beginning of Period B suggests it may be unrepresented in Period A levels only because of the relatively small rim sample available, is Form 2. Also making their appearance during Period A are Forms 1, 3, 6, 11, and 15, all of which are present in very small frequency. Absent are carinated bowls (Forms 8, 9, and 10) and cambered jars (Forms 20, 21, 22, 23). CUT A, 20-40 CM CUT H, 0-20 20-40 40-60 60-80 80-100 100-120 CUT A, 40-60 CUT F, 0-20 20-40 CUT A, 60-80 80-100 CUT H, 120- 130 CUT A, 100-120 CUT F, 40-60 CUT H, 130-140 CUT A, 120- 140 140-160 160-180 180-200 200-220 220-240 240-260 260-280 280-300 300-340 340-360 360-380 380-400 400-420 420-440 4530ñ55 4I20ñ65 4270œ60 4450ñ200 11111111111 ii rm CZ3 I ' 111111 ti I I INI I I I CZI CD I i LUIiMIIMII lllllllllllll 0 10 20 30 40 50 % HI ®x Ed < a O LU cc lo Søt 27o -- ^Q jO œ o tn "- Ul DO. _|0- << >Q > UJ G 84, CUT 1 40-50 CM 50-60 60-70 70-80 80- 90 90-100 100- 110 110- 120 120-130 _G_- 2 5 G-54, CUT 1,0-20 CM 20-40 40- 60 60-80 G-54, CUT 2, 20-30 CM 30-40 40-50 50-60 G-54, CUT I, 80-100 G-54, CUT 2, 60-70 70-80 80-90 90-100 100-110 110-120 G-54, CUT I, 100-120 120 -130 _G_-54, CUT 2, 120-130 G-88, CUT I, 30-50 CM G-88, CUT 2, 0-30 CM 30-60 G-88, CUT I, 50-70 70-90 90-110 G- 88, CUT 2, 60-90 G-88, CUT 3 CD CZj CCj I -I I =3 \/ // / / / / / / / 1 ////// 1 CD CD ? CD CD CD I I I I I I I I I I I I I / / A r7~7~i czzzn V / / A rzvi i z z I F7~3 I I I I 0 10 20 30 40 50% <0 u. _ o Q 3Q ujuj ?t- < a.< < e> CD >œ ace n* ?oi- 5z< / / / / / A \/ ////// /\ E Z ZDZ1 1/ z zzg EZZD PI CDd CD ! 01 1111111 111 MI f llllllllllll 0 10 20 30 40 50 % Q z UJ - H < >o 7^ o < 0. < DIVIA NICKE B-AND-NUBBI DIVIA NICK OAD-LINE IN DIVIA RED NED PUNCT DIVIA PLIQUE FIL JS _i - -Jœ UJ gui o - m -j ; j1" xo HALI NE 1 y z oR 0UJ '> '> "> '> ?> "> s '> "> ?> '> ø> > P '> '> '> ?> > *> ?> '> '> ?> '> -g o3 Ph '> '> 73 73 73 73 73 73 73 73 73 73 73 73 73 73 73 73 73 73 73 73 73 73 73 73 73 ?M 73 73 -t-J is is 13 13 c3 13 13 a 13 13 13 13 13 13 13 13 13 3 13 13 13 13 13 13 13 03 "5 c 3 13 13 San G 13 13 Ph > > > > > > > > > > > > > > > > > > > > > > > > > Ph > > Ph > > D D ? ? D E E E E E ? D ? E n n C D ? D n D D E (?) E E E E E ? ? ? D ? ? E D ? ? B ? n ? ? n ? ? D ? D E E E ? E ? ? D ? (?) ? ? D E E A E El EI E E El e El El E E D (?) n ? (?) D E E ? D ? Figure 55.-Chronological distribution and period of maximum frequency of Valdivia Phase pottery types. quency, ? presence of the type. denotes maximum fre- Analysis of the evidence concerning earliest occur- rence of the various common vessel shapes permits construction of a tentative scheme of vessel shape evolution during the Valdivia Phase (fig. 56). In certain groups of forms, the typological series reflects evolutionary change, characterized by increased angularity of the carination on bowls and of the camber on jars, and by loss of exterior thickening on everted jar rims. Also evident is the diminution in number and variety of vessel forms through time. Unique forms, such as the castellated rim vessel and spouted jar, are shown at the point where the level in which they occur fits the seriated sequence. In addition to changes like the foregoing, which can be measured and graphically expressed, there are certain gradual alterations in the ceramic complex that become obvious only if a group of sherds from Period D is compared with those of Period A. Most striking is the decline in body wall thickness, particu- larly after Period C, but discernable as a general tendency throughout the Phase. Another "drift" is toward simplification or decreasing variability, evi- dent in disappearance of exteriorly thickened and finger-pressed rims on jars and decline in frequency first of red slipping and then of polishing of the interior of jar rims and necks. Consciously produced un- evenness of the rim in the form of lobes and castella- tions in Periods A and B gives way to irregularities resulting from sloppy workmanship in Periods C and D. There is no similar trend clearly evident in excellence of decoration; execution varies from good to bad throughout the sequence. ® Recognition of these temporal differences in dec- oration and vessel shape makes it possible to esti- mate the chronological position of the two sites re- ported by other investigators: G-115 and G-117 (see locations on fig. 2). Although a more complete analysis of material from G-115: San Pablo will ultimately be available, the preliminary report (Zevallos and Holm, 1960), supplemented by our brief examination of the material in Ecuador during 1961, contains sufficient information on chronologi- cally significant features to permit a tentative evalua- tion. First, it is important to note that diagnostic Period A decorated pottery types are absent with the exception of Valdivia Modeled (fig. 57). In vessel shape, two typically early forms are represented: Form 14, jars with folded-over, finger-pressed rim, and lobed-rim varieties of bowls. However, tetrapod bases common in Periods A and B are not mentioned as occurring. These considerations appear to rule out an occupation during Period A. At the opposite end of the scale, Valdivia Carved, Valdivia Zoned Incised and Punta Arenas Incised, types dominant during Period D, are not described or illustrated. The general character of the pottery, including body wall thickness, surface finish and execution of the decoration, does not resemble that characteristic of 94 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 WHOLE VOLUME THE VALDIVIA PHASE 95 Period D. Enumeration of pottery types present at G-115 reveals that they include all those present during Periods B and G, and with three exceptions the types are those reaching their maximum popu- larity during these periods (fig. 57). The exceptions are: Valdivia Modeled, which is rare in Period B and absent thereafter; Valdivia Applique Fillet, which begins in Period G; and Valdivia Broad-line Incised, which is common throughout the sequence so that its late expansion is less chronologically significant than changes in frequency in other decorated types. Weighing of the evidence from vessel shape and decorative techniques and motifs leads to the con- clusion that G-115 was occupied some time during Period B, probably during the latter part when the early decorated types had largely disappeared, and continued in use through Period G. There is no evidence that it persisted into Period D. The sherd sample from BushnelPs site G-117 (see p. 22 and Bushnell, 1951, figs. 51, 52) is very small but several diagnostic decorated types are repre- sented. Valdivia Applique Fillet, Valdivia Red Zoned Punctate, Valdivia Pebble Polished and Val- divia Broad-line Incised are included, all dominant types during Period D except Valdivia Red Zoned Punctate, which is typical of Period G but may have survived into Period D. One design shows a small rectangle formed by short incisions not meeting at the corners, a motif found in Punta Arenas Incised. Superposition of incision on pebble polishing (op. cit., fig. 52a) is a rare practice during late Period C. Com- mon vessel shapes include cambered jars of Forms 21 and 22 and bowls of Form 8, all typical of Period D. When these data are considered in terms of the chronological position of the various traits, it seems reasonable to conclude that G-117 was occupied at the end of Period G or the beginning of Period D, before Valdivia Red Zoned Punctate disappeared but after Valdivia Applique Fillet became a major decorated type. Early in Period G, sherds of Machalilla Phase plain and decorated pottery types make their appearance in Valdivia Phase refuse, reflecting the arrival of this new population on the Ecuadorian coast. The decorated types represented provide a basis for cor- relation between the seriated sequences of the two phases (see pp. 147-148). Figurines Type Descriptions Two principal figurine classes can be differentiated by material of manufacture: stone and pottery. Stone figurines fall into three readily distinguishable types, which are descriptive entities as well as units with Decorated Types cised u 13 G o3 .o t-H 3 u 73 O v 5 fc lin 73 U ise Excised Incised Brushed Pebble RedZo Nicked Nicked Rocker Broad-1 03 03 03 03 08 03 2 2 03 2 03 Q > > > > > > '> "> "> '> g O 2 73 73 73 73 '-3 73 73 73 73 73 73 13 13 13 13 13 13 13 13 13 13 13 Ph > > > > > > > > > > > D ? D ? ? E E C ? ? E E E E E E ? ? B ? E E ? ? ? D ? A E ? ? ? ? Figure 57.-Pottery types represented at G-115 and their pe- riod distribution in the Valdivia Phase sequence, based upon figure 55. E denotes maximum frequency, ? presence of the type. probable chronological significance. Pottery figu- rines are more variable and the number of descrip- tive types recognized depends on the features selected for emphasis. After a number of attempts at more detailed breakdown, a fourfold division was selected as the only one that showed significant chronological association. Palmar Plain Material: Shale, fine-grained sandstone or mudstone. Method of Manufacture: Thin, flat slabs used in their natural form if approximately symmetrical, or with ends squared off to create a rectangular outline with slightly rounded corners. Surface Treatment: Typically waterworn producing a smooth and even or slightly undulating surface and rounded edges. Shaped edges are similarly rounded. Surfaces sometimes show lengthwise scratches or striations (pis. 117 g-h, 187 h-j). Anatomical Features (pis. 117 a-q, 187 h-j): No details of head or body are indicated in outline or by incision, and the identification of these objects as figurines is by analogy with the Palmar Notched type. Outline is rectanguloid with parallel straight or slightly convex sides and straight or slightly rounded ends; slightly irregu- lar, with nonparallel or converging sides (pi. 117 f, h); or asymmetrical, with one side straight and the other parallel for part of the length, then slanting to meet the other side at a rounded point (pi. 117 I-m). 96 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Dimensions: Length 3.8-8.5 cm., width 1.1-3.2 cm., thick- ness 0.3-1.5 cm. Chronological Position: Limited to Period A (figs. 50, 64). Palmar Notched Material : Mudstone or shale. Method of Manufacture: Naturally eroded, elongated slabs or flat stones carved into symmetrical shape. Surface Treatment: Where shaping has been limited to straightening of sides, the waterworn surface is smooth and sometimes even. When shaping has been more extensive, the surface may show prominent lengthwise striations. Anatomical Features (pis. 117 r-ee, 187g): The only anatomical detail is a central incision, nick or notch at one end, to give a suggestion of legs. Shaping in general is more consistent than in the Palmar Plain type, so that sides are more consistently parallel and straight to slightly convex. The upper or head end is square to rounded, and may be the same form as the lower end, or slightly broader or slightly narrower. One example has horizontal zig- zag lines where the face would be (pi. 187g). Cross sec- tion is rectanguloid with angular or rounded corners, or ovoid. The notch varies from a parallel sided, narrow slit (pi. 117 r, u) to a V-shaped cut (pis. 117 s-t, 187g). Occasionally there is an incision on both faces sometimes with a shallow nick at the edge (pi. 117 x, cc). The edges of the nick are not smoothed over or rounded. Dimensions: Length 3.7-7.0 cm.; width 1.1-3.2 cm.; thickness 0.5-1.5 cm., majority 0.5-0.8 cm. Chronological Position: Limited to Period A (figs. 50, 64). Palmar Incised Material: Pale greenish mudstone or pink shale. Method of Manufacture: Pebbles shaped to produce a symmetrical elongated form with slightly convex sides and flattened or slightly rounded ends. All examples show modification of the original outline and surface with correspondingly greater consistency in shape than in Palmar Plain and Palmar Notched types. In some cases, the lower end has been cut off usually at a slight slant rather than at right angles to the length. Surface Treatment : Generally even, but sometimes leav- ing lengthwise striations or scratches. In some examples, the surface is smooth to the touch, but small pits remain visible. Anatomical Features (pi. 118 a-j): All examples have incision on one surface typically showing recognizable facial features, and occasionally stylized arms. Incisions are 0.5-1.0 mm. in depth and up to 2 mm. wide. Execu- tion varies from good, with symmetrical spacing, to crude and irregular. The nose is typically parallel sided, join- ing the horizontal eyebrows in a "T." Eyes and mouth are oval or rectanguloid with a central horizontal slit. Shallow vertical incisions sometimes border the mouth or cover the cheeks. Arm stylization is typically in the form of several concentric angular or curved U-shaped elements sloping downward from both sides toward the center front. A vertical line may run down the center from chin to base. The lower end of the figurine is finished in two alterna- tive ways. Some examples have short legs like those of Palmar Notched type, formed by a parallel-sided narrow notch (pi. 118 d, j). More commonly, the lower end is cut off cleanly about at the waist, which usually corre- sponds approximately to the greatest thickness of the figurine. Although the surface is flat, it is generally not horizontal but sloping from front to back or side to side (pi. 118, b, e-f, g). The top of the head, by contrast, is flattened in a plane perpendicular to the main axis of the body. Except for one example in which a vertical incision runs up the center of the back, the reverse side is smooth and plain. Rare examples classified as Palmar Incised type have a band of geometric decoration instead of a face at one end. Dimensions: Length 2.8-9.3 cm.; width 1.3-4.2 cm.; thickness 0.4-2.5 cm. Chronological Position: Periods A and B (figs. 50, 64). Valdivia Material: Pottery; fine-grained sandy paste like that of Valdivia Plain, oxidized at the surface leaving a medium to dark gray interior. Method of Manufacture: Two small cylindrical coils laid side by side form the foundation of the body and head. Although the coils are pressed together, fracture is frequent along the junction, particularly in the head (pi. 119 a-q). At the lower edge of the torso, the coils remain separated to form the legs. Thin strips and small lumps of clay were added to the coil foundation to round out the body and head. The front of the hair, framing the face, was added last when the surface was already beginning to dry, judging from the frequency with which it sloughs off (pi. 119 r-z). Surface Treatment: Most of the body fragments show vertical smoothing tracks, and except for the crudest examples the surface is red slipped and partially or com- pletely polished. The face and hair are unslipped, except in a few examples of zoned-red hair treatment. Unslipped surfaces are light tan, light orange or light gray and may show fire clouding. Anatomical Features (fig. 65, a-i; pis. 120-124): Head: The neck is often disproportionately long and the face is slightly raised above the neck plane, producing a slight chin. The top of the head is rounded or flattened, the sides typically curved or parallel. Eyes and mouth are represented by slits or gashes, and eyebrows by a thin line like that forming the eye or by a longer, arched incision, sometimes curving down at the sides. The face is flat and the nose is never shown either by relief or incision. The hair forms a frame on three sides of the face, projecting above it so that the face is inset to a depth of several millimeters. The back of the head is convex. In constract to the highly stylized rendition of the face and body, hair styles are variable and sometimes whole volume THE VALDIVIA PHASE 97 elaborate. About 60 percent of the heads have a simple hair style best described as a long bob, with bangs above the forehead and the hair falling straight at the sides and back to the shoulders or below (pis. 120-121, 122 a-d, g-h, 123 a-d, f-t, v-ii). In crude examples, hair surface is left smooth; in better executed ones it is cov- ered with fine incisions suggesting individual hairs but not always running in a vertical direction. Occasionally a geometric pattern occurs that bears no resemblance to a hair arrangement (pis. 120 c-d, 121 a-b). About 20 percent of this type has a broad red-painted groove running from the top of the head down to the end of the hair at the back, suggesting a part (pi. 120 e-f, k-1). In some examples, hair is shorter at the sides, reaching to the shoulders, but falls to the waist at the back (pi. 120 a-b, e-f, k-1). Among more elaborate coiffures are the following: Large loops, or hollow centered "buns" at the sides (pi. 122 e-f, k-1). The surface is not incised. Hair hanging to the shoulder at one side but much longer on the other side and in back (pis. 122 i-j, 123 e, u). 3. A long bob, in which a broad incision like that usually employed to show a part, runs horizontally in an undulating pattern rather than vertically down the back (pi. 120 g-h). Fine incisions are slanting as if to indicate braiding. Hair short in front and long in back, with a promi- nent ridge running from the top of the head down the back giving the effect of an exaggerated pigtail (pi. 122 o-r). A coiffure of the common bobbed type, or of Variant 2, in which part or all of one side of the head is "shaved", red slipped and polished (pi. 121 g-h, o-p). A tall ring or crown-like ornament on the top of the head, below which the hair falls in a long bob. Rare examples are doubled-headed, the two heads typically rising from the shoulder one leaning slightly to each side (pi. 123 cc-ee). Both face forward. On all examples where both heads are present, they share a single body of the characteristic Valdivia type. One fragment appears to have had duplication of the torso as well (pi. 123 dd). Torso: All bodies are nude, although zoned red paint on a few suggests the possibility of body painting. Sides tend to be straight, with little distinction between waist and hips. With a few exceptions, the sex is identifiable as female by the presence of prominent rounded breasts, often large in proportion to general body size (pis. 120 a, e, k, 124 e-j). A few torsos have a trianguloid patch of stippling in the pubic area (pi. 124c). Some have a conical projection on the lower abdomen of similar size and form to the breasts (pis. 122q, 123u, 124 a, h), which might be interpreted as indicative of male sex except that breasts are sometimes also shown. The rear of the torso is the most uniform feature. In all cases, the lower back curves outward slightly and is then cut inward sharply to form angular buttocks, which are rendered more realistic by a vertical incision up the center (pis. 120b, f, j, 122r, 124b). Arms: Arms are treated in three principal ways: A rounded applique strip attached at the shoulder curves down and across the front below the breasts, where a nick usually separates the two hands. The applique may be superficially applied or smoothed onto the torso so that the junction is obliterated (pi. 120 a, c, e, i, k, 121 a, g, 122m). Expansion of the shoulder to a rounded nubbin or stub (pis. 120g, 122 i, o, 124 a, c). Rarely, shoulders are rounded and any vestige of an arm is completely absent (pis. 122q, 123d). Occasional aberrant examples occur. In a few cases, the right arm is in the normal position while the left one is raised so that the hand rests on the chin (pis. 121k, 123g). One shows traces of a former applique strip that ran down the sides and up the back. Legs: With very rare exceptions, the figurines are stand- ing, with the legs straight and separated. Leg form is approximately cylindrical except for a prominent bulge at the rear in the vicinity of the knee or just below it (pi. 124 e-f, 1-cc). Below this, the leg tapers to a rounded stub. Rarely, the end is broadened to produce a foot (pi. 124 y, ee). A few fragments suggest a bent knee sitting or kneeling position (pi. 124 g, dd-ee). Dimensions: Significant measurements are difficult to ob- tain because of the fragmentary condition of most ex- amples. Two complete figurines are 4.6 and 8.8 cm. long. Ten torsos, complete from neck to buttocks, range from 2.0-6.3 cm. in length. The smallest head is 1.7 cm. high; the largest 4.0 cm. high. Chronological Position: Possibly beginning at the end of Period A, this type is characteristic of Period B of the seriated sequence (figs. 49-50, 52, 64) and continues during early Period C. No correlation could be found between chronological position and crudity of execution. San Pablo Material : Pottery; fine sandy paste, breaking with a very irregular edge; incompletely oxidized surface, medium to dark gray core. Method of Manufacture : The double coil foundation is characteristic and there is a tendency to separate down the midline. Sloughing off of added applique is a fre- quent form of breakage (see Valdivia type for details). Surface Treatment: Finishing is variable, including polish, striated polish, even but unpolished or uneven surfaces. Smoothing is typically poorer on the back than on the front and the back of the head is often uneven. Less than one third of the examples are red slipped. When a slip is present it frequently covers the hair as well as the body, but the face is generally left plain. Un- slipped surfaces are tan, or light to dark gray. Gray fire clouds often occur on tan surfaces. Anatomical Features (figs. 58, 65 j-u; pi. 125): Head: The head often rests directly on the shoulders, or is separated by a short neck. In the former case, the chin is receding (pi. 125 a-b, o-p); in the latter, it is sharply defined (pi. 125 u-v). The top of the head in front view may be curved (pi. 125 a, c, e), squared (pi. 125 98 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 0 I 2 CM d Figure 58.-Typical figurine heads of the San Pablo type. i, k) or flared sideways so that the greatest breadth is across the top (pi. 125 m, s), which typically tilts back- ward. The face is convex and projects slightly to markedly forward beyond the plane of the hair. Typi- cally, both sides slope toward a central ridge, which runs from forehead to chin and gives the suggestion of a nose. In a few the projection is limited to the nose region. Eyes and eyebrows are curved incisions of similar form and contour, applied so that the inner edge is slightly or markedly lower than the outer, creating a slanted effect. The mouth is a shorter horizontal cut. The back of the head is slightly to markedly concave from top to bottom, and concave to slightly convex from side to side. Hair frames the face, but in contrast to the Valdivia type is rarely detailed on the back of the head. Hair surrounds the face and is delimited by a slight projection of the surface or an incised line giving the effect of bangs across the forehead and falling straight or curving around at the sides. The hair surface may be textured with fine incisions (pi. 125 k-1), left smooth (fig. 58c; pi. 125 s-t), red slipped (pi. 125 u-v), or decorated with punctation or broad-line incisions (fig. 58b). A central part is rarely shown. More elaborate hair treatment, in which incision is zoned leaving the center front plain (fig. 58 a, d), or a strand hanging down the front at one side (Zevallos and Holm, 1960, pi. 15, lower center), is exceptional. Torso: All bodies are nude. Width tends to narrow at the waist and increase in a realistic fashion at the hips. Small rounded breasts are typical, and stippling of the pubic area occasionally outlined by incision appears to be more common than in Valdivia type. The abdomi- nal bulge also appears to be more frequent. An incision may curve from side to side at the waist (Zevallos and Holm, 1960, pi. 15, lower left, pi. 22, top right, and pi. 23, top left). The upper back is rounded and tapers into the legs, omitting the well defined buttocks charac- teristic of the Valdivia Type (pi. 125 b, p). Arms: Arms are usually restricted to a rounded stubby projection at the shoulder (pi. 125 a, u) or are com- pletely absent, the neck blending directly into the body (fig. 65 t-u). Rarely, applique ribs are added to show arms extending down the body at the side from the shoulder and curving forward near the waist, or the left arm is raised to the chin (Zevallos and Holm, 1960, pi. 19 top left). Legs: Legs taper from a maximum diameter at attach- ment to the torso downward to a rounded end. Cross section is circular. Dimensions: No complete examples exist. Head length ranges from 2-6 cm. Chronological Position : Periods B and C (figs. 52, 64). Buena Vista Material: Usually, pottery; rarely stone. Fine-grained sandy paste with numerous pin-size air pockets; medium to dark gray core. Method of Manufacture: Double coil foundation is characteristic. Fracture between coil junctions is less common that in Valdivia and San Pablo types, but the two coils are often distinguishable on the irregularly broken edges. Small bits and fillets of clay were used to round out and expand the contours. A few stone examples are executed from naturally eroded pebbles of suitable shape, in which working was restricted to incising the facial features and headdress. Rarely, some shaping may have been employed to im- prove symmetry. Surface Treatment: Finishing is generally superficial, leaving unevenness and even pronounced roughness of the surface. A minority are better smoothed but a really smooth and even surface is rare. Polishing is minimal and usually confined to high portions of the uneven sur- face; most examples are unpolished. Red slipping is a very rare exception. Surface color is typically grayish, occasionally light brown or tan. Firing is poorly con- trolled causing a mottled effect. Anatomical Features (figs. 59, 65, v-gg; pi. 126): Head: The outline in front view typically expands from the neck upward, so that maximum width is across the flat or slightly curved top. Sideward expansion is generally accompanied by slight to marked backward curvature. The back of the head is the most poorly finished part of the surface and never continues the ornamentation applied on the front. The face is typi- cally formed by two sloping planes meeting in a ridge down the center, which constitutes the nose. This prominence is more exaggerated and sharply delimited than in the San Pablo type, often beginning abruptly WHOLE VOLUME THE VALDIVIA PHASE 99 just above the mouth and fading into the sloping fore- head above the eyes. Eyes and eyebrows are typically drawn with fine incisions, but are often not parallel, symmetrically placed or uniform in length and width. The mouth may be indicated by a shorter cut or is absent. In crude examples, eyes and eyebrows may also be absent (pi. 126k). If head or hair ornamenta- tion exists it is limited to a band across the top of the forehead; in many examples it is completely lacking. The simplest treatment is a horizontal or curved in- cision about half way between the eyes and the top of the head (fig. 59a; pi. 126 t-u). Sometimes the upper zone is decorated with parallel vertical incisions (pi. 126i), nicked applique ribs (pi. 126 e, g), or an incised line running down the center and toward one side. With rare exceptions, execution is crude. Fracture line in several examples suggests the original possessed two heads, rising side by side from a common torso (pi. 126 t-w). Stone examples of this type typically consist only of a head, trianguloid in outline with the apex downward (pi. 118 k-r). Eyes, eyebrows (often absent) and mouth are straight or slightly curved incisions. The top of the forehead is usually set off by a horizontal incision and the area above filled with parallel vertical lines. Torso: The torso is short in proportion to the head and legs, typically representing less than one-third of the total length. Small rounded breasts may occur, and when present occupy a major portion of the body sur- face. Punctation of the pubic area is rare. The back is superficially smoothed and shows no anatomical details (pi. 126 m, s). Arms: Arms are totally absent. Legs: Legs taper from a maximum diameter at the point of attachment toward a blunt point, and are circular in cross section. They are often bent slightly forward or slightly backward, and occasionally assume a sitting position. Length is frequently unequal (Zevallos and Holm, 1960, pi. 20 center, left and right). In a few examples, the body tapers from the head toward a central blunt end, producing the effect of a single leg. Dimensions: Length of complete pottery examples ranges from 3.5-8.0 cm., but several large heads suggest that maximum length may reach about 12 cm. Head length is 1.5-3.5 cm. Stone examples range from 3.5-7.3 cm. in length. a j L 2 CM Figure 59.-Typical figurine heads of the Buena Vista type. 100 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Chronological Position: Occurring rarely in Period B, this type is characteristic of Period C (figs. 52, 64). Punta Arenas Material: Pottery. Method of Manufacture: The small sample does not provide clear indications of presence or absence of a double coil foundation. Surface Treatment: Smoothed but not even, with minor defects, pits, ridges and other regularities remaining. Red slipping and polishing are both absent. Anatomical Features (fig. 65 hh-pp): Head: The two examples are flattened and concave at the top, which is the region of maximum width. There is no true neck, the head expanding slightly to markedly from the shoulders. Eyes are represented by deep circular punctations; in one case the mouth is a thin horizontal incision, in the other it is not shown. The face is slightly convex in one, the curvature following the contour of the surface (fig. 6511); in the other, the nose and mouth area is slightly rounded (fig. 65oo). No details of hair or headdress are shown. Torso: The two torsos show variation in anatomical de- tail, and are of normal proportions. Both are flat and ovoid in cross section. One (fig. 65 hh-ii) has two low rounded breasts outlined in fine punctation and with a single punctate at the center. The pubic area has a shallow nearly vertical incision surrounded by fine punctation. The other torso is plain (fig. 6511). v-- "?-'? Figure 60.-Valdivia Phase figurines of unclassified types. a, Pottery figurine resembling the stone Palmar Notched type, length 5.2 cm. b, Rare stone variety, length about 4.3 cm. (after Zevallos and Holm, 1960, pi. 20). Arms: A protrusion at each side on the shoulder is the only indication of arms. In one example, the front has two or three fine punctations (fig. 65ii); in the other, a perforation runs from front to back (fig. 6511). Legs: Legs are separated, rounded in cross section, and taper from a maximum diameter at the junction with the body toward the flattened end (fig. 65 hh-jj). The degree of taper is less than in the Buena Vista type. Dimensions : None is complete. One body, including torso and legs, is 5 cm. long. Chronological Position : Restricted to Period D (figs. 52, 64). Unclassified A few figurines differ sufficiently in style to put them outside the range of variation included in each of the types recognized for the Valdivia Phase. Two pottery figurines, one complete and one in- complete (lower half missing), resemble the stone Palmar Notched Type except for the addition of small conical breasts on the front about a third of the distance from the upper end (fig. 60a). Outline is similar to that of the stone examples, but thickness is greater. The surface is even and polished, leaving vertical striations. Length of the complete example is 5.2 cm. Both are from G-31 and date from Period B. A stone figurine, shaped from a thin sandstone slab, is of particular interest because of its unique eye treatment (fig. 61). Outline follows the rectanguloid form of the early stone types, with the addition of indentations at the sides to indicate the waist, and a shallow vertical incision to separate the legs. The chest is adorned with two large hemispherical breasts. The head is set off by cutting back the neck area, leaving an angular chin line. Gutting back at the sides and top outlines a hemispherical face that contains two ovoid eyes outlined by incisions that continue downward at the outside. A horizontal in- cision through the center completes the eye treat- ment. Sculpturing at the sides of the head and body defines a narrow ridge, which from the shoulder downward somewhat resembles the arm treatment on the Valdivia type but fails to continue across the center of the front. Detail on the back is restricted to a groove separating the head from the body. Existing length (head and torso) is 12 cm. The figurine is from Period B levels of G-31, Cut J. Several stone figurines have a cylindrical form slightly tapering to rounded ends (fig. 60b). The head is set off from the neck by a shallow groove. Two circular breasts are on the upper part of the front, the inner edges abutting. Legs may be set off by incisions that run from the hips downward toward the center where they meet and continue as a single line separating the legs. The head may be plain or WHOLE VOLUME THE VALDIVIA PHASE 101 Figurine Stools Two small zoomorphic pottery objects recovered from stratigraphic excavations at G-31 have been tenta- tively identified as figurine stools. They are of similar form, with flattened, polished tops or backs and four small feet. Projections at the ends, corresponding to head and tail are damaged or broken off. The smaller of the two examples, from G-31, Cut H, Level 1.20-1.30 meters, is the more realistic (fig. 63 d-f). Existing length is 5.2 cm., which excludes the broken-off tail. Width at the center of the back is 2.5 cm., height 1.5 cm. The back has been flat- tened, and forms an angular junction with the nearly vertical sides. The legs are slightly splayed outward and conical in form, coming to a blunt tip. The neck and head are bent slightly downward and the nose, which was directed downward, is missing. The surface is dark brown to gray-brown in color, un- polished except on the back, which is striated polished lengthwise. Fine, parallel, scratchlike incisions run down the sides and across the bottom, but run across the top only on the neck. Two similar incisions run lengthwise on the underside (fig. 63f). Broader lines, 0.5-1.5 mm. in width, were used to draw a cross on one end of the back and a semicircle at the center of each side. Execution of the decoration is slapdash, lines not being equally spaced, straight or evenly parallel. The object rests firmly on the feet so that the back and top are level. J I CM 0 1 2 CM Figure 61.-Unique Valdivia Phase stone figurine. slightly flattened from the midline toward the sides producing a ridge to mark the nose. Curved in- cisions may represent eyes. Length is about 4.5 cm. Examples date from Period C. A few pottery figurines have a cylindrical head shape, with the top flat or concave, set off from the body by a well defined neck (fig. 62). The body is like that associated with the San Pablo Type, tapered inward at the waist on the sides, with small rounded breasts. One lacks arms; the other has them raised against the sides of the head. Facial features are distinctive. Eyes and mouth are short horizontal cuts, while a continuous incision or ridge curves from the side up over the eyes to meet at the center and extend- ing down to eye level to form the nose. All are frag- mentary. Examples come from Periods B and G. A few unique stone examples apparently reflect random experimentation, perhaps the equivalent of doodling. In some cases, the form may have been suggested by the shape of the unworked stone. Figure 62.-Two examples of a rare unclassified type of Valdivia Phase pottery figurine. 102 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 The second example (fig. 63 a-c), from G-31, Cut F, Level 20-40 cm., is shorter in length but taller and broader than the one just described. Existing length is 4.5 cm., width at the center 3.0 cm., height ranging from 2.0 cm. at one side to 2.3 cm. at the other because of the slope of the back. The back is flattened and joins the convex sides at a well defined angle. The four small conoidal feet are more stubby and set closer together, but allow the object to rest firmly. Projections at both ends are broken off, one having been turned sharply upward (fig. 63b). The surface color is medium to dark brown, slightly fire clouded at one end, and striated polished lengthwise. Polish- ing is most complete on the back and upper sides. Around and between the legs, the surface is uneven. None of the defects on the surface are suggestive of decoration. Chronological Distribution and Evolution of Figurine Types Figurines were recovered from almost all excavations in Valdivia Phase sites, the principal exception being the lower levels of G-31, Gut A, where the excavated area was too small for the relatively rare early stone types to be represented. This temporally defined sample is supplemented by a large collection of frag- ments from surface collections and nonstratigraphic tests, which assist in establishing variation within the types and occasionally include kinds not present in the stratigraphic excavations. When occurrence of figu- rine types is plotted according to the seriated se- quences based on ceramic change, correlations between certain of the types and period subdivisions are clearly evident. On the first seriation chart, including the small early tests at G-31 (fig. 49), only figurines of the classic Valdivia type appear. The chart covers Periods A and B and the early part of Period C. Figurines make their appearance shortly after the beginning of Period B and continue for the rest of the sequence. Larger scale excavations, represented by Cut J, Sections D and E at the same site, show that all of the stone figurines types and all but the latest of the pottery ones are actually present at G-31 (fig. 50). Palmar Plain type is distributed from the bottom until the end of Period A, and although the sample is small per level it suggests a tendency toward decline from a maximum frequency at the beginning of the Period. Palmar Notched type, in which one end is notched, seems to begin slightly later. Palmar In- cised type, the most elaborate stone type, is very rare so that its apparent absence during the first half of Period A may not be a true reflection of its chronological duration. Three examples occur in Period B levels. Since no similar prolongation is evident in Palmar Plain and Palmar Notched types in spite of higher frequency of occurrence, it is likely that this represents a real persistence rather than up- ward migration through disturbance of the refuse of the kind manifested by many of the decorated pot- tery types. The pottery figurines are clearly asso- ciated with Period B, only three examples of the Valdivia type occurring in later Period A levels. It also seems clear that the Valdivia type is characteristic of Period B, half of the levels producing 7 or more fragments. Sporadic occurrence of fragments of the San Pablo type and a single example of the Buena Vista type suggest that the inception of these types may be of Period B date. The third chart (fig. 52), compiled by seriation of stratigraphic samples from single period sites, sheds further light on the chronological distribution of figurine types. The Valdivia type is the only one associated with Period B, and except for a single fragment is absent from Period G levels. The San Pablo type and Buena Vista type both occur through- out Period C, with the latter slightly more frequent. The seeming sporadic occurrence is attributable to absence of a typological breakdown for figurines from G-54, Cut 2. Period D is represented only by the Punta Arenas type, which is restricted to this part of the sequence. Coordination of the evidence from these three seria- tions into a single pattern produces a better picture because of differential coverage exhibited by the various excavations. Unfortunately, the interdigita- tion employed for the pottery sequence cannot be adopted without modification because of the small figurine sample from most levels. However, sub- stituting Period A levels from G-31, Cut J for those of Cut A supplies missing information on this part of the sequence. Because of differences in sample size resulting from differences in excavated area, and the impossibility of minimizing these by converting the totals to percentages due to the small samples, the graphic presentation employs a vertical line extending from earliest to latest occurrence (fig. 64). Where it is uncertain whether an early or late appearance reflects migration in the refuse or actual persistence of the type, the line is broken. Certain details of body finish and leg form also show temporal differences when plotted in terms of pres- ence and absence. Red slipped surface is restricted to Periods B and C, being characteristic of the Valdivia type and occasionally associated with the San Pablo type. Plain or unslipped surface is typical of the late figurines and is correlated with Periods C and D. The distinctive bulge at the rear of the leg is charac- teristic of the Valdivia type and shows a similar WHOLE VOLUME THE VALDIVIA PHASE 103 I I L 2 CM Figure 63.-Top, side and bottom views of two probable pottery figurine stools from the Valdivia Phase. 104 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Figure 64.-Temporal distribution of figurine types and details of body treat- ment during the Valdivia Phase. Bars connect earliest and latest occurrence. Dotted extensions reflect uncertainty as to whether terminal occurrence is the result of artificial mixture or actual presence. Levels are arranged in seriated order based on changes in pottery type frequency. temporal restriction, while the straight leg persists through Period D. Bent legs implying a seated posi- tion were recovered only from the latter part of Period C, as represented at G-54. In addition to evidence from our own investigations, more than 2,000 figurine fragments were recovered from excavations at G-115. Sixty-one of those illus- trated (Zevallos and Holm, 1960, pi. 15-24) are classi- fiable into types. Thirty-two represent the Valdivia type, twenty the San Pablo type, and nine the Buena Vista type. None of the early stone types appear to have been found. The Valdivia type is dominant during Period B but extends into early Period C, when the other two are diagnostic (Appendix 1, table 13), a chronological distribution paralleling that derived from the pottery type analysis for G-115 (p. 95). A number of seated fragments also occur (op. cit. pi. 25), a variant that appears to be limited to Period C. The single figurine from G-117 represents the Buena Vista type, also agreeing with the late temporal position indicated by the pottery types. Two small zoomorphic objects have been tentatively identified as figurine stools because their size corre- sponds to that of seated figurine fragments, and the flat top conflicts with the relatively naturalistic modeling of the rest of the body, suggesting a special- ized function. In form, these objects resemble some of the zoomorphic stools from later periods and cul- tures of northwestern and lowland South America. When the chronological distribution of the seven types of figurines is analyzed in terms of evolutionary trends, two principal foci of change are evident: one, from a dominance of stone types in Period A to a dominance of pottery types thereafter; and the other, from one type to another in a continuous series. Examination of each of these foci in more detail will bring to light some of the problems in explanation of the origin and development of the figurine complex. The three types of stone figurines form an obvious evolutionary series, from an unadorned smooth thin slab (Palmar Plain type), to a hint of anthropomor- phic significance in the form of rudimentary legs (Palmar Notched), to a stylized but unmistakable rendering of facial and occasionally anatomical fea- tures (Palmar Incised). Along with increased atten- tion to anthropomorphic detail goes increased shaping of the form, a worked rectanguloid block being sub- stituted for the naturally eroded, elongated pebble characteristic of Palmar Plain. The stone is always a soft variety, easily trimmed and cut. Identification of the unworked flat slabs as figurines of Palmar Plain comes in part from recognition of this evolutionary sequence, in which they are the logical WHOLE volume THE VALDIVIA PHASE 105 antecedent to Palmar Notched. However, additional support comes from the fact that natural rocks of this specific form are restricted to levels identified with Period A. If they were not selected, or were selected for some other function, then it might be expected that they would occur with similar frequency through- out the deposit. In the evolution of stone figurines, the most abrupt transition is between Palmar Notched and Palmar Incised types, in which the faceless slab is provided with a highly stylized and very well executed face. In a few examples a geometric pattern replaces the face, and it is possible that these represent a transi- tional form. On the other hand, vessels of Valdivia Broad-line Incised and Valdivia Excised occasionally are decorated with faces of very similar style, raising the question of whether one served as a prototype for the other. Valdivia Broad-line Incised examples do not come from seriated levels, but the chronological position of this motif in Valdivia Excised (Appendix 1, table 12) is approximately the same as that of Palmar Incised figurines, suggesting that the two may be expressions of the same influence dating from early Period A. Although the three stone types form a logical evolutionary series, and their periods of maximum frequency tend toward a sequential pattern, the small size of the sample and approximate coincidence of the initial appearances of Palmar Notched and Palmar Incised types permits an alternative view that they may represent contemporary variants of greater or lesser crudity, reflecting differences in the skill of the maker or in the use for which the figurine was in- tended, rather than change through time. The earliest pottery figurines, represented by the Valdivia type, are very different in appearance from the abstract stone ones, and no stylistically transi- tional form is evident. An interesting crossover is represented by two unclassified pottery figurines from Period B, which combine diagnostic characteristics of Palmar Notched with a pottery medium of execution, but this sheds no light on the origin of the Valdivia type. Except for a few examples in the late levels of Period A, which may be attributed to downward migration in the refuse, Valdivia figurines are re- stricted to Period B and early Period C. Judging from the frequency with which fragments occur, this is the time when figurines played their major role. The Valdivia type is also the most artistic and the best made of the figurine types, with typically polished red surfaces and a high degree of anatomical detail, particularly notable in the realistic rendition of the back. Although standardized in body detail, Val- divia figurines are outstanding for their variety of hair treatment and relatively realistic appearance (fig. 65 a-i). They constitute a unique and remarkable fea- ture of Period B Valdivia culture. In the seemingly random variation in excellence and alternative treatments of anatomical details, there is a drift in the direction of projection of the face and simplification of the back, which leads to the San Pablo type (fig. 65 j-u). Here the distinctive squared buttocks and bulging back of the leg are absent. The breasts of the Valdivia type are retained and stippling of the pubic area is more frequent. Whereas in the Valdivia type, hair execution is as detailed at the back as at the front, in the San Pablo type it is usually shown only on the front. The back of the head, formerly convex and curving inward toward the top, has become concave, with the top projecting slightly or markedly toward the rear. The diagnostic criterion for purposes of classification is the plane of the face, which projects above the framing hair, often to a grotesque degree. The nose, rarely even hinted at in the Valdivia type, is suggested by a medial ridge or represented by a hooklike prominence from forehead to mouth. Execution of eyes and eyebrows is variable, but there is a tendency to a slanting orientation, the inner end of the incision being lower than the outer one. Surfaces tend to be poorly polished, especially at the back, and red slipping is not typical. Further changes in the direction of simplification result in the Buena Vista type (fig. 65 v-gg). Here the hair ceases to frame the face, and if shown at all is depicted by a horizontal zone across the forehead. Increased stylization of the head and loss of anatomi- cal detail produce an unrealistic result. Some of the better executed examples incorporate simplified survi- vals of earlier characteristic features, such as occasional zoning of the hair (although without red slipping), and lateral projection of the shoulder area to suggest arms, but complete arms are never shown. The double coil foundation construction continues. Most double-headed figurines are of the Buena Vista type. In Period D, there is a notable diminution in the frequency with which figurine fragments are en- countered in the refuse. Only three examples repre- sent this part of the sequence, and they are variable among themselves and unlike those of Period C. Be- cause of this, the Punta Arenas type cannot be de- scribed with the same confidence as the other figurine types, but there are several apparently diagnostic features (fig. 65 hh-pp). The top of the head is flattened and slightly concave so that the sides tend to rise in points-a tendency discernible in some ex- amples of the Buena Vista type. Facial features, instead of formed by incisions, are small circular punctates. The body is unpolished and unslipped, and the legs are conical tapering to a flattened tip. 106 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 cc dd ff J, ø \J q ^Sx t 3 k u I-I-I I 2 3 CM ee hh kk nn oo mm Figure 65 -Evolutionary changes in figurine style during the Valdivia Phase, a-i, Valdivia type, j-u, San Pablo type, v-gg, Buena Vista type, hh-pp, Punta Arenas type. WHOLE VOLUME THE VALDIVIA PHASE 107 Miscellaneous Worked Sherds Aside from figurines, the only artifacts of pottery re- covered from sites of the Valdivia Phase are disks fash- ioned from flat sherds representing almost exclusively plain pottery types. Of the 44 examples tabulated, 19 are Valdivia Polished Plain, 11 Valdivia Polished Red, 5 Valdivia Striated Polished Plain, 4 Valdivia Plain, 2 San Pablo Plain, 2 Valdivia Brushed and 1 Valdivia Incised. Diameter ranges from 1.9 to 5.5 cm. Four have central perforations, and one more is incompletely perforated. One is from the end of Period B at G-88; the remainder are from Period G levels at G-54. Similar disks are reported from G-115 (Zevallos and Holm, 1960, p. 10). The rarity of perforated examples appears to rule out use as spindle whorls. Unfired Clay Objects Six small flattened lumps of well dried but unfired clay (pi. 19 k-n) were recovered from different levels during excavations at G-54. Outline is generally circular but asymmetrical, as indicated by the dimen- sions of unbroken examples: 2.0 by 2.4 cm.; 2.5 by 2.8 cm.; and 3.5 by 3.8 cm. The margin is typically interrupted by several V-shaped notches, where it was "torn" when the lump was flattened. Thickness ranges from 5 to 15 mm., with only one thicker than 10 mm. Surfaces are not smoothed or polished, and one shows traces of what look like finger prints. DIAGNOSTIC FEATURES AND PERIOD SUBDIVISIONS OF THE VALDIVIA PHASE Although the Valdivia Phase exhibits considerable change through time, it is possible to describe general characteristics that are applicable throughout its time span, and which distinguish it from other archeologi- cal complexes on the Ecuadorian coast. Sites are located at the margin of salitres, within a few kilometers from the present shore. The principal component of the refuse is marine mollusks, with the Venus clam {Anomalocardia) the dominant genus. Crab and fish remains are also characteristic. Occa- sional lenses containing charcoal and numerous fire- cracked stones are residue from hearths. There was no direct evidence of house type, but the markedly sloping terrain at some of the sites, and occasional fragments of sun dried mud with stick, twig and grass impressions suggest that floors may have been raised on piles, and walls may have been of wattle and daub construction. Evidence for reconstruction of the subsistence pat- tern of the Valdivia Phase is unfortunately incomplete, both because of failure of some excavators to preserve for analysis faunal remains from many of the excava- tions and because of destruction by the climate of vegetal materials. Restriction of sites to the shore, the abundance of shell and fish remains in the refuse, and the predominance among recognizable stone and shell artifacts of objects relating to fishing all lead to the conclusion that the sea was the principal subsist- ence resource. Knowledge of practices among recent food gathering groups and comparative evidence from the adjacent dry coast of Peru in the preceramic period support the inference that terrestrial plant 767-S41-65 9 food sources must also have been exploited. The presence of fishhooks and weights, either net or fish- line, implies cord. Since wild cotton is found along the Guayas coast today, and domesticated cotton makes its earliest appearance on the north Peruvian coast about 2500 B.C. (Engel, 1963a, p. 10), it can be postulated that cotton was known by peoples of the Valdivia Phase. Whether it was gathered or planted remains an unanswered question. Other plants as- sociated with incipient agriculture in Peru could have been grown, perhaps more readily, in the wetter climate of the Ecuadorian coast. Occasional grinding stones are the only direct evidence, however, of utilization of some kind of plant food. Animal bones are frequent only in Period C, but their scattered occurrence throughout the sequence attests to what could have been postulated, namely that terrestrial fauna were exploited as an additional source of food. How these animals were captured remains question- able. The absence of stone projectile points can be interpreted as reflecting manufacture ol these in a perishable material, and a single bone point supports this inference. It is probable, however, that snares or other trapping devices were also employed. Un- doubtedly a combination of food sources was tapped, with the abundant seafood providing a sufficiently reliable staple to permit long residence in the same place for populations as small and scattered as those of the Valdivia Phase. As might be expected of a food gathering tradition, stone, bone and shell artifacts are relatively crude and nondescript, with the exception of implements asso- 108 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 ciated with fishing. Most stone tools are shaped only to the extent needed to produce a working edge, and although differences in length and contour suggest classification into knives, blades, scrapers, choppers and other general categories, this functional subdivi- sion is not characterized by shape uniformities imply- ing manufacture in accord with a culturally defined norm. Greater standardization of sandstone saws and bulbous reamers results in large part from characteris- tics of the stone. Net or line weights are the most consistent in shape and size, but here uniformity not only has a functional advantage but is created simply by selection of similarly shaped waterworn pebbles. The only numerous shell artifacts are fishhooks of a distinctive circular form, in which the point is sepa- rated by a narrow gap from the bevelled and slightly expanded head. Size is variable, but small in com- parison with hooks of the Machalilla Phase, suggesting a difference in the kind of fishing done by the two groups. Identification of species represented by fish remains indicates that fishing was done not only from land or in shallow water, but also in deeper water, implying some kind of watercraft. Bone utensils are the rarest, and limited to a few antler tip awls or punches. Since fish and animal bones occur in the refuse, this scarcity of bone tools cannot be accounted for wholly on the ground of poor preservation, although bone is the most perishable of the raw materials represented. On the basis of what has survived, it could be con- cluded that the Valdivians made little use of personal ornament. Shell beads and pendants are rare, per- haps reflecting an attitude that shells were "garbage", and consequently not suitable for jewelry. A few whole shells were perforated for attachment or sus- pension, and rarely, fragments of larger shells were worked into pendants. Aside from these objects, per- sonal ornamentation must have been by perishable methods. There is no direct evidence of weaving. If spindle whorls were employed, they must have been of gourd or other perishable material; sherd disks that might have served this purpose are typically unperforated. Cord imprints on pottery are restricted to a single rim with a simple twisted cord and several sherds with imprints identified tentatively as braided cord, all from Period B. If the identification as braid is correct, variations in impressions on different sherds indicate that complicated types of braiding were created. Among the objects of daily use, pottery vastly out- numbers all other kinds of material. A variety of vessel shapes and decorative techniques form the basis for identification of the relative chronological position of sites. Surfaces are even and often polished; 21 percent of the undecorated sherds have a polished red slip. Decoration is limited to plastic techniques, but nearly all known methods are represented. Incision, either broad or narrow, is most common. Other techniques include corrugation, fingernail and finger- tip impression, excision, applique fillet, shell scraping, shell stamping, finger grooving, rocker stamping, punctation, drag-and-jab with a multiedged tool, "pebble polishing" and combing. Broad-line incisions and excised areas are sometimes filled with red ocher. Different combinations of decorative technique and surface treatment create a varied result that is dis- tinctive, attractive and generally competently ex- ecuted. Decoration is common throughout the Phase, but especially during Period C, when few vessels were left entirely plain. Lobed rims and folded-over finger-pressed rims are typical of Periods A and B. Compared with later Ecuadorian archeological com- plexes, bowls and jars are small. Wide mouthed forms are characteristic, and distinctive features in- clude angular (carinated) shoulders on bowls and cambered rims on jars. Small tetrapod feet are characteristic of rounded bowls of Periods A and B; well defined concave bases of jars of Periods B and C. A unique Period G trait is a spout at the shoulder. Bottles are totally absent in any form throughout the Phase. There is no evidence, direct or indirect, that pottery vessels served any but a domestic function. The only evidence of ceremonial life is in the form of small anthropomorphic figurines of stone or pottery. Stone figurines, restricted to Period A, are stylized and standardized as compared to pottery examples, which begin with Period B. Although considerable modi- fication takes place in the pottery figurine style during Periods B and C, the double coil foundation and small size remain characteristic. The earliest pottery figu- rines are the best executed, although variation from fine to crude is a continuum. The abundance of fragments in the refuse suggests that a figurine may have been manufactured for use only once, and having served its purpose became valueless and was dis- carded. Such a situation exists among present day Chocd, Chami, Embera and Cuna Indians of Colom- bia, as described by Reichel-Dolmatoff (Evans, Meggers, Estrada, 1959, p. 10). Here, wooden figurines are carved for curing ceremonies, in which they play an important role. As soon as the cere- mony is over, however, they lose their potency and are often thrown on a trash heap immediately. Since curing ceremonies are the most frequent kind of religious activity, many figurines are produced and discarded during the course of a year. If the figurines were of pottery instead of wood, the remains would accumulate as they did in Valdivia Phase sites with the passage of time. WHOLE VOLUME THE VALDIVIA PHASE 109 The figurines, particularly the Valdivia type which is the most realistic, are of interest for the light they shed on dress and hairstyle. Except for an occasional belt, bodies are nude and seldom show any patterning in surface finish that could be interpreted as represent- ing body paint. Coiffures, by contrast, are varied and some times complicated. A "pageboy" style is most common, with bangs at the front and hair falling to the shoulder at the sides and back, sometimes evenly, other times longer at the back. Other styles include a single braidlike extension at one side or down the back; "shaved" sections on the head, red slipped and polished; a large coil or "bun" at each side; and a crownlike effect, concave at the center. The hair area is typically covered with fine incision, often conforming to the vertical direction of real hair, but sometimes applied in abstract rectilinear patterns. In San Pablo figurines, and more markedly in Buena Vista figurines, execution becomes more abstract and simplified, until they bear no resemblance to any persons living or dead. Burial in the habitation site was one method of dis- posal of the dead, but the relatively small number of skeletons encountered suggests that other methods or other burial grounds were also employed. All of the skeletons were adults, interred in graves excavated to sterile soil. Discovery of a jar containing remains of an infant suggests that urn burial may have been the practice for very young children. Where position was recognizable, legs were tightly flexed and arms ex- tended along the sides. Orientation was not consist- ent, and although most of the skeletons were lying on the right side, at least one lay on the left side. In many cases, disarrangement and decomposition of the bones were too great for reconstruction of original position. Sherds and occasionally a figurine frag- ment came from the adjacent earth, but evidence of grave goods was unquestionable only in one burial, where a polished stone ax was beneath the bones of the lower arm (pi. 12b). Although the continuity of Valdivia Phase culture is obvious, certain changes take place during the long term of its existence that permit characterization of the complex by periods. A few decorative techniques and vessel shapes are excellent time markers because of their relatively restricted temporal distribution, and in a small surface sample a single sherd of one of these types can suggest the period of occupation at a Val- divia Phase site. In terms of such "period markers", and ignoring continuities exhibited by the majority of the cultural elements, the periods can be charac- terized as follows: Period A: Site location: Junction of salitre and shore. Stone artifacts: (None diagnostic). Shell artifacts: Waterworn scoops. Bone artifacts: (None diagnostic). Pottery types: Valdivia Fine-line Incised, Valdivia Red Incised, Valdivia Combed, Valdivia Shell Stamped, Valdivia Finger Grooved, Valdivia Corrugated, Valdivia Cut-and-Beveled Rim. Vessel shapes: Castellated jar rim, folded-over and finger- pressed jar rim, lobed bowl rim, tetrapod base. Figurine types: Palmar Plain, Palmar Notched and Palmar Incised. Period B: Site location: Junction of salitre and shore. Stone artifacts: Sinkers, paint stones, saws. Shell artifacts: Waterworn scoops, bowl. Bone artifacts: (None diagnostic). Pottery types: Valdivia Multiple Drag-and-Jab Punctate, Valdivia Cord Impressed. Vessel shapes: (None diagnostic). Figurine types: Valdivia. Period C: Site location: Inner edge of salitre. Stone artifacts: Choppers. Shell artifacts: Shaped scoops, pecten pendant. Bone artifacts: Antler and sawfish tooth awl, ander projectile point (?) Pottery types: Valdivia Nicked Broad-line Incised, Valdivia Nicked Rib or Nubbin, Valdivia Rocker Stamped, Valdivia Red Zoned Punctate. Vessel shapes: Cambered jars, spouted jar, convex-walled carinated bowl. Figurine types: San Pablo and Buena Vista. Other: Sherd disks. Period D: Site location: Inner edge of salitre. Stone artifacts: (No evidence). Shell artifacts: (No evidence). Bone artifacts: (No evidence). Pottery types: Valdivia Carved, Valdivia Applique Fillet, Valdivia Zoned Hachure, Punta Arenas Incised. Vessel shapes: Concave-walled carinated bowl. Figurine types: Punta Arenas. At the beginning of Period C, pottery of Machalilla Phase types begins to appear in the Valdivia Phase refuse. Valdivia Phase sherds in Machalilla Phase sites attest to an exchange that was probably not limited to pottery vessels. Amicable relations appear to have been maintained between the two populations until the advent of a new group revolutionized life on the Ecuadorian coast and brought an end to both the Valdivia and Machalilla Phases. The Machalilla Phase DESCRIPTION OF SITES AND EXCAVATIONS Description of the Machalilla Phase is based on sur- face collections and stratigraphic excavations at two sites, M-28: Machalilla and G-110: La Cabuya, sup- plemented by a small surface collection from G-112. The area occupied by these sites extends along the coast from south Manabi Province into northern Guayas Province (fig. 2). Machalilla Phase pottery has been reported by other investigators from the Santa Elena Peninsula. G-110: La Cabuya The coast south of the Valdivia Valley curves west- ward to a point, forming the southern limit of the broad Valdivia Bay. The land rises steeply behind a narrow beach, becoming a nearly vertical cliff at the point, where it is subject to constant erosion by the sea (pis. 1, 2b). At the point, the ground is thickly strewn with large boulders, completely obliterating the sandy beach. South of the point is a smaller bay that today is a good fishing ground. The land slopes gradually backward away from the brink of the cliff, and is covered at the present time by dense xerophytic vege- tation, including cactus. Sherds are sparsely scattered over a generally circular area extending from the brink backward (fig. 66). Tests indicated that the depth and concentration of the deposit was greatest at the edge of the cliff, where a large sherd sample was gathered (fig. 66, test). Three stratigraphic excava- tions and a trench were made to secure information on composition of the refuse deposit. Cut 1, 1.5 by 1.5 meters, was placed 3 meters in from the edge of the cliff near the center of the point (fig. 66). The upper 5 cm. of soil was lumpy; below, it be- came loose, powdery clay of a light gray color. Shells, principally representing the Venus clam {Anomalocardia 110 subrugosa Sby.), were present in about the same propor- tion as at the Valdivia Phase site of G-54. Sterile dark gray clay was encountered at a depth of 70 cm. At the east side of the excavation, a poorly preserved human skeleton (Skull M-l; pi. 196) appeared at the bottom of the refuse deposit. Arrangement seemed to be flexed, but the condition of the bones was too frag- mentary to permit detailed observations. No artifacts were associated as burial goods. Cut 2, 2 by 2 meters, was excavated 2 meters in from the edge of the cliff and 5 meters southwest of Cut 1. Soil and refuse conditions were similar to those in Cut 1. Sterile clay appeared at a depth of 40 cm. toward the bank and 50 cm. on the inward side. Cut 3, 2 by 3 meters, was placed at the edge of the summit north of Cut 1 (fig. 66). Below the surface root mass, sherds were abundant. Small lumps of red and yellow ochre were recovered from Level 20-30 cm. Sterile soil appeared at 40 cm. In order to check the composition of the deposit on the inland side, a test trench was excavated 25 meters east of Cut 2 (fig. 66). Originally 3 by 1 meters in a northeast-southwest direction, it was later extended about 2 meters toward the southeast. Depth of the refuse was 50-70 cm.; composition was similar to that at the opposite edge of the site. A burial (M-2) encountered by workmen and left in situ was removed under our direction at the southwest side of Cut 2. Bones were badly crushed and poorly preserved, lying only 10-20 cm. below the surface. The head was toward the west, and the position of tibia and fibula suggested a flexed position of the leo-s. Sherds were wedged around the bones and several fish- hook fragments were found in close proximity, but it WHOLE VOLUME THE MACHALILLA PHASE 111 0 3 6 9 M CABUYA BAY Figure 66.-Sketch map of G-110: La Cabuya, a Period C site of the Machalilla Phase, showing extent of the refuse and location of excavations. could not be determined whether or not this associa- tion was accidental. G-112 A small surface collection made by Estrada from a site at the margin of the salitre behind the present Ayangue Bay (fig. 2) included pottery types identify- ing it as of Machalilla Phase origin. No tests were made to determine the area or depth of the deposit. M-28: Machalilla Cemetery The modern village of Machalilla occupies a small bay along the coast of Manabi Province (fig. 2). The present area of occupation is between the shore and a group of low hills (Estrada, 1958, mapa no. 5). Be- hind and to the north and south are three archeological sites representing different cultures, indicating that the bay was a favored place of habitation over a long period of time. The northernmost site, where Machalilla Phase pot- tery is found, is the only one located on high land. It extends over the flanks of the hill partly appropriated for the modern cemetery, placing it at the northern edge of the former beach, judging by the situation along the Guayas coast. Six small stratigraphic tests made under the supervision of Julio Viteri indicate that the deposit varies in depth between 40 and 120 cm., with sherds sparse below about 80 cm. No burials or other unusual features were encountered in the excavations. A large surface collection was made to increase the sherd sample from the site. Data from Other Investigations Although details are incomplete, data from excava- tions by Bushnell and Disselhoff indicate that the distribution of the Machalilla Phase extends to the northern side of the Santa Elena Peninsula. No sites have yet been reported farther to the south. Bushnell (1951, pp. 17-21) describes pottery and other artifacts recovered from Engoroy cemetery, near La Libertad. Illustrated sherds belong to Ayangue Incised and the description mentions Cabuya Black- on-White. A figurine head fragment believed to be associated is somewhat more realistic than those from G-110, but shows the Machalilla type of coffee-bean eye and prominent nose, as well as a series of perfora- tions at each side (op. cit., fig. 3n). Disselhoff (1949, pp. 347, 351) investigated 5 shell middens at La Carolina. His work followed that of Bushnell, and he indicates that it includes the same sites. The illustrations show principally Regional De- velopmental Period material, but both late Valdivia Phase sherds (op. cit., Tafel I) and late Machalilla Phase sherds (op. cit., Tafel I, II, IV) can be identified. The Machalilla Phase is represented by Ayangue In- cised, characteristic of Period C. Motifs 1, 2, 3 and 5, and vessel forms 1 and 2 can be recognized. Of these, Motif 5 is a late Period G diagnostic (see p. 119). 112 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY DESCRIPTION OF ARTIFACTS VOLUME 1 d J I 0 I 2 CM Figure 67.-Gravers from the Machalilla Phase. Stone Artifacts On the basis of rock material, type of artifact, or method of manufacture, it is impossible to distinguish stone artifacts of the Machalilla Phase from those of the Valdivia Phase. In order to reduce repetition, only unusual features and dimensions are indicated here. If the description is the same only the type is listed with the figure and plate reference. The reader is referred to descriptions of Valdivia Phase stone arti- facts for details of material, form and method of manu- facture (pp. 26-34). Provenience and frequency of stone artifacts are given by type in Appendix 1, table 17. Abraders. Blades or Knives. Choppers. Cores. Gravers (fig. 67). Grinding Stones (pi. 127).-The only grinding stone is a large piece of roughly pecked, coarse sand- stone conglomerate formed into a loaf-shaped mano that was rubbed back and forth in a rocking motion that produced well marked wear on the long edge. Finger and thumb grips were created by pecking pits 1.5 cm. deep and 2.0-2.5 cm. in diameter on one surface and a generally oval region 10 by 5 cm. and 1.0 cm. deep on the opposite surface. The pecked edges have also been partially rubbed over and smoothed. The amount of wear implies that this mano was rubbed across a large grinding slab, but none was found during the limited excavations at G- 110. The hand stone came from miscellaneous testing at the site conducted by local workmen and was assigned to the surface collection. It measures 25.5 cm. long, 10.5 cm. high, and 7.0 cm. wide, and weighs 2.4 kilograms. Hammerstones (fig. 68). "Jaketown Perforators." Paint Stones.-Irregular fragments of hematite (red) and limonite (yellow) may have been a source of pigment. (Only hematite came from Valdivia Phase refuse.) Pebble Polishing Stones (pi. 128). Reamers (fig. 69). Saws (fig. 70). Scrapers (fig. 71). Miscellaneous Stone.-A variety of burnt frag- ments of sandstone, sandstone conglomerates, and indurated clay (mudstone) was found throughout the refuse of the Machalilla Phase, probably representing rocks used in supporting cooking pots. WHOLE VOLUME THE MACHALILLA PHASE 113 Chronological Distribution of Stone Artifacts Since the sample of stone artifacts is limited to Period C of the Machalilla Phase, and to a single site, G-l 10, it is not possible to read any chronological significance into the seriated position of the artifacts (Appendix 1, table 17). Nor is the absence of a few types associated with the Valdivia Phase likely to be reliable indication of a difference between the two Phases. For example, all of the excavations at Valdivia Phase sites produced only 4 notched sinkers; given the same relative fre- quency of these objects in the Machalilla Phase, the amount of digging is too small to expect any to be found. The only unusual artifact is a large two-handed, loaf-shaped grinding stone (pi. 127) showing advanced wear from rubbing against a resistant surface, prob- ably a flat stone. No object of similar shape has been reported either from the contemporary Valdivia Phase or from later coastal Ecuadorian cultures, so that it seems likely to be a diagnostic Machalilla Phase imple- ment. There is no evidence as to what material may have been ground. Shell Artifacts Limited excavations in Machalilla Phase sites pro- duced the same types of worked shell objects as occur in the Valdivia Phase. The major difference is that fishhooks appear to be more common and are typically larger. Pendants, beads and other objects of personal adornment are very rare. To avoid repetition, classi- fication follows categories established for Valdivia Phase shell artifacts (see pp. 37-40), and only unusual features are described here. Similarities and differ- ences in the objects associated with the two Phases can be evaluated by comparing figure 19 and pis. 21-24 with pis. 129-130. Abraders and Polishers Group A, cut and abraded fragments, mostly Spondylus or conch {Strombus) shells, includes most examples. Some of these fragments almost appear as if incompletely worked after they had been sawed from a larger section of shell. Five fragments are from various stratigraphic levels in the seriated sequence (Appendix 1, table 8) and three are from the surface of G-l 10. Only one fragment of Group B, a waterworn piece of conch shell used as a polisher, came from the surface collection at G-l 10. Bead One small shell has the edges rounded and the center perforated. The perforation is 1.3 cm. in diameter and the object is oval, measuring 2.2 by 2.5 cm. It does not appear to be a shell fishhook blank since the type of shell and natural contours, with the sides tapering upward to a smaller diameter at the top than at the bottom, does not occur among fishhooks. Although worn and reshaped so that identification of species cannot be certain, it is clearly one of the class Gastropoda. Bracelet Plate 129q, An 8 mm. wide gently curving section has been cut out of a shell with natural grooves on the surface. The 0 I 2 CM Figure 68.-Small hammerstones from the Machalilla Phase. 114 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 b J I 0 I 2 CM Figure 69.-Fishhook reamers from the Machalilla Phase. WHOLE VOLUME THE MACHALILLA PHASE 115 a b :.-J.^:..::. ;.--.-!y.-*,--/-.-i.-.- 7^ e g h Figure 70.-Sandstone saws from Machalilla Phase. J L 0 I 2 3 CM 767-841-65 10 116 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 fragment is 5 cm. long, suggesting that the recon- structed oval bracelet could have been as large as 8 by 5 cm. Natural thickness of the shell makes the bracelet 5 mm. thick. The edges are well smoothed, erasing saw marks. The piece is too reworked and shaped for identification of type of shell. b J 0 I 2 CM Figure 71.-Snub nosed scrapers from the Machalilla Phase. Disks One disk fragment (Appendix 1, table 8), well polished on all sides and edges, measures 4.5 by 4.0 by 1.0 cm. thick. It is broken on one side, suggesting an originally ovoid contour. The shell is pearl oyster {Pinctada mazatlanica Hanley). Fishhooks Plate 130 The complete size range of fishhooks made of pearl oyster found in the Valdivia Phase occurs here. Also there is a category of thick, large hooks that are unique to the Machalilla Phase. Five of the hooks resemble the small to medium Valdivia Phase variety and 22 represent the larger type (Appendix 1, table 8). Two lips of land snail shells {Porphyrobaphe iostoma Sby.) may have been deliberately shaped into fishhooks, but they also could be an accidental shape resulting from natural fracture and erosion. The large fishhooks are oval, with exterior diameters of 6.5 by 5.0, 5.0 by 4.0, 4.0 by 3.5 cm. Shank and back range from 7-11 mm. thick and 1.0-1.4 cm. wide, tapering to a short point and a flattened shank. Fishhook Blanks Plate 129 g-p, r-s Eight blanks from G-l 10 surface and 16 from levels of the seriated sequence show all the degrees of work- ing exhibited by Valdivia Phase specimens, from rough shaping to circular with edges abraded to incompleted perforation (pi. 129 g-p, r-s). All were shaped from pieces of pearl oyster {Pinctada mazatlanica Hanley). Pendant Blanks Plate 129 a-f Although no finished pendants were encountered in Machalilla Phase sites, a number of elongated blanks unsuitable for fishhook manufacture are sufficiently similar in form to shaped pendants from the Valdivia Phase (pi. 23 c, j-k) to suggest they may be unfinished examples of similar ornaments. One conch fragment has well smoothed edges and is symmetrical except for a slanting termination at the wide end (pi. 129c). The others are Spondylus and pearl oyster. Chronological Distribution of Shell Artifacts Because of limitations of the sample to Period C of the Machalilla Phase, it is not possible to suggest temporal differences in the occurrence of any of the shell artifact types. The complex is generally similar to that of the Valdivia Phase, but shows two significant differences. The most striking is in the size of the fishhooks, three- quarters (22 of 29 specimens) of which are considerably larger than the maximum Valdivia Phase size. The use of larger hooks may be correlated with either a different mode of fishing or emphasis on a different kind of fish. In view of the small size of the Machalilla Phase shell artifact sample, the occurrence of a fragment apparently belonging to a bracelet is noteworthy, particularly since none was identified in Valdivia Phase remains. The absence of such Valdivia Phase items as scoops and pendants, on the other hand, may be the result of inadequate sample size rather than failure of these kinds of artifacts to be made and used in the Machalilla Phase. Bone and Tooth Artifacts The only artifact in this category is a cut and drilled canine tooth of a medium sized carnivore, possibly a dog (fig. 72). The surfaces have been greatly modified by polishing, removing diagnostic features. The tooth is cut off straight across the root end, which measures 1.0 by 0.8 cm. A 2 mm. diameter hole drilled from WHOLE VOLUME THE MACHALILLA PHASE 117 each side, 5 mm. below the cut edge, created a perfora- tion that would have permitted stringing of the tooth to be worn as a pendant. Total length is 2.8 cm. Although antler fragments came from the refuse at G-110, none showed working or evidence of use. 0 I CM Figure 72.-Carnivore tooth perforated for suspension from the Machalilla Phase. Pottery Artifacts Pottery Type Descriptions The pottery of the Machalilla Phase has been classified into six undecorated and sixteen decorated types, on the basis of differences in temper, surface treatment, technique and motif of decoration. Although incor- porating a certain amount of variation in care and skill of workmanship, the pottery complex exhibits several diagnostic characteristics that can be summa- rized briefly. Construction is by coiling. In general, surface finish is even, smooth, and shows some polish. Body walls are typically thin, especially on decorated or polished vessels, numerous examples falling between 2.5 and 4.0 mm. Such thin-walled sherds are not fragile or easily broken, indicating a high tensile strength for the paste. However, wall thickness is variable over small areas and rims are slightly to markedly unlevel. It seems likely that complete vessels, if existent, would show some degree of asym- metry, particularly when undecorated. Principal techniques of decoration are fine incision and red banding, the latter comprising about half of the decorated sherds. Similar rectilinear motifs, such as cross-hatch, zoned parallel lines and slanting paral- lel lines, are typical regardless of technique of execu- tion. Several minor decorated types combine incision or punctation with red bands or zones. A rare late type is blackv-on-white painting. The other common embellishment is nicking, gashing or addition of small nubbins to the angular shoulder of bowls or jars, often in combination with other techniques of decoration. Decoration is less frequent than in the late Valdivia Phase, decorated sherds averaging 18-22 percent of the total per level. Vessel shapes are characterized by angular shoulder treatment, evident in several varieties of carinated bowls and by a tendency to abrupt change of contour at the shoulder of jars. Spouts, both cylindrical and stirrup, are rare but diagnostic. Absence of recog- nized basal sherds implies bases were rounded or slightly flattened, continuing the general contour of the body wall. Pottery types have been designated by the binomial system and descriptions are arranged in alphabetical order with the Spanish name in parenthesis for ease in cross referencing. Information on frequency and provenience of the various types is given in Appendix 1, table 16. Ayangue Incised {Ayangue Inciso) Paste: Method of manufacture: Coiling. Temper: Very fine sand, possibly natural constituent of the clay, often indistinguishable but sometimes visible as white specks in a gray paste. Rare particles up to 0.5 mm. in diameter, usually pinpoint in size. Evenly distributed. Color: Majority have a medium to dark gray core, oxidized adjacent to both surfaces in an even tan to dark brown band. Rarely, sherds are completely oxidized. Texture: Fine grained, compact; breaks cleanly at right angle to the surface. Firing: Incompletely to completely oxidized. Surface: Color: Typically brown to dark gray, occasionally tan to cream. Interior similar shade to exterior except where fire clouded. Occasional sherds have a thin red slip. Treatment: Exterior polished or striated polished leaving horizontal tracks, producing an even, smooth and some- times glossy finish. Better finished on decorated area. Interior typically less polished, occasionally remaining slightly uneven, with polishing tracks more prominent. Surfaces tend to be less polished than average on bowls of Form 1. Hardness: 4-4.5. Form: Rim: Exteriorly thickened, direct, everted or cambered, with rounded, flattened or tapered lip. Body wall thickness: 2.5-6.0 mm. on unthickened areas; increasing to a maximum of 10 mm. at the shoulder on some forms. Base: Probably rounded or slightly flattened. Principal vessel shapes reconstructed from sherds: Open bowl with rounded bottom, high angular shoulder produced by marked body wall thickening, short vertical rim with rounded or flattened lip. Rim diameter 10-22 cm. (fig. 73-1). Decoration re- stricted to exterior between rim and shoulder. Carinated bowl with rounded bottom, angular slightly thickened shoulder, insloping upper wall, up- turned rim with rounded or flattened lip. Mouth 118 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 diameter 14-20 cm. (fig. 73-2). Decoration re- stricted to exterior above shoulder. . Carinated bowl with rounded bottom, slightly thickened angular shoulder, insloping upper wall slightly concave on the exterior, direct rim, rounded lip. Mouth diameter 16-22 cm. (fig. 73- 3). Decoration restricted to exterior above shoulder. . Carinated bowl with rounded bottom, angular slightly thickened shoulder, nearly vertical upper wall slightly concave on exterior, direct rim, rounded lip. Rim diameter 20-24 cm. (fig. 73-4). Decoration re- stricted to exterior above shoulder. . Carinated bowl with rounded bottom, prominent shoulder produced by abrupt thickening of the body wall, sloping inward sharply before joining the out- flaring upper wall, direct rim, rounded or tapered lip. Rim diameter 16-24 cm. (fig. 73-5). Decoration restricted to exterior above shoulder. Bowl with rounded bottom, angular shoulder, short nearly vertical wall and everted rim producing a broad insloping band. Rounded lip. Exterior rim diameter 24-30 cm. (fig. 73-6). Decoration restricted to rim top. Jar with rounded bottom, high angular shoulder, constricted mouth and everted rim with rounded or tapered lip. Maximum body diameter 18-28 cm. (fig. 73-7). Decoration restricted to exterior between base of neck and shoulder. Jar with rounded body, slightly constricted waist, cambered rim, insloping to rounded lip. Interior wall 0 4 8 12 CM VESSEL SCALE 1 , 1 i . i 0 1 2 3 CM RIM SCALE Figure 73.-Rim profiles and reconstructed vessel shapes of Ayangue Incised. WHOLE VOLUME THE MACHALILLA PHASE 119 is typically angular, exterior rounded at "waist". Rim diameter 14-18 cm. (fig. 73-8). Decoration re- stricted to exterior between lip and camber. 9. Globular jar with rounded shoulder, insloping upper wall, direct or slightly expanded rim, rounded or tapered lip. Mouth diameter 12 cm. (fig. 73-9). Decoration restricted to upper exterior. Decoration (pis. 131-134): Technique: Fine, sharply incised lines drawn when the surface was sufficiently dry to leave a clean cut. Edges are often dragged. Incisions vary in width from pencil- line thickness to 0.5 mm. Short lines are typically straight; longer ones are often slightly curved. Execu- tion ranges from equally spaced to markedly irregular and unparallel; overshot or undershot junctions are typical of cross hatch. Occasional designs incorporate fine punctates with incision. Motif: With rare exceptions, designs are composed of straight parallel lines. The exceptions are represented by typical designs with the addition of a row of con- centric arcs (pis. 131 c, g, m, 132e) and occasional sub- stitution of zoned punctate for zoned hachure (pis. 133p, 134 1-n). Typical motifs include: Triangles placed apex up or apex down, or alternat- ing one up and one down, filled with cross hatch or parallel hachure (pis. 131, 132 a-h). Groups of four to six parallel lines slanting in alter- nating directions with terminations overlapping or separated (pi. 132 i-y). Diamonds delimited by single or multiple intersect- ing slanting lines (pi. 133 a-n). Cross hatch in a solid band or in slanting zones (pi. 133 o-v). Checkerboard composed of alternate vertical and horizontally hachured squares (pi. 134 a-d, g). Although complicated patterns may include several motifs, a typical vessel illustrates only one. Associated techniques: Occasionally, carinated bowls have embellished shoulder treatment (pis. 132 a, c, i-j, 1-n, 133t, 134x). Temporal Differences Within the Type: With the pos- sible exception of Motif 5, which is largely restricted to the end of Period C, no temporal differences are evident in decoration. In vessel shape, Form 5 is limited to the latter part of Period C and consequently diagnostic of the Late Machalilla Phase. Chronological Position of the Type: Absent from the early part of the Machalilla Phase, Ayangue Incised be- comes a popular decorated type in Period C, and con- tinues with increasing frequency into the succeeding Chorrera Phase (fig. 89). Cabuya Black-on-White {Cabuya Negro sobre Blanco) Paste: Method of manufacture: Coiling. Temper: Fine sand or occasionally white sand, with size of particles ranging from very fine to pinpoint size. Some 0 4 8 12 CM VESSEL SCALE I i I 0 I 2 3 CM RIM SCALE Figure 74.-Rim profiles and reconstructed vessel shapes of rare Machalilla Phase decorated types. Top, Cabuya Black-on-White. Bottom, Cabuya Finger Pressed Rim. 120 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 sherds have no visible temper. Texture: Compact, typically fine grained. Color: Orange or gray-brown through the cross section; rarely, orange along both surfaces leaving a gray core of even width. Firing: Incomplete to complete oxidation. Surface: Color: Slipped surface: Cream to whitish. Unslipped surface: Light orange, light tan, light gray, whitish; gray tone typical of polished surfaces. Treatment: Slipped surface: White slip varying from paper thin coat- ing obliterating the underlying surface and easily seen in cross section to a streaky wash of uneven thickness incompletely and unevenly masking the underlying orange color resulting in a mottled effect. Polishing striations running parallel to the rim remain visible but not prominent; polishing is most complete on thickest slip and results in an even, smooth finish. Unslipped surface: Even but unpolished, or striated polished but not always sufficiently to remove pits and scars. Hardness: 4-4.5. Form: Rim: Direct with rounded lip. Body wall thickness: 2-5 mm. Base: Probably rounded or slightly flattened. Principal vessel shapes reconstructed from sherds: Shallow bowl with upcurving wall, direct rim and rounded lip. Rim diameter 16-22 cm. (fig. 74-1 top). Decoration on exterior. Several small body sherds with unpolished plain interiors may be from jars, but in the absence of rims this identification is uncertain. Decoration on the exterior. Decoration (pi. 135 a-e): Technique: Narrow black lines on a white slipped surface. Color is clearly black on about 50 percent, dark reddish- brown on the remainder. Line width varies from 0.5- 2.5 mm., with consistency of width more characteristic of narrower lines. Seven examples have narrow lines, between 0.5 and 1.0 mm.; three have wider lines, between 1.0 and 2.5 mm. Variation in width is caused by fail- ure to maintain consistent pressure or even flow of paint, and a single line may increase and decrease markedly in width over a short space. This defect is more evident in wider lines, which may also terminate in an expanded "blob". Lines are relatively straight and parallel, again more precisely drawn at the narrow end of the range. Polish subsequent to painting does not cause streaking as in Machalilla Red Banded, but results in a differential effect of weathering so that eroded lines frequently present a dotted appearance. Two sherds have a red band or zone in addition to black-on-white decoration. Because of their rarity, and because the vessel shape is the same as that associated with black- on-white painting (Form 1) they have been included here rather than left as unclassified. Motif: Bands composed of four to six straight parallel lines, or rectilinear zones filled with straight parallel lines are characteristic. Temporal Differences Within the Type: None discernible. Chronological Position of the Type: Absent from the early part of the Machalilla Phase sequence (fig. 89). Cabuya Finger Pressed Rim {Cabuya Decorado con Dedo) Paste: Method of manufacture: Coiling. Temper: White sand (cf. Cabuya Plain). Texture: Compact, fine grained. Color: Gray, gray-brown or orange-brown through the cross section. Firing: Incompletely oxidized. Surface: Color: Medium gray, tan or gray-brown, often variegated rather than solid because of fire clouding and polishing striations. About half have a streaky red slip leaving exposed portions of the underlying surface. Treatment: Striated polished, with well defined marks running parallel to the rim. Variation extends from total coverage to incomplete polish leaving narrow un- polished streaks between the striations. Either treat- ment leaves an even surface, smooth but not slick to the touch. Form: Rim: Everted, expanded, or interiorly thickened with rounded lip. Body wall thickness: 2-4 mm.; increasing to 6 mm. at carination. Base: Probably rounded or slightly flattened. Principal vessel shapes reconstructed from sherds: Bowl with rounded bottom, walls curving upward to everted or expanded rim, rounded lip. Rim diameter 20-22 cm. (fig. 74-1 bottom). Decoration on rim top. Carinated bowl with rounded bottom, angular shoulder, slightly insloping upper wall, everted rim with rounded lip. Rim diameter about 24 cm. (fig. 74-2 bottom). Decoration on rim top. Bowl with rounded bottom, walls sloping upward to interiorly thickened rim, rounded lip. Rim diameter 22 cm. (fig. 74-3 bottom). Decoration on lip and rim interior. Decoration (pi. 135 f-i): Technique: Pressing along the top of expanded or everted rim to produce a continuous row of ovoid depressions. Pressure usually causes horizontal expansion resulting in slight lobing of the lip. Depressions are shallow (less than 0.5 mm. deep) and polished like the adjacent sur- face. Dimensions vary between 5 by 9 mm. and 6 by 10 mm. Margins are usually not sharply defined rounding on to the adjacent surface. Motif: Row of depressions along the center of the flat rim top (pi. 135 g-i), or rarely on the lip (pi. 135f). Associated techniques: Incisions of the type associated with WHOLE VOLUME THE MACHALILLA PHASE 121 Machalilla Incised may border the depressions (pi. 135 f, h). Temporal Differences Within the Type: None discernible. Chronological Position of the Type: Absent from the early part of the Machalilla Phase sequence (fig. 89). Although the sample is too small to be conclusive, it appears probable that Cabuya Finger Pressed Rim is limited to Period C. Cabuya Plain {Cabuya Ordinario) Paste: Method of manufacture: Coiling. Temper: Angular, opaque white sand particles from pin- point size to 5 mm. in diameter and rarely to 8 mm. are characteristic of 75 percent of the sherds. The remain- ing 25 percent have sand with few white particles. In both varieties temper is abundant, and the contrast between the white inclusionsand^a dark paste is striking. Mixture is insufficient to produce even distribution, and larger grains show a tendency to clustering. Occasional sherds contain finely ground shell particles in addition to sand. Texture: Sandy but not friable, compact, with very rare fissure-like air pockets lying parallel to the surface. Color: Light to dark orange throughout the cross section; medium to dark gray throughout the cross section; occasionally fired orange along one or both surfaces leaving the remainder gray. Firing: Incompletely to completely oxidized. Surface: Color: Buff, light to medium orange, red-orange, brown, gray-brown, medium to dark gray. Interior may be similar in shade to exterior or at opposite end of the range. Fire clouding may produce abrupt contrasts in color. Treatment: Scraped to remove gross irregularities and obliterate coil junctions, but leaving an uneven finish with frequent scars, pits and rough areas. A few sherds have limited polish, but remain too uneven to be classi- fied as polished. Temper grains remain visible espe- cially on the exterior. The interior may show fine parallel scraping marks. Superficial smoothing is reflected in inequality of wall thickness over small areas. Hardness: 4-4.5. Form: Rim: Direct or everted with rounded or tapered lip. Rim is unlevel and poorly circular, making diameter esti- mates difficult on small sherds. Body wall thickness: Typically 2.5-9.0 mm.; occasionally reaching 1.2 cm. Base: Probably rounded or slightly flattened. Principal vessel shapes reconstructed from sherds: Bowl with rounded bottom, slightly angular slightly thickened shoulder, nearly vertical convex upper wall, direct rim with tapered lip. Rim diameter 12-20 cm. (fig. 75-1). Carinated bowl with rounded bottom, angular shoulder, concave insloping upper wall, direct rim and rounded lip. Rim diameter 16-20 cm. (fig. 75-2). Rounded bowl with constricted direct rim, rounded lip. Rim diameter 16 cm. (fig. 75-3). Wide mouthed jar with flattened bottom, rounded shoulder, constricted mouth, everted rim with tapered or rounded lip. Junction of rim and neck may form a sharp angle on the interior. Rim diameter 20-26 cm. (fig. 75-4). Jar with rounded body, constricted neck and everted rim with rounded, tapered or (rarely) flattened lip. Rim diameter 6-14 cm. (fig. 75-5). Temporal Differences Within the Type: The range of variation in surface finish remains similar throughout the duration of the Machalilla Phase, but more poorly smoothed examples show a decline in frequency. Chronological Position of the Type : The principal un- polished plain type of the Machalilla Phase (fig. 89). M >> 0 1 œ 3 CM RIM SCALE 0*4 8 12 CM VESSEL SCALE Figure 75.-Rim profiles and reconstructed vessel shapes of Cabuya Plain. Chorrera Incised {Chorrera Inciso) Paste: Like Chorrera Plain (p. 122), see that type descrip- tion for details. Surface: Color: Interior: Light to medium gray, or light tan. Exterior: Medium to dark gray, or red. Treatment: Interior: Even, unpolished. Exterior: Well polished producing a smooth, even finish 122 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 slick to the touch; occasionally polishing striations may remain visible. Hardness: 4-4.5. Form: Only four sherds are represented. One is the everted rim of a constricted neck jar (rim diameter 10 cm.); the rest appear to be from bottles or spouted jars, by analogy with similar Chorrera Phase examples. Decoration (pi. 135 j-n): Technique: Incisions ranging from fine scratches to 0.5 mm. in width cutting through the polished surface to the lighter subsurface. Depth is less than 0.3 mm. Lines are sharply defined, occasionally with dragged places along the margin, but this is not typical. Width varies slightly because of variable pressure. Parallelism is good, but spacing of parallel lines not equal. Overshot or undershot intersections are rare. Motif: Straight parallel lines 3-10 mm. apart filling zones bounded by lines of similar width. Zones are separated by plain areas (pi. 135 1-n). Temporal Differences Within the Type : None. Chronological Position of the Type: Beginning in Period C of the Machalilla Phase, this type becomes characteristic of the following Chorrera Phase (fig. 89). Chorrera Plain {Chorrera Ordinario) Paste: Method of manufacture: Coiling. W 0 4 8 12 CM VESSEL SCALE I . I . I . i 0 1 2 3 CM fffii Figure 76.-Rim profiles and reconstructed vessel shapes of Chorrera Plain. Temper: Fine sand similar in color to the paste and dis- cernible only on close inspection. Rare particles may reach 2 mm. but size is typically less than 1 mm. Rare isolated white grains may occur. Texture: Compact, fine grained, not sandy to touch. Color: Medium to dark gray through the cross section or fired orange along one or both surfaces leaving a gray or brownish core. A few are orange or orange-brown throughout the cross section. Firing: Incompletely to completely oxidized. Surface: Color: Light orange, orange-brown, brown, gray-brown or medium to dark gray, with interior and exterior generally similar in shade. Some mottling results from fire clouding. Treatment: Smoothed leaving even but unpolished, or incompletely smoothed leaving defects and tracks. Smoothing is better than in Cabuya Plain, producing more uniform wall thickness. Hardness: 4-4.5. Form: Rim: Direct, tapering or everted, with rounded or flattened lip. Body wall thickness: 3-6 mm. Base: Probably rounded or slightly flattened. Principal vessel shapes reconstructed from sherds: Bowl with rounded bottom and high angular shoulder created by decreasing body wall thickness and slightly insloping the upper wall to a tapered, rounded or flattened lip. Rim diameter 14-24 cm. (fig. 76-1). Carinated bowl with rounded bottom, slightly angular shoulder, insloping convex upper wall, direct rim, flattened lip. Rim diameter 20 cm. (fig. 76-2). Globular bowl with slightly constricted direct or interiorly thickened rim, flattened lip. Rim diameter 14-24 cm. (fig. 76-3). Wide mouthed jar with flattened bottom, rounded shoulder, slightly constricted mouth, everted rim with tapered, rounded or flattened lip. Rim diameter 20-28 cm. (fig. 76-4). Jar with rounded body, constricted neck and everted rim with rounded lip. Rim diameter 8-14 cm. (fig. 76-5). Temporal Differences Within the Type : None dis- cernible. Chronological Position of the Type: A minor plain type during the Machalilla Phase, Chorrera Plain becomes an important type during the succeeding Chorrera Phase (fig- 89). Machalilla Burnished Line {Machalilla Bninido) Paste: On Chorrera Plain or Cabuya Plain (p. 121); see those type descriptions for details. Surface: Color: Light tan, orange-brown, gray-brown, reddish brown, or medium gray, with mottling sometimes resulting from polishing variation. Treatment: Even and smooth or slightly abrasive to the touch. Striated polished on bowl exteriors; unpolished on jar interiors. Hardness: 4-4.5. whole volume THE MACHALILLA PHASE 123 Form : Restriction of patterning to one surface indicates that bowls and jars are represented, but there are no rims. Body wall thickness is 4-6 mm. except for one jar sherd with a thickness of 11 mm. Decoration (pi. 136 a-j): Technique: Burnished lines on interior of bowls and ex- terior of jars varying from narrow (0.5-1.0 mm.) marks leaving a distinct groove to a 1-2 mm. wide line, polish- ing but not depressing the surface. Textural contrast is enhanced by strong color differentiation resulting from firing, burnished lines being much darker gray or reddish brown than the adjacent unburnished surface. Motif: Burnished lines of variable width cross and inter- sect in a random and haphazard manner, occasionally accidentally resolving into crude Crosshatch in small zones. Most of the lines cluster along three general axes, but spacing is unequal, resulting in uneven size and distribution of burnished and unburnished zones. Temporal Differences Within the Type: None discernible. Chronological Position of the Type: Scattered errati- cally throughout the seriated sequence of the Machalilla Phase (fig. 89), this type has been recognized because of its possible significance as an ancestor of burnished line decoration in later coastal Ecuadorian ceramic complexes. Machalilla Double-line Incised {Machalilla Inciso Doble Lined) Paste and Surface: About 50 percent red slipped on deco- rated surface and in the case of bowls on the interior also; remainder polished or striated polished. See type descriptions of Machalilla Polished Red (p. 130), Macha- lilla Polished Plain (p. 129) and Machalilla Striated Polished Plain (p. 137) for details. Form: Rim: Direct, exteriorly thickened or everted with rounded, tapered, or flattened lip. Body wall thickness: 0.3-1.1 cm. Base: Probably rounded or slightly flattened. Principal vessel shapes reconstructed from sherds: Carinated bowl with flattened bottom, high promi- nent shoulder produced by exterior thickening of the wall, upper wall sloping inward to exteriorly thickened rim with rounded lip. Mouth diameter 22 cm. (fig. 77-1). Decoration on exterior above carination. Carinated bowl with flattened bottom, angular shoulder not thickened at carination, insloping upper wall, exteriorly thickened rim with rounded lip. Mouth diameter 14 cm. (fig. 77-2). Decoration on exterior above carination. Carinated bowl with flattened bottom, angular un- thickened to slightly thickened shoulder, insloping up- per wall slightly concave on the exterior, tapering rim, rounded lip. Mouth diameter 18-22 cm. (fig. 77-3). Decoration on exterior above carination, Carinated bowl with rounded bottom, angular shoulder slightly thickened to produce a curved in- terior wall, insloping upper wall slightly convex on the exterior, direct rim, tapering or rounded lip. i i i i i i i 0 4 8 12 CM VESSEL SCALE / . . >?, ?. >. ;, \ ?.... < \ I'y yy s~ yy c4 yy y " \ \ 6 y I ? I ? I ? I 0 I 2 3 CM RIM SCALE Figure 77.-Rim profiles and reconstructed vessel shapes of Machalilla Double-line Incised. 124 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Mouth diameter 18-20 cm. (fig. 77-4). Decoration on exterior above carination. Rounded bowl with slightly incurving upper wall, constricted direct rim, rounded or tapered lip. Mouth diameter 16-20 cm. (fig. 77-5). Decoration on upper exterior, extending below region of maximum diameter but not covering the bottom. Jar with rounded bottom, angular shoulder, in- sloping upper wall, exteriorly thickened rim, flattened lip. Mouth diameter 14 cm. (fig. 77-6) . Decoration on exterior rim thickening and upper wall, probably not extending below shoulder. Jar with rounded bottom, slightly angular shoulder, constricted neck, everted rim and tapered lip. Rim diameter 6 cm. (fig. 77-7). Decoration beginning at base of neck and extending downward, probably ending at or slightly below shoulder. Decoration (pis. 137-138): Technique: Paired incisions not always equally spaced or evenly parallel, but sometimes too regular to have been drawn independently, suggesting use of a double-ended tool with the points not rigidly fixed allowing some in- dependent fluctuation. Incisions are very fine to 1 mm. in width and shallow, often resembling scratches. Ap- plication when the surface was dry is indicated by jagged margins, chipping off the adjacent surface. In- tersections of incisions are often overshot and poorly executed. Spacing of paired lines may be so close as to obliterate their dual nature, producing the effect of a series of parallel lines. Typical designs are spaced so as to emphasize the double line effect. Motif: Straight paired lines of different lengths are com- bined to produce a variety of patterns, of which the following are most frequent: Crosshatch (pi. 138a). Zigzag (pis. 137c, h, j, 138 b-d). Diagonal zones filled with vertical lines (pi. 138 e-f) Parallel, horizontal, or vertical lines and short diago- nal strokes (pis. 137g, 138 h-i). Zoned parallel lines (pi. 137 d-e). Concentric rectangles (pi. 138 o-p). Associated techniques: Double-line incision may be combined with red banding, the incision applied either to red or to plain areas. Some bowls have embellished shoulder treatment. One jar body sherd has low relief ridges and nubbins incorporated in the red banded and in- cised design. Temporal Differences Within the Type : None discerni- ble in the small sample available. Chronological Position of the Type: Restricted to Period C in the seriated sample available (fig. 89), but probably beginning in Period B (see p. 141). Machalilla Embellished and Red Zoned {Machalilla Rojo en %pnas) Paste and Surface : Like Cabuya Plain except that surface finish on exterior is more even and may show a slight polish; see Cabuya Plain (p. 121) for detailed description. Form: Rim: Everted with rounded or tapered lip. Body wall thickness: 4-1 mm., increasing to 8-12 mm. at junction of neck and body. Base: Probably rounded or slightly flattened. Vessel shape reconstructed from sherds: 1. Jar with globular body, constricted neck and everted rim. Junction of neck and body wall may be angular on the interior. Interior neck diameter 4-6 cm. Decoration (pi. 142 g-k): Technique: Rim red slipped beginning at the angular in- terior junction with the body wall and continuing down the exterior to the base of the neck or about 1 cm. farther along the body wall. Slip is like that characteristic of Machalilla Polished Red, and polished horizontally leaving parallel striations. Embellishments are of three types: Irregular punctates of generally ovoid form and variable size (pi. 142 i-j). Ovoid nubbins, 3-5 mm. long and rarely exceeding 1 mm. in elevation (pi. 142g). "Molded" circular nubbins 3 mm. in diameter and less than 1 mm. in elevation (pi. 142 h, k). Motif: A single row of embellishments typically occurs just below the limit of the slip on the exterior. One example has a double row of punctates; another has the embellishments just inside the border of the slip. Temporal Differences Within the Type: Techniques 1 and 2 are restricted to Period C, Technique 3 to Period A (table B), paralleling temporal differences evident in Machalilla Embellished Shoulder (table C). Table B.-Temporal distribution and frequency of techniques oj Machalilla Embellished and Red %oned decoration Period Seriated sequence Technique of Embellish- ment 1 2 3 C G-l 10, Cut 1, 10-20 cm. G-l 10, Cut 2, 0-20 cm. G-l 10, Cut 1, 20-30 cm. G-l 10, Cut 1, 30-50 cm. G-l 10, Cut 1, 50-70 cm. G-l 10, Cut 2, 20-40 cm. 1 1 1 1 to I 1 2 1 1 1 1 1 1 1 B G-112 - - - A M-28, Cut 1, 0-20 cm. M-28, Cut 1, 20-40 cm. - - 1 Chronological Position of the Type: Very rare but present throughout the Machalilla Phase (fig. 89). Machalilla Embellished Shoulder {Machalilla Decorado Plastico) Paste and Surface: On Machalilla Polished Plain (p. 129), Machalilla Striated Polished Plain (p. 137) or rarely Machalilla Polished Red (p. 130); see those type descriptions for details. whole volume THE MACHALILLA PHASE 125 ' ' ' ' ? ' ' 0 4 8 12 CM VESSEL SCALE l i i_ 0 I 2 3 CM RIM SCALE Figure 78.-Rim profiles and reconstructed vessel shapes of Machalilla Embellished Shoulder. 126 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Form: Rim: Exteriorly thickened, direct or everted, with rounded or tapered lip. Body wall thickness: 2.5-7.0 mm., increasing to a maximum of 1.3 cm. at carination on bowls of Form 1. Base: Probably rounded or slightly flattened. Principal vessel shapes reconstructed from sherds: Carinated bowl with rounded bottom, high angular shoulder produced by pronounced body wall thicken- ing, short insloping upper wall, upturned rim with rounded lip. Rim diameter 8-28 cm. (fig. 78-1). Decoration restricted to shoulder. Carinated bowl with rounded bottom, angular shoulder slightly thickened to produce curved in- terior wall, insloping upper wall, exteriorly thickened rim, rounded lip. Rim diameter 14-30 cm. (fig. 78- 2). Decoration restricted to shoulder. Carinated bowl with rounded bottom, angular unthickened shoulder, insloping upper wall, exteriorly thickened rim, rounded or tapered (rarely flattened) lip. Rim diameter 12-26 cm. (fig. 78-3). Decora- tion restricted to shoulder. Carinated bowl with rounded bottom, angular shoulder, insloping upper wall concave on the exterior, direct rim, rounded or tapered lip. Rim diameter 16-24 cm. (fig. 78-4). Decoration restricted to shoulder. Carinated bowl with rounded bottom, slightly angular shoulder, insloping upper wall convex on exterior, direct rim, rounded or tapered lip. Rim diameter 16-24 cm. (fig. 78-5). Decoration restricted to shoulder. Cambered jar with rounded bottom, upper wall sloping outward to angular junction, then insloping to exteriorly thickened rim, rounded or tapering lip. Rim diameter 8-26 cm. (fig. 78-6). Decoration restricted to angular camber. Jar with rounded bottom, angular shoulder, insloping upper wall, everted rim, rounded lip. Rim diameter 12-26 cm. (fig. 78-7). Decoration restricted to shoulder. Stirrup spout jar with rounded bottom, high angular shoulder, stirrup spout with everted rim, rounded lip. Shoulder diameter 16-22 cm. (fig. 78-8). Majority have a second "tier" above the shoulder; decoration applied to both angles. Decoration (pis. 139-142 a-f, 143): Technique: Embellishment of the angular shoulder in one of six techniques: Nubbins, small polished bumps from barely visible to 7 mm. in diameter and 2 mm. elevation, occurring continuously or in pairs (pi. 139). Gashes, 2-4 mm. wide, less than 1 mm. deep, applied continuously but not equally spaced (pis. 140, 141 a-b, g-j). Nicks or cuts, typically about 1 mm. wide, 2-5 mm. in vertical dimension, sharply defined, spaced 3-6 mm. apart (pi. 141 c-f). "Molded" flat-topped nubbins, typically approxi- mately 2 mm. in diameter, occasionally 5 mm. diameter, but uniform in size on a single specimen, applied in a continuous row spaced 1-5 mm. apart, with spacing varying little on a single sherd (pi. 142 a-f). 5. Slight undulation, probably resulting from poorly defined gashes. 6. Row of punctation just above angular shoulder. Motif: Row of embellishments of one of the six techniques along the carination or shoulder of the vessel. Only one technique occurs on a single specimen. Associated techniques: Embellishments may occur on the angular shoulder of carinated bowls with incised or painted decoration; in such cases, the sherd has been classified in one of the above types rather than as Machalilla Embellished Shoulder. A few sherds have large bosses projecting from the angular shoulder (fig. 78, insert; pi. 143). These are approximately 2.5 cm. in diameter, typically flattened on the upper side, and project 0.5-1.5 cm. from the shoulder. Interior is hollow and opens into the interior of the vessel. Temporal Differences Within the Type: Analysis of frequency of techniques according to the seriated sequence shows that Technique 4 is restricted to Period A (table C). Technique 6 is absent during the same period. The remaining techniques occur throughout the Phase, and frequencies are too small to be reliable indications of changes in popularity. Table C.- Temporal distribution and frequency of techniques oj Machalilla Embellished Shoulder decoration Period Seriated sequence Technique 1 2 3 4 5 6 C G-l 10, Cut 1, 10-20 cm. G-l 10, Cut 2, 0-20 cm. G-l 10, Cut 1, 20-30 cm. G-l 10, Cut 1, 30-50 cm. G-l 10, Cut 1, 50-70 cm. G-l 10, Cut 2, 20-40 cm. 2 5 2 3 18 14 2 5 1 26 1 8 2 4 2 4 - 2 2 2 2 1 1 1 5 2 B G-112 A M-28, Cut 1, 0-20 cm. M-28, Cut 1, 20-40 cm. 27 17 36 16 30 10 17 5 15 5 - Chronological Position of the Type: Machalilla Em- bellished Shoulder occurs in the maximum frequency of 9 percent in the early part of the Machalilla Phase, and continues to decline in popularity thereafter (fig. 89). Machalilla Finger Punched {Machalilla Modelado) Paste and Surface: One Machalilla Striated Polished Plain (p. 137), one Machalilla Polished Red (p. 130); see whole volume THE MACHALILLA PHASE 127 those type descriptions for details. One appears to have been shell tempered plain. Surfaces are poorly smoothed and uneven. Form: Rim: Direct with tapered or flattened lip. Body wall thickness: 4-6 mm. Base: Probably rounded or slightly flattened. Principal vessel shape reconstructed from sherds: 1. Shallow bowl with wall curving outward and upward to direct rim, tapered or flattened lip. Rim diameter estimated at 18-20 cm. (fig. 79-1). Decoration (pi. 136 k-1): Technique: Wall punched inward from exterior just below lip producing an indentation that varies from a deep hole to a concavity in the rim profile. In one example the depression is circular, 1.5 cm. in diameter, 3 mm. deep (pi. 1361); in another, it is ovoid, 2.0 cm. long, 1.1 cm. wide and 7 mm. deep (pi. 136k); in a third, pressure has pushed the lip inward over a 2.5 cm. wide area without a clearly defined depression. Motif: A single depression at or about 2 mm. below the lip; incompleteness of vessels makes it impossible to judge whether another may have been placed on the opposite side. Temporal Variation Within the Type : None discernible. Chronological Position of the Type: Present with extreme rarity throughout the duration of the Machalilla Phase (fig. 89). 0 I 2 3 CM RIM SCALE i ' i i i i i 0 4 8 12 CM VESSEL SCALE Figure 79.-Rim profiles and reconstructed vessel shapes of Machalilla Finger Punched. Machalilla Incised {Machalilla Inciso Pulido) Paste and Surface: Similar to fine sand tempered, well polished examples of Machalilla Polished Plain. In- completely oxidized firing typical. Form: Rim: Direct or slightly expanded with rounded lip. Body wall thickness: 1-4 mm. Base: Probably rounded or slightly flattened. Principal vessel shapes reconstructed from sherds: 1. Rounded jar with constricted mouth, direct or expanded rim, rounded lip. Some examples have a slight shoulder. Mouth diameter 6-12 cm. (fig. 80-1). Decoration restricted to incision adjacent to lip; rarely applied to small zones on the body. Carinated bowl with rounded bottom, slightly thickened angular shoulder, concave upper wall, slightly everted rim with rounded lip. Rim diameter 12-14 cm. (fig. 80-2). Decoration restricted to exterior above shoulder. Carinated bowl with rounded bottom, high angular thickened shoulder, upper wall insloping to slightly thickened rim, rounded lip. Rim diameter 16 cm. (fig. 80-3). Decoration restricted to exterior above shoulder. Rare vessel shapes reconstructed from sherds: 1. Jar with rounded body, constricted neck and everted rim. Interior neck diameter 8 cm. Decoration on exterior below neck. Decoration (pi. 144 a-v): Technique: Shallow polished incisions 1.0-2.5 mm. wide, drawn with a smooth ended tool that left a sheen like that on the adjacent surface. Parallel lines are evenly spaced and straight on some examples; uneven in width and not regularly parallel on others. Intersections may be overshot. A few examples exhibit red coloring in the incision, a treatment that may have been typical but is fugitive because of application subsequent to firing. Motif: Groups of parallel, straight or curved lines, closely (2-3 mm. apart) or widely (10-13 mm. apart) spaced, may cover the design area (pi. 144 i-p, s-t) or occupy zones separated by plain sections (pi. 144 c, q). On vessels of Form 1, a single incision typically borders the exterior of the lip (pi. 144 b, d). More elaborate patterns occasionally occur (pi. 144 h, u-v). Associated techniques: Vessels of Form 2 may have em- bellished shoulder treatment. Temporal Differences Within the Type : Vessel Form 2 is typical of the early part of the seriated sequence; Form 1 of the latter part. Chronological Position of the Type: Rare but present throughout the duration of the Machalilla Phase (fig. 89). Machalilla Incised and Punctate {Machalilla Inciso y Punteado) Paste and Surface : Similar to fine sand-tempered variety of Machalilla Polished Plain (p. 129); see that type description for details. Form: Rim: None identified; body contour suggests everted rim or a stirrup spout. Body wall thickness: 3-6 mm. Base: Probably rounded or slightly flattened. Principal vessel shape reconstructed from sherds: 1. Jar with angular shoulder, constricted neck and everted rim or possibly stirrup spout. Shoulder diameter 14-20 cm. Decoration (pi. 144 w-z): Technique: Combination of fine incision and punctation. Incised lines range in width from fine scratches to 0.5 mm. and are too shallow to be evident in cross 128 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 I I I I 1 I I 0 4 8 12 CM VESSEL SCALE I i I i I r l o I 2 3 CM RIM SCALE Figure 80.-Rim profiles and reconstructed vessel shapes of Machalilla Incised. section. The bed may be polished or unpolished. Lines are not perfectly straight or evenly parallel, and may vary in width. Punctates are shallow and most frequently circular, but may be elongated or triangu- loid. Diameter is from 0.5 to 2.0 mm., with size being relatively consistent on a single example. Spacing is from adjacent to 3 mm. apart. Motif: Diagonal bands composed of three rows of punctates alternating with incised lines (pi. 144 x-y); variations include two rows of punctation with an intervening incision (pi. 144w), and four rows of puncta- tion separated by four incised lines (pi. 144z). Decora- tion is restricted to the exterior above the shoulder. Associated techniques: The angular shoulder may have embellished shoulder treatment, either nicks or gashes. Temporal Differences Within the Type: None dis- cernible. Chronological Position of the Type : Rare and restricted to the early part of the Machalilla Phase (fig. 89). Machalilla Incised and Red Zoned {Machalilla Inciso y Rojo en %pnas) Paste: Fine sand tempered, see Chorrera Plain (p. 122) for details. Surface : Color: Unslipped areas tan or light to dark gray; small fire clouds typical. Slipped areas dark, rich red. Treatment: Exterior slightly more even and better finished than interior. Interior of bowls usually red slipped, generally giving a striated appearance because of incomplete coverage. Unslipped interiors are striated polished, or on jars unpolished. Exterior polished on red slipped areas, unpolished where not slipped on about 75 percent of examples. On the remainder, slipped and unslipped zones are equally polished. Hardness: 4-4.5. Form: Rim: Direct, expanded or everted, with tapered, flattened or rounded lip. Body wall thickness: 2.5-7.0 mm. Base: Probably rounded or slightly flattened. Principal vessel shapes reconstructed from sherds: Carinated bowl with rounded bottom, angular to rounded shoulder, insloping upper wall convex on the exterior, direct or slightly expanding rim with rounded or flattened lip. Mouth diameter 12-24 cm. (fig. 81-1). Zoned decoration restricted to exterior above shoulder. Rounded jar with incurving upper wall, constricted mouth, direct or expanded rim, rounded or tapered lip. Mouth diameter 12-24 cm. (fig. 81-2). Zoned decoration restricted to upper exterior. Jar with rounded bottom, angular shoulder, con- stricted mouth, everted rim with rounded lip. Rim diameter 22-34 cm. (fig. 81-3). Zoned decoration restricted to area between neck and shoulder, and upper rim interior. Decoration (pi. 145 a-k, m-o, t-u): Technique: Zones containing incision separated by zones of solid red or red banding. Incisions average wider than in Ayangue Incised, often about 1 mm. Width tends to be greatest at the center, tapering off toward both ends. Edges sometimes dragged or thrown up leaving a low ridge. Lines are straight, generally par- allel but not equally spaced, with overshot ends frequent. Motif: Two principal combinations of incision and red zoning occur with about equal frequency. In one, a single or double incision borders the red zone, either at the junction or slightly inside the unslipped area (pi. 145 b, e-g). In the other, several slanting parallel lines run in alternating directions generally leaving a plain area between them (pi. 145 a, c-d, h-i). A rare variant alternates a plain band containing an incised line running parallel to its length and a red band of similar width (pi. 145 k, m-o) either on the everted rim interior of jars or on the exterior body. Occasionally, a con- tinuous incised line alternates with a punctate (pi. 145j) or a dashed one (pi. 145 t-u). Associated techniques: Vessels of Forms 1 and 3 may have nicks or gashes as embellishments of the shoulder, but these appear to be typically discontinuous and placed to coincide with the incised areas, rather than con- tinuous as in Machalilla Embellished Shoulder. WHOLE VOLUME THE MACHALILLA PHASE 129 Temporal Differences Within the Type: The only motif represented in the early part of the sequence is the rare alternating band on the exterior body (table D). It is perhaps significant for reconstructing the evolution of this type that the alternating band treatment frequently incorporates both incised lines and rows of punctate, whereas incision and punctation are not associated in examples from Periods B and C. Table D.-Combined temporal distribution of minor motifs of Machalilla Incised and Red Zoned and Machalilla Punctate and Red %oned Period Seriated sequence Jar rim interior Alter- nating bands C G-l 10, Cut 1, 10-20 cm. G-l 10, Cut 2, 0-20 cm. G-l 10, Cut 1, 20-30 cm. G-l 10, Cut 1, 30-50 cm. G-l 10, Cut 3, 0-20 cm. G-l 10, Cut 1, 50-70 cm. G-l 10, Cut 2, 20-40 cm. 1 1 3 1 2 1 8 B G-112 - - A M-28, Cut 1, 0-20 cm. M-28, Cut 1, 20-40 cm. - 2 Chronological Position of the Type : Machalilla Incised and Red Zoned is characteristic of Periods B and C. It is absent during Period A, except for two sherds representing an untypical variety of the type, in which red bands alternate with incisions (fig. 89; pi. 145k-1). Machalilla Plain {Machalilla Ordinarid) Paste: Method of manufacture: Coiling. Temper: Coarse sand containing both angular and water- worn particles, frequently pale orange or whitish in color making a strong contrast with the paste color. Particles of 2-4 mm. abundant and even small sherds contain one or more 5-8 mm. in diameter (pi. 146). Large particles may lie in the center of the cross section or adjacent to the surface. Texture: Very coarse but compact and not friable; appearance of concrete aggregate. Color: Usually brown, gray-brown or orange through the cross section; sometimes oxidized along one surface, usually exterior, and dark gray along the other. Rarely, medium or dark gray throughout. Firing: Incompletely to completely oxidized. Surface: Color: Brown, gray-brown, medium or dark gray; occasionally orange-brown or grayish tan. Extremely variable over small areas because of fire clouding and unequal polishing. Interior and exterior typically similar shade except when interior is blackened. Treatment (pi. 146) : Smoothed leaving uneven finish with pits, scars, rough areas and protruding temper grains, especially of larger sizes. Coarse crackle radiates from exposed temper particles, which typically lie with a flat facet parallel to the surface. Polishing striations are often visible parallel to the diameter, especially on the exterior. Polishing may produce a low luster in spite of poor smoothing. Hardness: 4-4.5. Form: Rim: Direct or everted with rounded or tapered lip. Body wall thickness: 4.5-8.0 mm. Base: Probably rounded or slightly flattened. Principal vessel shapes reconstructed from sherds: Wide mouthed jar with flattened bottom, rounded shoulder, slightly constricted neck and everted rim with tapered or rounded lip. Rim diameter 18-24 cm. (fig. 82-1). Rounded bowl with slightiy constricted direct rim, rounded lip. Mouth diameter 24-26 cm. (fig. 82-2). Temporal Differences Within the Type : None evident. Chronological Position of the Type : Typical of Period A of the seriated sequence of the Machalilla Phase; rare thereafter (fig. 89). Machalilla Polished Plain {Machalilla Pulido) Paste: Method of manufacture: Coiling; occasional coil junction breaks leaving a slightly convex or concave edge. Temper: Fine to medium white sand, occasional particles to 3 mm. Moderate to abundant quantity, with larger particles more frequent in sherds with abundant tem- per. About 25 percent have fine paste with no visible sand grains. Texture: Compact, fine grained, with occasional pinhole air pockets. Breaks typically straight and clean, at right angle to surface. Color: Variation from orange or light tan through the cross section to solid medium gray, wifli intermediate examples showing oxidation along both surfaces leaving a gray core of even width. Firing: Incomplete to complete oxidation. Surface: Color: Polished surfaces tan, light orange, orange-brown, gray-brown, light to dark gray. Fire clouding may produce extreme variation on a single sherd and often causes a mottled appearance. Reddish-brown tones are sometimes difficult to distinguish from red-slipped surfaces. Unpolished surfaces (jar interiors) may be similar to die exterior or at the opposite end of the color range. Treatment: Exterior of jars and both the exterior and in- terior of bowls polished producing a slick texture and gloss on uneroded examples. Polishing striations are obliterated or unobtrusive. Surface is even to touch, although minor flaws may remain visible. Floating is frequently not sufficient to conceal fine white temper particles. Bowl interiors occasionally striated polished; jar interiors unpolished. Hardness: 4-4.5. 130 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Form: Rim: Exteriorly thickened or everted, with rounded, tapered or flattened lip. Body wall thickness: 2-6 mm., occasionally reaching a maximum of 10 mm. at thickened shoulder of Form 1. Base: Probably rounded or slightly flattened. Principal vessel shapes reconstructed from sherds: Bowl with flattened bottom, angular shoulder slightly thickened to produce a curve on the interior, concave insloping upper wall, slightly everted rim with flattened or tapered lip. Rim diameter 14-24 cm. (fig. 83-1). Bowl with flattened bottom, slightly angular shoul- der, nearly vertical slightly convex upper wall, direct rim with rounded or tapered lip. Rim diameter 14- 22 cm. (fig. 83-2). Carinated bowl with rounded bottom, unthickened to slightly thickened angular shoulder, exteriorly thickened rim with rounded or tapered lip. Rim diameter 14-18 cm. (fig. 83-3). Globular jar with constricted mouth, exteriorly thickened or slightly everted rim, rounded lip. Mouth diameter 12-18 cm. (fig. 83-4). Jar with rounded body, constricted neck and everted rim with rounded or tapered lip. Rim diameter 8- 16 cm. (fig. 83-5). Temporal Differences Within the Type : Tan to orange surface color is most frequent in the early part of the seriated sequence. Chronological Position of the Type : One of the com- mon plain types during the Machalilla Phase, continuing into the succeeding Chorrera Phase (fig. 89). Machalilla Polished Red {Machalilla Rojo Pulidd) Paste: Method of manufacture: Coiling. Temper: Abundant fine to coarse, typically whitish sand. Particles typically less than 0.5 mm., but occasionally reaching 4 mm. in diameter. Larger grains are angu- <\ vs r 111' 111 0 4 8 12 CM VESSEL SCALE ' I 0 I 2 3 CM RIM SCALE Figure 81.-Rim profiles and reconstructed vessel shapes of Machalilla Incised and Red Zoned, and Machalilla Punctate and Red Zoned. whole volume THE MACHALILLA PHASE 131 lar with flat facets. A minority of sherds has fine sand blending into paste color. Texture: Sandy, generally compact except for occasional fine fissures at coil junctions. Temper grains show a tendency to clustering but distribution is not markedly uneven. Color: About 50 percent are light to medium orange through the cross section; the remainder are fired orange to 1 mm. from both surfaces or along the ex- terior only leaving a medium gray core. Typically an oxidized band is sharply defined, and even in width. Firing: Incompletely to completely oxidized. Surface: Color: Slipped surface: Dark rich red to reddish brown, varying because of differences in thickness of coating and firing. Unslipped surface: Light orange, tan, buff; light, medium or dark gray. Treatment: Slipped surface: Slipping ranges from thin but even coating to streaky application incompletely obliterat- ing the underlying surface. Polishing varies from complete but typically leaving clearly defined parallel striations, to superficial and limited to high spots on an unlevel surface. Best polished surfaces are cor- related with most evenly applied slip, but even on these white temper grains may remain visible. Although even surfaces may feel smooth and slick, differential effects of polishing are expressed in variega- tion of coloring. Fine crackle is characteristic of vari- egated surfaces. Rare sherds exhibit vertical rippling reminiscent of Valdivia Pebble Polished in addition to horizontal striations. Slip applied to exterior and neck interior of jars, exterior and interior of bowls. Unslipped surfaces: Even but unpolished, with temper grains typically less visible than on slipped surface. Occasionally striated polished on the interior of bowls slipped only on the exterior. Some jar in- teriors show fine brushlike scraping marks. Hardness: 4-4.5. Form: Rim: Direct or everted with rounded, tapered or flattened lip. Body wall thickness: 3-8 mm., occasionally reaching 10-15 mm. on the body of large jars or thickened portions of neck. Base: Probably rounded or slightly flattened. Principal vessel shapes reconstructed from sherds: Large bowl with flattened bottom, high angular thickened shoulder, tapering inward to slightly upturned rim, rounded lip. Rim diameter 22-30 cm. (fig. 84-1). Red slipped on interior and exterior. Large bowl with flattened bottom, high angular shoulder tapering inward to rounded lip. Rim diameter 22-30 cm. (fig. 84-2). Red slipped on interior and exterior above shoulder. Carinated bowl with rounded bottom, angular unthickened shoulder, convex insloping upper wall, direct rim with rounded lip. Rim diameter 8-22 cm. (fig. 84-3). Red slipped on interior and exterior. Carinated bowl with rounded bottom, angular unthickened to slightly thickened shoulder, insloping concave upper wall, direct rim with rounded or flattened lip. Rim diameter 14-22 cm. (fig. 84-4). Red slipped on interior and exterior, or lip and exterior. Globular jar with slightly constricted mouth, direct rim, rounded lip. Mouth diameter 12-18 cm. (fig. 84-5). Red slipped on interior and exterior. Jar with flattened bottom, rounded to slightly an- gular shoulder, constricted neck and everted rim with rounded lip. Junction of neck and body is typically angular on the interior. Rim diameter 8-16 cm. (fig. 84-6). Red slipped on exterior, continuing over onto interior as far as angular junction of neck with body. Wide mouthed jar with flattened bottom, rounded to slightly angular shoulder, insloping upper wall, 0 I 2 3 CM RIM SCALE VESSEL SCALE Figure 82.-Rim profiles and reconstructed vessel shapes of Machalilla Plain. 132 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 everted rim with rounded or flattened lip. Junction of rim and body wall typically angular on the interior. Rim diameter 22-32 cm. (fig. 84-7). Red slipped on exterior, continuing over onto rim interior as far as angular junction with body. Rare vessel shape reconstructed from sherds: 1. Carinated bowl with rounded bottom, thickened angular shoulder, insloping upper wall, exteriorly thickened rim with rounded lip. Rim diameter 10-22 cm. Red slipped on interior and exterior. Temporal Differences Within the Type: None dis- cernible. Chronological Position of the Type: Present through- out the Machalilla Phase with slightly increasing fre- quency ; continuing to increase in the succeeding Chorrera Phase (fig. 89). Machalilla Punctate {Machalilla Punteado) Surface: On Machalilla Striated Polished 137), see that type description for details. Paste and Plain (p. Form: Rim: Direct with rounded or flattened lip. Body wall thickness: 2.5-7.0 mm. Base: Probably rounded or slightly flattened. Principal vessel shapes reconstructed from sherds: Bowl with rounded bottom, angular unthickened to slightly thickened shoulder, insloping upper wall, direct rim with rounded or flattened lip. Rim di- ameter 14-18 cm. Decoration on exterior above shoulder. Jar with rounded body, constricted neck and everted rim. Interior neck diameter 4-8 cm. Decoration on exterior above shoulder. Decoration (pi. 147 a-i): Technique: Three distinct kinds of punctates are repre- sented : Small, circular impressions 1.0-1.5 mm. in diameter, 1-2 mm. deep, arranged 1-5 mm. apart in horizontal rows (pi. 147 a-d). Rectanguloid marks, 1.0-2.5 mm. long and 1-2 mm. wide, squarish in outline, often slightly shallower at one end, arranged 1-2 or 2-3 mm. apart (wider spacing correlated with larger punctates) in relatively straight rows (pi. 147 g-i). Long, narrow, trianguloid marks filling zones bor- dered by fine incised lines. Punctates 2-3 mm. long, 1 mm. maximum width tapering toward one end, w wi uv> (((® < II I ? I I I ? ' 0 4 8 12 CM 0 1 2 3 CM VESSEL SCALE RIM SCALE Figure 83.--Rim profiles and reconstructed vessel shapes of Machalilla Polished Plain. WHOLE VOLUME THE MACHALILLA PHASE 133 placed 1-3 mm. apart in straight rows. Incisions fine, sharply defined, with dragged edges; long lines tend to be crooked (pi. 147 e-f). Motif: Techniques 1 and 2 are typically not accompanied by incision; punctations occur in double parallel rows or occupy areas of the vessel wall, the extent unde- terminable because of the small size of the sherds. Technique 3 occurs in rectilinear bands or zones bounded by one or two incised lines. Associated techniques: Two bowls have embellished shoulder decoration in the form of gashes. One of these and another sherd have large "bosses" at the shoulder. Temporal Differences Within the Type: Technique 1 is limited to Period A of the seriated sequence; Tech- niques 2 and 3 are limited to Period C (table E). Chronological Position of the Type: Present in minor frequency throughout the Machalilla Phase seriated sequence (fig. 89). V >m y>y H i i ' i ' ' i 0 4 8 12 CM VESSEL SCALE Y ^)W i. i. 111 0 I 2 3 CM Figure 84.-Rim profiles and reconstructed vessel shapes of Machalilla Polished Red. 134 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Table E.- Temporal distribution and frequency of techniques of Machalilla Punctate decoration Period Seriated sequence Tech- nique 7: Small circular Tech- nique 2: Rectan- guloid Tech- nique 3: Trian- guloid Zoned C G-l 10, Cut 1, 10-20 cm. G-l 10, Cut 2, 0-20 cm. G-l 10, Cut 1, 20-30 cm. G-l 10, Cut 1, 30-50 cm. G-l 10, Cut 3, 0-20 cm. G-l 10, Cut 1, 50-70 cm. G-l 10, Cut 2,20-40 cm. - 1 1 1 B G-112 - - - A M-28, Cut 1, 0-20 cm. M-28, Cut 1, 20-40 cm. 2 1 - - Machalilla Punctate and Red Zoned {Machalilla Punteadoy Rojo en Zonas) Paste: Fine sand tempered; see Chorrera Plain (p. 122) for details. Surface: Similar to Machalilla Incised and Red Zoned; see that type description for details. Form: Same as Machalilla Incised and Red Zoned (fig. 81 and p. 128). Decoration (pis. 145 1, p-s, 148): Technique: Alternation of zones of polished red with zones bordered by or filled with punctates. Red usually solid, sometimes pattern of bands 3-8 mm. wide and streaked by polish subsequent to application. Punc- tates are in the form of elongated gashes, 3-6 mm. long and generally 1 mm. wide. Rarely, width may reach 2 mm. Arrangement in straight rows is end to end, with separation from adjacent to 4 mm. apart. In curved rows, horizontal orientation may be retained, with succeeding punctates in slightly staggered position (pi. 148 a, n). Rows tend to be regular when bounding areas but more random when filling zones. Marks are deep, sharply defined and rarely show pushed up edges or ends. Motif: Typically, one or two parallel rows of punctate border a red band or solid red area, falling at the junc- tion or just inside the unslipped area (pi. 148 a, d, f-g, j, n, t). Occasionally the unslipped zone may be almost filled with punctation (pi. 148 c-h, k, 1, s, u). Two rare variants are a single or double row of puncta- tions between two red bands on the rim interior of jars of Form 2 (pi. 148 o-p), and alternating bands of red and punctate on the body of large jars (pi. 145 1, p-s). Associated techniques: Carinated bowls may have embel- lished shoulder treatment. Temporal Differences Within the Type: None discern- ible. Chronological Position of the Type: Like Machalilla Incised and Red Zoned, this type is characteristic of Periods B and C (fig. 89). A single example from Period A represents the rare decorative variant in which a red band alternates with a band containing a dotted line. Machalilla Red Banded {Machalilla Bandas Rojas) Paste: On Chorrera Plain (p. 122) or Cabuya Plain (p. 121); see those type descriptions for details. Surface: Color: Polished surfaces typically tan, buff, orange- brown or gray-brown with a leather-like tone, rather uniform on a single specimen except in areas with dark gray fire clouds. Unpolished surfaces light tan, also with little range of variation in tone. Treatment: Decorated surfaces (exterior of jars, interior and exterior of bowls) even and striated polished, horizontal polishing striations usually remaining clearly visible. Faint vertical rippling reminiscent of Valdivia Pebble Polished sometimes occurs. Floating is not as complete as in Ayangue Incised, leaving small defects, and temper grains protruding. Uneroded surfaces may have a varnish-like sheen. Undecorated surfaces (jar interior) were scraped with a tool that left fine parallel brush-like marks that remain faintly to prominently visible. Bowl interiors and lower exteriors may be covered with a red slip, often streaky in appearance. Hardness: 4-4.5. Form: Rim: Exteriorly thickened, direct, everted or cambered, with rounded, flattened or tapered lip. Body wall thickness: 2-8 mm., increasing to a maximum of 1.4 cm. at the shoulder of vessels of Form 1 and 4. Base: Probably rounded or slightly flattened. Principal vessel shapes reconstructed from sherds: Bowl with flattened bottom, high angular shoulder created by pronounced body wall thickening, slight upturned rim with rounded or tapered lip. Rim diameter 10-26 cm. (fig. 85-1). Wide banded decoration restricted to upper interior and exterior. Carinated bowl with rounded bottom, angular slightly thickened shoulder, insloping upper wall, exteriorly thickened rim with rounded or flattened lip. Rim diameter 12-18 cm. (fig. 85-2). Narrow banded decoration on interior and exterior. Carinated bowl with rounded bottom, angular unthickened shoulder, insloping upper wall, exteriorly thickened or slightly everted rim, rounded lip. Rim diameter 18-22 cm. (fig. 85-3). Wide or narrow banded decoration on interior and upper exterior. Large shallow bowl with rounded bottom, high angular shoulder, tapering rim, rounded lip. Inslop- ing rim creates an approximate 90 degree angle at the shoulder. Rim diameter 14-28 cm. (fig. 85-4). Wide banded decoration on upper exterior. Carinated bowl with rounded bottom, angular un- thickened or slightly thickened shoulder, concave in- sloping upper wall, direct rim with rounded lip. Rim WHOLE VOLUME THE MACHALILLA PHASE 135 0 I 2 3 CM RIM SCALE Figure 85.-Rim profiles and reconstructed vessel shapes of Machalilla Red Banded. 136 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 diameter 18-22 cm. (fig. 85-5). Decoration on upper exterior. Mildly carinated bowl with rounded bottom, slightly angular shoulder, insloping upper wall convex on exterior, direct or slightly expanded rim with rounded, flattened or tapered lip. Rim diameter 10-20 cm. (fig. 85-6). Decoration restricted to upper exterior. Open bowl with flattened bottom, walls curving outward and upward to direct rim with flattened or rounded lip. Rim diameter 12-28 cm. (fig. 85-7). Decoration on exterior and interior. Rounded bowl with rounded bottom, curved shoul- der and slightly constricted direct rim with rounded or tapered lip. Rim diameter 14-24 cm. (fig. 85-8). Decoration restricted to exterior. Globular jar with rounded body, constricted, direct or slightly expanded rim, rounded or tapered lip. Rim diameter 7-14 cm. (fig. 85-9). Decoration on upper exterior. Cambered jar with rounded body, joining outflar- ing neck at a slight constriction frequently angular on the interior, cambered rim produced by wall thick- ening, rounded lip created by constriction and re- expansion of the wall. Rim diameter 12-16 cm. (fig. 85-10). Decoration restricted to exterior. Wide mouthed jar with flattened bottom, rounded shoulder, constricted neck, strongly everted rim with rounded lip. Rim diameter 18-26 cm. (fig. 85-11). Wide banded decoration on upper exterior. Narrow mouthed jar with rounded bottom, rounded shoulder, constricted neck, everted rim with flattened or rounded lip. A variant has a steplike projection about half way between the base of the neck and the shoulder. Rim diameter 8-22 cm. (fig. 85-12). Decoration on exterior and upper rim interior. Decoration (pis. 149-153): Technique: Thick rich dark red slip applied in sufficient thickness to produce a ridge readily visible and easily felt in the majority of cases, measuring up to 1 mm. thick in some cases. Polishing subsequent to applica- tion of red bands typically causes streaking of the pig- ment blurring the edges of the bands when they are not parallel to the direction of polish, which is horizontal. This polishing after decoration leaves the surface beneath the pigment unpolished with the result that sherds in which the paint has flaked off may present a zoned polished appearance. Bands occur in two general classes: 1) Narrow bands, 1-3 mm. wide, rarely reaching 4 mm. and 2) wide bands, 0.5-1.2 cm. wide, rarely reaching 2 cm. About 10 percent of the sherds are intermediate in band width, and occasional designs incorporate wide and narrow bands. Some bowls have wide bands on the interior and narrow bands on the exterior. Either wide or narrow bands may occur on jar exterior, jar rim interior, bowl interi- or or bowl exterior. Execution ranges from lines of approximately even width, parallel and relatively evenly spaced to those of variable width (2-5 mm.), crooked and unevenly spaced. Blobs may occur where a full brush was applied to the surface. Broad bands have markedly irregular margins, caused partly by the painting technique, which involved repeated application of the brush, and partly by sub- sequent polishing. Motif: Narrow bands (pis. 149-151, 153 a-m, o-p) are applied in groups of straight parallel lines often slanting in alternate directions with a blank zone between (pi. 149 a, d-e, i), horizontal parallel lines with an undulating line between (pis. 150 a-b, d-f, h, 151c), Crosshatch (pi. 149k), diamond (pi. 150 k, t), or zoned parallel line arrangements (pis. 149j, 150s). Parallel lines are 2-8 mm. apart. Except for an occasional wavy line, curvilinear elements are rare (pis. 149f, 150 c, m, 151 a-b). Wide bands (pis. 152, 153 n, q) typically run parallel horizontally, 0.5-1.0 cm. apart on the interior of everted jar rims (pi. 152a) or on the neck exterior (pi. 152 f-g, i, m-o), or vertically on the body exterior (pi. 152q). Wide bands are applied to the lip and carination of bowls (pis. 152 b, d, j, 153 n, q). Occasionally bands expand into zones of solid red (pi. 152f). Two sherds show ovoid spots of type produced by finger painting (ph 152p). Associated techniques: Carinated bowls may have embel- lished shoulder treatment. Temporal Differences Within the Type: No clear temporal distinction can be discerned between the fre- quency of wide and narrow banding, the latter being approximately twice as common as the former throughout the duration of the seriated sequence during the Machalilla Phase (fig. 89). Chronological Position of the Type : The most common decorated type throughout the duration of the Machalilla Phase (fig. 89). Machalilla Red Incised {Machalilla Rojo Inciso) Paste: On Machalilla Polished Red (p. 130), see that type description for details. Surface: Color: Unslipped surface: Brown, light tan or orange. Slipped surface: Rich red. Treatment: Unslipped surface (interior of jars) : Even but unpolished. Slipped surface: Smooth and even to touch but streaky in appearance because of polishing striations. Slip of variable thickness but not leaving bare spots. Hardness: 4-4.5. Form: Rim: Exteriorly thickened, direct or slightly expanded with rounded, flattened or tapered lip. Body wall thickness: 4-10 mm. Base: Probably rounded or slightly flattened. WHOLE volume THE MACHALILLA PHASE 137 ? ' 0 I 2 3 CM VESSEL SCALE RIM SCALE Figure 86.-Rim profiles and reconstructed vessel shapes of Machalilla Red Incised. Principal vessel shapes reconstructed from sherds: Carinated bowl with flattened bottom, angular shoulder, slightly thickened to produce curved interior wall, slightly concave upper wall, direct or expanded rim with flattened or rounded lip. Rim diameter 14-20 cm. (fig. 86-1). Decoration on exterior above shoulder. Carinated bowl with flattened bottom, high angular shoulder slightly thickened to produce curved interior wall, insloping upper wall, exteriorly thickened rim with rounded lip. Rim diameter 18-20 cm. (fig. 86-2). Decoration on exterior above shoulder. A few small thick (8-10 mm.) walled sherds may be from jars; no rims could be identified. Decoration (pi. 147 j-r): Technique: Sharply defined incisions cutting through the red slip to reveal the lighter tan to orange paste. Incisions less than 0.5 mm. wide, applied when the surface was dry enough to produce a "scratched" effect with dragged edges. Strokes are crooked, not equally spaced or evenly parallel. Longer lines may be drawn with several strokes, the ends of which do not abut. Overshot junctions are typical. Motif: Rectilinear patterns of straight parallel lines spaced 1-8 mm. apart, with variation of several millimeters typical on a single design. Diamonds produced by crossing or intersection of several parallel lines are most frequent (pi. 147 n-p, r); other motifs include zoned parallel lines (pi. 147 j, k, q), single erratic zigzag lines (pi. 1471), and cross hatch (pi. 147m). Associated techniques: Bowls may have wide red banded decoration on the interior or embellished shoulder treat- ment. Temporal Differences Within the Type: None dis- cernible. Chronological Position of the Type : Present with minor frequency throughout the Machalilla Phase (fig. 89). Machalilla Striated Polished Plain {Machalilla Pulido en Lineas) Paste: Method of manufacture: Coiling. Temper: Fine to coarse sand, typically white. Most char- acteristic is abundant fine white sand less than 0.5 mm. in diameter standing out prominently in dark paste. Large inclusions may reach 5 mm. and occasionally extend through the cross section; usually, however, they occur in thicker walled sherds. A minority of sherds have no visible temper. Texture: Typically sandy, with pinpoint or thin fissure- like air pockets. Not friable or easily broken even when heavily tempered. Temper is not evenly dis- tributed and larger particles tend to cluster adjacent to the surface. Untempered sherds are compact and fine grained. Color: Orange or medium to dark gray through cross section, or orange adjacent to both surfaces leaving a gray core. Rarely, gray along one surface and orange along the other. Core width is less uniform than in Machalilla Polished Plain and may be discontinuous. Firing: Incompletely to completely oxidized. Surface: Color: Tan, light orange, light brown, red-brown, gray- brown, medium to dark gray, with considerable varia- tion on most sherds. Uneven firing conditions and differential absorption by polished and unpolished areas result in a typically streaky or variegated effect. Fre- quent fire clouds. Treatment (pi. 154): Polished leaving well defined par- allel striations running parallel to the rim. Coverage varies from complete, with tracks abutting or over- lapping, to incomplete leaving narrow unpolished in- tervals between adjacent tracks. Polishing typically covers more than 75 percent of the surface. Less com- pletely polished examples tend to be uneven, witii scars and visible temper grains, although even these generally feel smooth. Most surfaces have a low luster. Striated polish is applied to exterior of jars and both surfaces of bowls (if one surface was completely polished, the sherd was classified as Machalilla Polished Plain). Hardness: 4-4.5. Form: Rim: Exteriorly thickened, direct or everted with rounded, tapered or flattened lip. Rims are markedly unlevel and frequently noncircular. Body wall thickness: Typically 2.5-7.0 mm., increasing to 10 mm. at thickened carinations. Based: Probably rounded or slightly flattened. Principal vessel shapes reconstructed from sherds: 1. Carinated bowl with rounded bottom, high angular 8 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 thickened shoulder, short slightly thickened rim with rounded lip. Rim diameter 14-18 cm. (fig. 87-1). Large bowl with flattened bottom, high angular thickened shoulder, tapering rim, rounded lip. Rim diameter 22-30 cm. (fig. 87-2). Carinated bowl with rounded bottom, angular slightly thickened shoulder, slightly concave insloping rim, rounded or tapered lip. Rim diameter 16-30 cm. (fig. 87-3). . Carinated bowl with flattened bottom, angular unthickened shoulder, convex insloping upper wall, exteriorly thickened rim with rounded or tapered lip. Rim diameter 16-36 cm. (fig. 87-4). . Carinated bowl with flattened bottom, angular thickened shoulder, insloping upper wall, slightly everted or exteriorly thickened rim with rounded lip. Rim diameter 16-24 cm. (fig. 87-5). . Globular jar with slightly constricted direct rim, ¯>) W > )) 0 4 8 12 CM VESSEL SCALE C < or < a. - VALDIVIA PHASE TRADE Figure 89.-Seriation of Machalilla Phase sites on the basis of changes in pottery type frequency (Appendix 1, table 16). G-l 14, Cut 1, representing the early Chorrera Phase, is added to show persistence of Machalilla Phase pottery types. G- 110, CUT I, 10-20 CM G-IIO, CUT 2, 0-20 CM G-IIO, CUT I, 20-30 30-50 50-70 G- 110, CUT 2, 20-40 1T-54, SURFACE ~m"-28, CUT I, 0-20 CM 20-40 PERI 00 B PERIOD 10 20 30 40 50% I I I 0-H SAMPLE OVER 50 1=3 SAMPLE UNDER 50 ^m TRADE SAMPLE Figure 90-Temporal distribution of vessel shapes of Machalilla Phase pottery types. Sites and levels are arranged in the seriated order indicated by changes in pottery type frequency. Shapes represent grouping of similar rim profiles in the various pottery types (Table F and Appendix 1, Table 15). Vertical bars connect earliest and latest occurrence of unusual appendages. The sample of Machalilla Phase trade pottery from the Valdivia Phase site of G-54 replaces G-l 12 to represent Period B, since no rim analysis was available for the small sample from G-l 12. WHOLE VOLUME THE MACHALILLA PHASE 141 Valdivia Applique Fillet (pi. 116a). Valdivia Brushed (both varieties) (pi. 116 b-c). Valdivia Broad-line Incised (pi. 116 d-e, h). Valdivia Corrugated (pi. 116 f-g). Valdivia Incised (pi. 116 i, m-n). Valdivia Pebble Polished (pi. 116 j-k). Valdivia Pseudo-^Corrugated (pi. 116o). Valdivia Punctate (pi. 116p). Pottery of Probable Trade Origin Four sherds are ornamented with deep circular punctates in zones bounded by bold incised lines. Punctates are 2-3 mm. in diameter, about 2 mm. deep, with a flat bottom, and spaced randomly 2-3 mm. apart, 3-6 mm. apart, or 6-9 mm. apart on different examples. Incisions are 1-2 mm. wide, about 1 mm. deep, U-shaped in cross section, boldly drawn with a single stroke or series of short strokes. Designs are composed of curvilinear zones outlined with incision and filled with punctations on the ex- terior of open bowls (pi. 158 g-h) or rounded jars (pi. 158i), both of forms unrepresented in other pottery of the Machalilla Phase. Surface is brown to black, with frequent fire clouding. Finish is uneven although polished or striated polished, with striations running vertically on the exterior of the bowls. Body wall thickness is 7-9 mm. Bowl rims are outsloping, direct, tapering slightly toward the rounded lip. Diameter is 10-14 cm. Two bowl sherds from M-28, Cut 1, Level 0-20 cm., one bowl sherd from G-l 10, Cut 2, Level 20-40 cm., one jar body sherd from G-l 10, Surface. The Seriated Ceramic Sequence and Its Implications The seriated sequence of the Machalilla Phase is based on pottery type frequencies from levels of M- 28, Cut 1; G-110, Cuts 1 and 2, and three samples not of stratigraphic origin. G-l 10, Cut 3 is repre- sented only by Level 0-20 cm., because a preliminary classification of this cut in Ecuador drew slightly different boundaries between plain types than were finally adopted, making the existing data unusable in the seriated sequence. The material collected from G-l 10, Trench, although not segregated by artificial levels, has been interdigitated to provide a basis for evaluating the general chronological position of shell, bone and stone artifacts from this provenience. The surface collection from G-l 12, although badly eroded, provides an indication of a transitional period be- tween early and late parts of the sequence. In assembling the seriation chart, the levels of the various cuts were first arranged in stratigraphic order within each cut. Since the cuts at G-l 10 showed similar frequencies, they were interdigitated. Next, the samples from G-l 12 and G-l 10, Trench were placed where they seemed best to fit the trends. When M-28, Cut 1 was added, it clearly belonged at the bottom of the sequence. The resulting graph (fig. 89) shows relatively little change in the plain types, with the exception of Machalilla Plain, which is most abundant at the beginning of the sequence. In decoration, Machalilla Red Banded, both narrow and wide varieties, and Machalilla Embellished Shoulder are the most common types throughout the sequence, although Machalilla Embellished Shoulder declines slightly from a maximum popularity in the earliest levels. Ayangue Incised shows a small but steady growth from early to late. Of the minor deco- rated types, several appear to be absent in the earliest levels, although in some cases this may be the result of small size of the sample. However, the failure of Machalilla Double-line Incised, Ayangue Incised, Machalilla Incised and Red Zoned, and Machalilla Punctate and Red Zoned to be represented (except for a single aberrant sherd of each of the latter three types, to be discussed later) at M-28 is probably indicative of the true situation, since these four decorated types are also absent from a much larger sample from surface collections and other tests. The abrupt beginning of Machalilla Double-line Incised and Ayangue Incised in the earliest seriated level at G-l 10 suggests a gap in the existing sequence. Absence of the former type at G-l 12 might be ex- plained by the small size of the sample, which lacks several other minor types present both earlier and later (such as Machalilla Incised, Machalilla Finger Punched and Machalilla Burnished Line, identified with question marks on fig. 91). On the other hand, the abundance of Machalilla Double-line Incised among sherds of Machalilla Phase origin from the Valdivia Phase site of G-54 suggests that it was a popular decorated type at the time of contact. Since the levels included in the seriation show a declining trend, the only place where greater frequency would fit the chart is prior to the occupation of G-110, or precisely in the region where the existing sample is most inadequate. Since Machalilla Red Incised, which combines single-line incision with a red-slipped surface, is present from the beginning of the Phase, Machalilla Double-line Incised can be postulated as a local innovation principally characterized by the substitution of a double for a single incised line. The origin of Ayangue Incised may reflect outside influence. It is not represented among trade sherds at G-54, suggesting that it was either not present or not important at that time. By the occupation of G-110, it has become one of the principal decorated types. A single sherd of aberrant execution from M-28 could be interpreted as a crude early form, but the absence of intervening examples makes its insig- 142 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 nificance difficult to evaluate. The close similarity of Ayangue Incised technique, motif and vessel shape to incised styles of Mesoamerica (see p. 173) and absence of the type from the early part of the Mach- alilla Phase, seem to favor attributing its appearance to foreign contact. Absence of antecedents in the form either of white slip or black painting argue for a similar explanation for the appearance of Cabuya Black-on-White at about the same time. Whether Chorrera Incised and Cabuya Finger Pressed Rim are local developments or not is arguable but not demonstrable at the present time. Table F.-Correlation between vessel shapes of Machalilla Phase pottery types and generalized forms recognized for the Machalilla Phase Generalized Form Constituent Forms Generalized Form Constituent Forms 1: Carinated Bowl, High Thickened Shoulder Ayangue Incised, Form 1 Machalilla Embellished Shoulder, Form 1 Machalilla Red Banded, Form 1 Machalilla Polished Red, Form 1 Machalilla Striated Polished Plain, Form 1 9: Carinated Bowl, Low Shoulder Ayangue Incised, Form 4 Machalilla Incised, Form 2 Machalilla Red Incised, Form 1 10: Everted Rim Bowl Ayangue Incised, Form 6 Cabuya Finger Pressed Rim, Form 2 Ayangue Incised, Form 2 Machalilla Double-line Incised, Form 1 Machalilla Embellished Shoulder, Form 2 Machalilla Incised, Form 3 Machalilla Red Banded, Form 2 Machalilla Red Incised, Form 2 Machalilla Polished Plain, Form 3 Machalilla Striated Polished Plain, Form 5 2: Carinated Bowl, Thickened Shoulder 11: Narrow-mouthed, Everted Rim Jar Cabuya Plain, Form 5 Chorrera Plain, Form 5 Machalilla Double-line Incised, Form 7 Machalilla Red Banded, Form 12 Machalilla Polished Plain, Form 5 Machalilla Polished Red, Form 6 Machalilla Striated Polished Plain, Form 9 12: Shouldered-neck Jar 3: Carinated Bowl, Unthickened Machalilla Double-line Incised, Form 2 Machalilla Embellished Shoulder, Form 3 Machalilla Red Banded, Form 3 Machalilla Striated Polished Plain, Form 4 Machalilla Embellished Shoulder, Form 8 Machalilla Red Banded, Form 12 Shoulder 13: Wide-mouthed, Everted Rim Jar Ayangue Incised, Form 7 (?) Cabuya Plain, Form 4 Chorrera Plain, Form 4 Machalilla Embellished Shoulder, Form 7 Machalilla Incised and Red Zoned, Form 3 Machalilla Punctate and Red Zoned, Form 3 Machalilla Red Banded, Form 11 Machalilla Plain, Form 1 Machalilla Polished Red, Form 7 Machalilla Striated Polished Plain, Form 8 4: Carinated Bowl, Concave Upper Wall Ayangue Incised, Form 3 Cabuya Plain, Form 2 Machalilla Double-line Incised, Form 3 Machalilla Embellished Shoulder, Form 4 Machalilla Incised and Red Zoned, Form 1 Machalilla Punctate and Red Zoned, Form 1 Machalilla Red Banded, Form 5 Machalilla Polished Plain, Form 1 Machalilla Polished Red, Form 4 Machalilla Striated Polished Plain, Form 3 14: Cambered Rim Jar Ayangue Incised, Form 8 5: Carinated Bowl, Cabuya Plain, Form 1 Chorrera Plain, Form 2 Machalilla Double-line Incised, Form 4 Machalilla Embellished Shoulder, Form 5 Machalilla Red Banded, Form 6 Machalilla Polished Plain, Form 2 Machalilla Polshed Red, Form 3 Machalilla Red Banded, Form 10 Convex Upper Wall 15: Constricted Rim Jar Ayangue Incised, Form 9 (?) Cabuya Plain, Form 3 Chorrera Plain, Form 3 Machalilla Double-line Incised, Form 5 Machalilla Incised and Red Zoned, Form 2 Machalilla Punctate and Red Zoned, 6: Angular Bowl Ayangue Incised, Form 5 Form 2 Machalilla Red Banded, Forms 8 and 9 7: Carinated Bowl, Triangular Rim Machalilla Red Banded, Form 4 Machalilla Polished Red, Form 2 Machalilla Striated Polished Plain, Form 2 Machalilla Polished Red, Form 5 Machalilla Striated Polished Plain, Form 6 Cross Section Unclassified Cabuya Finger Pressed Rim, Forms 1 and 3 8: Open Bowl Direct Rim Cabuya Black-on-White, Form 1 Machalilla Finger Punched, Form 1 Machalilla Red Banded, Form 7 Machalilla Plain, Form 2 Machalilla Striated Polished Plain, Form 7 Chorrera Plain, Form 1 Machalilla Double-line Incised, Form 6 Machalilla Incised, Form 1 Machalilla Polished Plain, Form 4 Machalilla Striated Polished Plain, Form 10 WHOLE VOLUME THE MACHALILLA PHASE 143 Decorated Types Plain Types med T) o N T) V a c :tate ulder Red Zom o N T) -a Plai Pum Sho and J Ml* Pi T) Incis Rim shed a. T3 icised and hed V ded inger Punc ed Incised urnished L icised and C BJ ine X T) u J=3 0 Ph Pla 3 1I0NS unctate mbellis mbellis! a Incised ed Ban unctate ised iouble-1 ised k-on-W er Presi lain triated olished l fl olished IVIS i-i BJ Ph bJ W Bl Pi ro P* M ® k=i n3 crl cj nj Ph to Inc a TO lad bo Ph TO TO TO ST"1 Pla Ph BJ Q Q 3 13 alill a 13 alill 1 alill alill alill ami alill Ayangue alill era TO 1 BJ 3 13 alill 3 Ph 3 * era alill O X! x; X! X X X X XI X XI X X a 1 X X X 5* IH X 2 Mac Mac Mac Mac Mac Mac O P P P BJ TO TO BJ s s s s Mac Mac Choi Cabi Cabi Mac Mac Mac Cabi Choi Mac c ? n El Bl El ? ? El El El E El El El E D E El E E E B D n D (?) a (?) ? (?) D ? (?) (?) ? E E D ? D A E ? El ? D 0 ? ? ? H E e n ? ? Figure 91.-Chronological distribution and period of maximum frequency of Machalilla Phase pottery types. Machalilla Incised and Red Zoned and Machalilla Punctate and Red Zoned seem explicable as local evolution. Both the technique of red banding or zoning and the techniques of incision and punctation occur earlier in the seriated sequence. Rare sherds from M-28 represent the aberrant later variety, in which a single incision or row of punctates or both alternates with a red band. This is a logical ancestor for both types, which are distinguished principally by the fact that one employs punctation and the other incision. Decorative motifs and vessel shapes are essentially the same, and resemble those of Machalilla Red Banded. In order to observe whether any important changes took place in vessel shape, totals of the principal forms represented in the various pottery types were compiled (table F) and the relative frequency per level calculated. When the levels were arranged in the order indicated by the pottery type seriation, a few significant differences became apparent (fig. 90; Appendix 1, table 15), in spite of the small size of the rim sample from many levels. Carinated bowls of Forms 2 and 3 are most popular during Period A, as is jar Form 12. Unfortunately, no rim analysis is available for the sample from Period B. Bowls of Forms 6, 7, 9 and 10 are restricted to Period C, and can be viewed as diagnostic of the late Machalilla Period. Bowls of Forms 5 and 15, although present throughout, are more popular in Period C than in Period A. Stirrup and cylindrical spouted jars are also present during the entire duration of the Machalilla Phase. Calculation of the relative frequency of decorated sherds during Periods A and C shows a slight differ- ence, but one that may not be culturally significant. In Period A levels of M-28, decorated sherds represent 21.6 percent of the total pottery, while in Period C levels at G-l 10 they account for 18.2 percent. The frequency of sherds of Valdivia Phase trade pottery declines from 0.7 percent in Period A to 0.2 percent in Period C. The distribution of the various pottery types on the seriated chart, supplemented by the above inferences, makes it possible to suggest a three-fold subdivision of the Machalilla Phase sequence (fig. 91). Period A, represented at M-28, is characterized by the presence of Machalilla Incised and Punctate, and the maximum popularity of Machalilla Embellished Shoulder. Five other decorated types present through- out the Phase have their inception at this time: Machalilla Punctate, Machalilla Embellished and Red Zoned, Machalilla Incised, Machalilla Red Banded, Machalilla Finger Punched, Machalilla Red Incised, and Machalilla Burnished Line. All plain types are present, the diagnostic one being Machalilla Plain. Period B, represented only by the small surface collection from G-l 12, is transitional and not char- acterized at present by exclusive occurrence of any 144 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 decorated or plain type. Machalilla Incised and Punctate is absent, and two late types (Machalilla Incised and Red Zoned, Machalilla Punctate and Red Zoned) make their appearance in characteristic form. Machalilla Double-line Incised is probably present; the occurrence of Ayangue Incised is less certain. All plain types continue to be made, with little change in relative frequency, to judge by pro- portions at beginning of Period C. Period C is characterized by further proliferation in decorated types. Most popular are Ayangue Incised, and Machalilla Double-line Incised, while Chorrera Incised, Cabuya Finger Pressed Rim and Cabuya Black-on-White occur sporadically. All Period B decorated and plain types continue, Chorrera Plain, Cabuya Plain and Machalilla Polished Red showing slight increases in popularity. Figurines Pottery figurines are rare, but distinctive in ap- pearance. None is complete. Available fragments suggest a high degree of standardization in execution. Machalilla Material: Pottery containing sand temper, grains to 1 mm. diameter. Method of Manufacture: Slab of clay worked into rectanguloid form, flattened at the top to produce a head that is wider in front view but thinner than the body in profile. Breaks at torso and legs do not show clean edges indicative of addition, but kneading may have been sufficiently thorough to obliterate such evidence. There is no longitudinal fracture or other indication of double-coiled construction utilized for manufacture of Valdivia Phase figurines. Eyes and nose are added as applique and may slough off cleanly. Surface Treatment: Superficially smoothed to even but not polished. Vertical smoothing tracks sometimes remain visible. Finish is best on painted examples; others may remain uneven and sandy in texture because of protruding temper grains. Incomplete oxidation in firing leaves the surface light orange to medium gray. Anatomical Features (fig. 92; pi. 158 a-d): Head: Flattening of the upper end produces a head slightly wider than the body in front view but thinner in profile and tapering to a rounded edge. In outline, the head is rounded at the sides and rounded or flattened at the top. Maximum width ranges from 2.5-4.7 cm. Eyes are circular to ovoid pellets, 8-10 mm. in diameter and 3-4 mm. high, appliqued to the surface, but not smoothed over to obliterate the junction. A shallow to deep incision extends from one side to the other approxi- mately through the center of each pellet. Placement of the eyes is high, leaving almost no forehead. The nose is a narrow ridge, broadest at the central base and taper- ing toward ends and summit. Length is 1.3 cm., maxi- mum width 4-5 mm., height 4-5 mm. One fragment suggests a mouth similar in form to the eyes; on two other examples no mouth is indicated. A row of up to five small perforations may run along the sides of the head. One example has narrow red banded decoration on the back of the head, a broader band across the forehead, and three stripes running down from the eye (fig. 92). Torso: The only example shows a torso greatly shortened in proportion to the head. Width is 2.2 cm., thickness (front to back) 1.8 cm. Legs: Two legs represent figurines of larger size than the heads. Both are cylindrical, expanding slightly at the bottom to the flattened foot. Leg diameter is 2.4 by 2.7 cm. Vertical red-banded decoration occupies the sides and rear of one. a Figure 92.-Fragmentary head of a Machalilla type figurine from M-28 (after Estrada, 1958, fig. 34-2). Dimensions: No complete examples have been recovered. Head width ranges from 2.5-4.7 cm. Relative proportions of existing examples suggest a minimum size range of 4-15 cm. Chronological Position of the Figurine Type: Present throughout the Machalilla Phase. Although painted fragments are so far represented only at the earliest site, M-28, the sample is too small to conclude that they are confined to this part of the seriated sequence. Figurines of Trade Origin Two fragments of figurines recovered from G-l 10, Cut 2, Level 20-40 cm. are of indisputable Valdivia Phase origin. One is part of a torso showing the characteristic curve of the side from shoulder to hip and the diagnostic polished red surface associated with the San Pablo Type (pi. 158e). The other is the lower part of a figurine of the Buena Vista Type, with an unpolished gray surface and two legs tapering toward the foot (pi. 158f). Worked Sherds Worked sherds fall into two classes: scrapers and disks, the latter by far the most common. WHOLE VOLUME THE MACHALILLA PHASE 145 Disks Plate 159 a-m Seventy sherd disks were recovered from Machalilla Phase excavations. Nearly half represent Machalilla Striated Polished Plain, another 25 percent Machalilla Polished Red, with the remainder made up of Machalilla Red Banded, Machalilla Polished Plain and Cabuya Plain. Form is generally circular, but only 16 have smooth edges ground perpendicularly to the faces; the remainder vary from irregular to jagged. Size is a constant gradation from a maximum diameter of 7.2 cm. to a minimum of 2.3 cm. Sherds with slight curvature were selected so that most disks are not markedly concavo-convex. Thickness is correlated with pottery type and location on the body wall, and ranges from 5 mm. for the thinnest disks to 8 or 10 mm. for the thickest. Function of the disks is speculative, but it may be significant that the Machalilla Phase marks the first appearance of narrow mouthed jars for which sherd disks were often used as stoppers in later periods on the Ecuadorian coast. Scrapers Plate 159 n-q Five sherds of Machalilla Polished Plain and Machalilla Striated Polished Plain have one straight slightly beveled edge produced by abrasion. The bevel slopes toward the polished exterior surface. Width of the edge is 2.5-6.5 cm. Sherds are rectanguloid in outline but not consistently shaped. DIAGNOSTIC FEATURES AND PERIOD SUBDIVISIONS OF THE MACHALILLA PHASE The only two sites where the Machalilla Phase has been reported in detail are M-28 and G-110. The former occupies a hill near the shore; the latter is on a high cliff overlooking the ocean. Refuse is 40-80 cm. in depth and contains large amounts of shell, fireburnt rocks, and mudstone fragments in addition to sherds, stone and shell artifacts. With the excep- tion of one deer bone and three fragments identified as sea turtle, all the bones were from fish. Wattle and daub construction is indicated by fragments of clay with twig and cane impressions. No post holes, hearths or other features indicative of habitation structures were observed. These characteristics resem- ble those of Valdivia Phase refuse, and many of the inferences about perishable associated elements may be valid for the Machalilla Phase also. Stone tools are unshaped except for a working sur- face or edge, and mostly take the form of cobbles and cobble fragments used as hammerstones, choppers, and blades. A few types, such as "Jaketown perfo- rators", gravers and abraders may have been used for bone and woodworking, and other types, such as reamers and saws, have been equated with shell fish- hook manufacture. None of the tools show pressure flaking to improve the working edge. Sinkers are not represented in the stone artifact inventory, but since they are exceedingly rare in Valdivia Phase refuse in spite of the large amount of digging, this cannot be interpreted as reliable evidence of their absence in the culture. One large grinding stone suggests processing of some kind of seed or grain. The most distinctive shell artifact is a large circular fishhook, 4.0-6.5 cm. in maximum diameter, cut and shaped from pearl oyster. Smaller hooks resembling those of the Valdivia Phase occur, but are less com- mon. Shell was rarely used for personal ornament, the only objects in this category being a single disk- shaped bead and a fragment identified as part of a narrow bracelet. The only bone or tooth object is the curved end of a canine of some carnivore cut and perforated for suspension. This scarcity suggests that personal adornment was principally with perishable materials, an inference that is supported by hunks of hematite and limonite that could have been used for body painting. The outstanding category of material culture represented in the archeological remains is pottery. Both plain and decorated sherds are typically thin walled, but with high tensile strength and not easily broken. Surfaces tend to be smooth and are often polished. Decoration is by a few well defined tech- niques, including incision, punctation, gashing or nicking, applique nubbins, and red banding or zoning. Red banding, produced by application of bands of thick red clay (probably the same material used for red slipping of the entire surface) and subsequently polishing with horizontal strokes often causing "blur- ring" of the margins of the bands, is the most popular method of decoration throughout the duration of the Phase. Also common is embellishment of the shoulder of bowls and jars with a row of nicks, notches or small bumps. Red zones not bounded by incisions alternate with zones of punctation or incised lines on the shoulders of some bowls. Bowl and jar forms cluster into a small number of well defined shapes charac- terized by angularity of the shoulder. Both cylindrical 146 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 and stirrup-spouted jars are characteristic. Jars in general tend to have small orifices, suggesting that the large number of sherd disks may have served as stoppers in accord with the practice in later periods on the Ecuadorian coast. Evidence of ceremonial elements in Machalilla Phase culture is minimal. Although 18-22 percent of the sherds bear decoration, there is no indication that decorated vessels were intended for any but domestic functions. Pottery figurines are rare, crude and stylized. The surface may be plain or decorated with narrow red bands. The lesser frequency of figurines here than in the Valdivia Phase may imply a difference in function, although the presence of several fragments of Valdivia Phase types in Machalilla Phase refuse might be interpreted as evidence that similar beliefs were prevalent. Fragmentary remains of two burials from the refuse of G-l 10 suggest that here as in the Valdivia Phase interment in the village was one method of disposal of the dead. One skeleton had leg bones sufficiently well preserved to show a tightly flexed position. Perhaps the most interesting information provided by these burials is deformation of the skull, a practice not represented in any of the Valdivia Phase skulls (see Appendix 2, p. 229). No grave offerings were associated. All of the Machalilla Phase sites produced sherds of Valdivia Phase pottery types, but although vessel shapes and decoration are quite different, there is no indication of any acculturation in Machalilla Phase pottery as a result of this familiarity with an alien style. Although the duration of the Machalilla Phase is relatively short compared with that of the Valdivia Phase, there is sufficient change in pottery types to permit subdivision into three periods. The inter- mediate Period B is poorly defined, but differences between Period A and Period C are clearcut. Mach- alilla Incised and Punctate is limited to Period A, while Ayangue Incised, Machalilla Double-line In- cised, Machalilla Finger Pressed Rim, Machalilla Zoned Burnished and Cabuya Black-on-White are limited to Period C. Bowl Forms 6, 7, 9, and 10 are also time markers for Period C. The termination of the Machalilla Phase seems to have been accomplished by an influence received from Mesoamerica, bringing new traits that amal- gamated with those of the Machalilla Phase to produce a new configuration, designated as the Chorrera Phase. The continuation of the earlier traditions in potterymaking is clearly evident when early Chorrera Phase levels are added to the top of the Machalilla Phase seriated sequence (fig. 89). A few Machalilla Phase decorative techniques and motifs survive even into the later Regional Develop- mental ceramic complexes, providing a thread of continuity in the diversity of regional styles. Relative and Absolute Dating of the Valdivia and Machalilla Phases Two kinds of dating can be used to estimate the beginning and end of the Valdivia and Machalilla Phase seriated sequences, and to measure the duration of their contemporary existence on the Ecuadorian coast. One is the archeological evidence from strati- graphy and pottery exchanged between the two Phases. The other is a large series of carbon-14 dates from Valdivia Phase levels, representing Periods A-C, and a smaller number from the late Machalilla Phase. These two sources of information agree very well in establishing the relative durations of the two Phases. All the seriated sequences for the Valdivia Phase show sherds of Machalilla Phase pottery types in levels corresponding to Periods C and D (figs. 49-50, 52-53). In addition to plain sherds, including stirrup spout fragments, types identified from early Period C are Machalilla Incised and Machalilla Red Banded (both wide and narrow). During late Period C, the following additional types are represented: Machalilla Incised and Red Zoned, Machalilla Punctate and Red Zoned, Machalilla Embellished Shoulder, Machalilla Doubled-line Incised, and Ayangue Incised. Of these only Machalilla Red Banded and Machalilla Em- bellished Shoulder were identified in Period D refuse at G-84. Ayangue Incised is very rare, and restricted to G-31, where it can be equated with a late Period C or early Period D reoccupation of that site. The chronological position of the above pottery types in the Machalilla Phase sequence (fig. 89) is similar. Machalilla Incised and Machalilla Red Banded are present throughout, as is Machalilla Embellished Shoulder. Machalilla Incised and Red Zoned and Machalilla Punctate and Red Zoned have their inception in Period B and increase in popularity during Period C. Machalilla Double-line Incised and Ayangue Incised are restricted to Period C in the present sample, but their relatively high frequency in the earliest level equated with this Period suggests a beginning during Period B. These correspondences -12 767-841-6 suggest that Machalilla Period A can be correlated with the first half of Valdivia Period C, Machalilla Period B with the second half of Valdivia Period C, and Machalilla Period C with the end of Valdivia Period C and Period D. The absence of Ayangue Incised in the large sample of trade sherds from G-54 is inexplicable if the contact here dates from Machalilla Period C, since it is one of the most attractive and common of the decorated types at that time. On the other hand, abundance of Machalilla Double-line Incised rules out correlation of the occupation at G-54 with Machalilla Period A, where it does not occur. The order of appearance of these pottery types in the Machalilla Phase seriated sequence differs from that in the Valdivia Phase in one principal respect: in the former, Machalilla Double-line Incised and Ayangue Incised appear to begin simultaneously with Period C, while among trade types Machalilla Double- line Incised is earlier and more abundant than Ayangue Incised. In fact, Machalilla Double-line Incised was so popular with residents at G-54 that a larger sample is available from this Valdivia Phase site than from all the Machalilla Phase sites put together. Unfortunately, Machalilla Period B is at present represented only by a small and badly eroded surface collection, in which neither of these decorated types occurs. While a final verdict must await further investigation, the evidence strongly suggests that trade relations at G-54 were with an undiscovered Machalilla Period B site, where Machalilla Double- line Incised was being made, but not Ayangue Incised. Additional support for correlation between Machalilla Period B and late Valdivia Period C comes from analysis of the vessel shapes represented by sherds of Machalilla Phase origin at G-54. When relative frequency is plotted, the result seriates best between the upper level of Period A and the lowest level of Period C, filling the gap left by absence of 147 148 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 information on vessel shape at the single Period B site (fig. 90). Although trends in some of the popular shapes are not conclusive, absence of late Forms 6, 7, 9, and 10 favors this seriated position for the G-54 trade material. When we turn to an examination of the Valdivia Phase pottery types represented in Machalilla Phase refuse, we find that Valdivia Incised, Valdivia Broad- line Incised, Valdivia Brushed and several of the unslipped plain types occur throughout the Machalilla Phase seriated sequence. One sherd of Valdivia Punctate and several corrugated sherds of possible Valdivia Phase origin come from Machalilla Period A levels. Valdivia Applique Fillet and Valdivia Pebble Polished are restricted to levels seriating in the latter part of Period C (fig. 89). Motifs and vessel shapes include several diagnostic of the late Valdivia Phase. Nothing in this combination of pottery types is in contradiction with a Valdivia Period C dating for their origin. Valdivia Brushed, Valdivia Incised and Valdivia Broad-line Incised are the dominant decorated types at this time, while Valdivia Pebble Polished attains its maximum popularity during Period C. Valdivia Applique Fillet is most typical during late Period C and Period D. Plain corruga- tion, while diagnostic of Periods A and B continues during Period C, as does the type of punctation exemplified by the trade sherd. The presence of these latter two types could be used to argue that con- tact between the two Phases began toward the end of Valdivia Period B, and until more sites are repre- sented in the Machalilla Phase seriated sequence this possibility cannot be excluded. Analysis of the trade sherds suggests a correlation between the Valdivia and Machalilla Phases similar to that shown on figure 93. The beginning of the Machalilla Phase seems to equate approximately to the transition between Valdivia Periods B and C, and coexistence continues until the end of both Phases. During Valdivia Period C considerable amounts of pottery were exchanged, but there is no evidence during this time of any acculturation in either ceramic tradition as the result of exposure to strikingly differ- ent vessel shapes and decorative styles. During this period, the two Phases were in close geographical association, facilitating trade relations. With the ad- vent of Period D, the Valdivia Phase population moved toward the south and became centered in the vicinity of the modern town of Posorja, and although the Machalilla Phase also extended its zone of occupa- tion southward, it appears to have reached only the Santa Elena Peninsula. This geographical separation probably partly accounts for the decline in abundance and variety of trade sherds in late levels of both Phases. In the Machalilla Phase, frequency drops from 0.7 percent in Period A to 0.2 percent in Period C. Figure 93.-Period distribution of pottery types of Machalilla Phase origin in Valdivia Phase sites and of Valdivia Phase origin in Machalilla Phase sites. (Distribution of sherds by levels is given on figs. 49-50, 52-53, 89.) WHOLE VOLUME RELATIVE AND ABSOLUTE DATING 149 The seriated sequences of the two Phases and their correlation provide a chronological framework for judging the validity of the large series of carbon-14 dates (figs. 49, 50, 52). When the dates are arranged in the order indicated by seriation of pottery type frequencies (table G), the result is not consistent. Twelve dates represent Period A, five from charcoal samples and seven from shell samples. There is al- most no overlap; four of the charcoal dates being older than all but one of the shell dates. Further- more, the oldest date in each group comes from the uppermost level assigned to the Period. Of the six samples between 4100 and 4190 years, one is from Period C, two are from Period B, and the remaining three represent late, middle and early Period A. Although an increasing antiquity can be recognized if the dates for each Period are averaged (3900 years ago for Period C, 4268 for Period B, 4399 for Period A), such a solution does not take into consideration possibilities for error that may make some dates less reliable than others for determining the span of time represented by different portions of the Valdivia Phase sequence. One means of evaluating the dates is to place them in their actual stratigraphic position with reference to one another. The dated levels represent four excava- tions at G-31 and one at G-54. Three of the G-31 excavations (Cut A, Cut J-D and Cut J-E) are ad- jacent and depths are more or less equivalent, al- though their exact correlation was not recorded during the fieldwork. Cut H was some distance away, and can be aligned with the other three G-31 cuts only on the basis of pottery type seriation. It in turn can be aligned with G-54, Cut 1, by the same means. When these relationships are diagrammatically ex- pressed (fig. 94), apparent inconsistencies in the dates can be appraised. One potential source of error is immediately sug- gested by the sloping contour of the period alignment of levels in the refuse of G-31, Cuts A and J. These are in the deepest part of the site, at the edge of the natural knoll that underlies the rest of the area of occupation. As the refuse rolled down the slope, it built gradually outward, an inference indicated by the pronounced disconformity between Sections D and E at the end of Period A, which decreases slightly Table G.-Carbon-14 dates for complexes of the Early Formative Period. Provenience Material Date Phase Subperiod {Levels in Seriated Sequence) {B.P.) G-110 Cut 2 Level 0-20 cm. Charcoal 3320ñ170(SI-107) Machalilla C Level 20-40 cm. Shell 2830 ñ45 (SI-67) Machalilla C Level 20-40 cm. Charcoal 2980ñ160 (SI-108) Machalilla c G-54 Cut 1 Level 60-80 cm. Shell 3450 ñ50 (SI-69) Valdivia c Level 100-130 cm. Shell 4040ñ55 (SI-71) Valdivia c G-31 Cut H Level 20-40 cm. Shell 4140ñ60(SI-80) Valdivia C(B)* Level 80-100 cm. Shell 3970 ñ65 (SI-78) Valdivia C Level 120-130 cm. Shell 4050ñ200 (W-630) 7 Valdivia B Level 120-130 cm. Shell 4170ñ65 (SI-85) J Valdivia B Level 130-140 cm. Shell 4390ñ60 (SI-84) 4540 ñ150 (re-run SI-84) Valdivia B(A) G-31 Cut A Level 120-140 cm. Shell 4190 ñ200 (W-632) Valdivia B Level 160-180 cm. Shell 4530ñ55 (SI-83) - Valdivia A Level 180-200 cm. Shell 4120ñ65(SI-82) Valdivia A(B) G-31 Cut J Section E Level 360-390 cm. Charcoal 5150ñ150(M-1320) Valdivia A G-31 Cut A Level 200-220 cm. Shell 4270ñ60(SI-81) Valdivia A(B) G-31 Cut J Section E Level 390^20 cm. Shell 4220ñ100(SI-16) Valdivia A(B) Level 390-420 cm. Shell 4100 ñ140 (M-1321) Valdivia A(B) Level 390-420 cm. Shell 4230ñ100 (SI-18) Valdivia A(B) Hearth, 400 cm. Charcoal 4620ñ140(M-1322) Valdivia A G-31 Cut A Level 400-420 cm. Shell 4450 ñ200 (W-631) Valdivia A G-31 Cut Jf Section D Level 270-300 cm. Charcoal 4480ñ140(M-1317) Valdivia A Level 300-330 cm. Charcoal 4170ñ140(M-1318) Valdivia A(B) Level 300-330 cm. Charcoal 4450ñ90(SI-22) - Valdivia A * Identifications in parentheses indicate the probable subperiod represented by the date, reflecting mixture of the shell with sherd refuse of earlier or later Period affiliation. 150 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 by the end of Period B. Section E contains fewer levels corresponding to Period A than the adjacent excavations, and more levels corresponding to Period B than Section D, based on the pottery type seriation (fig. 50). When the carbon-14 dates are equated with the level from which the sample was obtained, there is a tendency toward clustering that permits rejection of some as less reliable than others. For example, although there are three determinations for shell samples from Cut J, Section E, Level 390-420 cm., and agreement between them is excellent, all three dates are far too recent for their stratigraphic context. Instead, they are within the range of dates derived from Period B and C levels of other excava- tions. Since these dates are from shell samples, it can be postulated that shells rolled down or were dumped down the slope during Period B to become mixed with the shallow Period A refuse in the area of Section E. The earlier date from this level as well as the one from the next level above, both derived from charcoal, can be accepted as valid for early Period A. Another unacceptably recent date comes from Cut J, Section D, Level 300-330 cm. Two charcoal samples from this level were processed by different laboratories, and the resulting dates disagree by 280 years. The older one (4450 ñ90) is close to a date from the next level above (4480 ñ140), and both are in the range expected from their stratigraphic position in relation to acceptable dates from the bottom of Cut J, Section E. Consequently, the younger date can be rejected as not referring to Period A. The remaining discrepancy in the Period A dates is the unusually early determination (4530 ñ55) for a shell sample from Cut A, Level 160-180 cm., which cor- responds ceramically to the end of Period A. It is within the range for the Period, however, and con- stitutes a problem only because it is stratigraphically above two levels with more recent dates (4120 ñ65, 4270 ñ60). This reversal can be explained by sup- posing that mechanical mixture caused concentration of earlier shells in the level. The two recent dates seem best equated with the beginning of Period B. An interesting situation presents itself in the dates from Cut H, Level 120-140 cm. Analysis by two laboratories of a shell sample from the upper half of this level gave closely similar results: 4050ñ200 and 4170 ñ65. A sample from the lower half of the level produced a date of 4390 ñ60, increased on rerun to 4540 ñ150. The earlier pair of dates falls within the range accepted for Period A and can be explained by the hypothesis that during this occupation shells were scattered up the hillside where they eventually became incorporated into later refuse. The more recent pair is in close agreement with a Period B date from Cut A, Level 120-140 cm. as well as with dates from levels higher up in Cut H, and consequently may be accepted as referring to the Period B occupation. Three of the four dates for levels representing Period C are more recent than any other of the Valdivia Phase dates, and differ by a magnitude that seems reasonable in terms of the seriated sequence. The remaining date (4140 ñ60) is only slightly older but its stratigraphic occurrence above a date of 3970 ñ65 for the lower part of Cut H, along with its similarity to the more recent pair of dates from Level 120-140 cm. of the same excavation suggests it refers to the beginning rather than the middle of Period C. The two dates from G-54 are in proper stratigraphic order, and the more recent one has been accepted as approxi- mating the end of Period C. Adjustment of the discrepancies and projection of the acceptable dates onto a time scale suggests the Valdivia Phase periods are approximately as follows: Period D 3400-3000(?) years ago Period C 4000-3400 years ago Period B 4300-4000 years ago Period A 5000-4300 years ago The only three dates representing the Machalilla Phase are all from the Machalilla Period C site of G-l 10, one from Level 0,-20 cm. and two from Level 20-40 cm. of Cut 2 (table G). Ceramically, these levels seriate in the upper and lower portions of Period C, but pottery type frequency curves suggest little difference in time. The three dates encompass about 500 years, however. The two charcoal samples, representing a charred incrustation on the interior of sherds probably belonging to the same vessel, gave determinations of 3320 ñ170 years (SI-1'07) and 2980ñ160 years (SI-108). Austin Long, who produced the dates, indicates that dis- crepancies of this magnitude can be expected about 30 percent of the time in multiple dating of a single sample and recommends a statistical averaging result- ing in 3150 ñ160 as the most reliable estimate. The shell sample is close to the charcoal date for the same level, being 2830 ñ45 years (SI-67). The shell species, Anomalocardia subrugosa, is the one utilized for the Valdivia Phase shell samples. In terms of the Valdivia Phase sequence, the earliest Machalilla Phase date appears to be generally correct. Assuming that the 3450 ñ 50 date corresponds approximately to the end of Valdivia Phase Pe- riod C, the date of 3320 ñ170 for the inception of occupation at G-l 10 coincides with the ceramic evidence represented by the pattern of trade sherds. The Valdivia Phase site of G-54, from which the WHOLE VOLUME RELATIVE AND ABSOLUTE DATING 151 above date was obtained, produced no sherds of Ayangue Incised among the trade material of Machalilla Phase origin. Since this decorated type is common at G-l 10, it was postulated (pp. 147-148) that occupation here postdated abandonment of G-54, an inference that is supported by all of the carbon-14 dates. The time span suggested by the remaining Machalilla Phase carbon-14 dates at G-l 10 cannot be taken at face value in view of the contradictory nature of the results for samples from the same level and the lack of agreement between antiquity and relative stratigraphic position. The results of the two charcoal samples are reversed, the oldest date coming from the upper level and the youngest from the lower level. The shell date, associated with the lower charcoal sample, is 150 years more recent than the latter. The archeological context is thus an ambiguous G-54 CUT I G-31 CUT J SECTION E CUT J SECTION D CUT A CUT H < 0 20 40 60 80 , o o (4140+60) o o 0 20 40 - GUANGAL, //// 4 PHASE RE CUSE /// 3970+65 ' 100 j 120 140 E > L 60 / 4050+200 \ 4170 + 65 OJ / / y CD 80 (4390+60) V4540+I50)/ 000 100 120 140 300-4 4190*200 v 160 r 4530+55 t I > ^ 200 (4120+65) (4270i60) 220 240 260 B.P o 280 4480tl40 o -43 320 (4170+140) 4450190 o o 340 360 380 5I50ñI50 400 420 (4100+140) (4220il00) (4230ñI00) \ 4620il40 / 4450+200 f 3450+50 4040+60 20 40 60 80 100 130 Figure 94.-Stratigraphic origin of carbon-14 samples from G-31 and G-54. Valdivia Phase period divisions derived from seriation of changes in pottery type frequency indicate a sloping surface during the refuse accumulation. Dates in parentheses are interpreted as not referring to the cultural period in which they stratigraphically occur. (See text pp. 149-150 for detailed explanation.). 152 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 one and evaluation of the dates depends on other evidence. Rejection of the two most recent dates seems justified because of their close approximation to a date of 2800 ñ115 for the late Chorrera Phase from Veliz in central Manabi (table H). This coincidence might be rationalized by viewing the Machalilla Phase occupation at G-l 10 as an isolated local survival were it not for the presence of early Chorrera Phase sites in the immediate vicinity. G-l 14, representing a transition from the Machalilla to the Chorrera Phase (fig. 89), and G-48, representing the fully developed Chorrera Phase, are both within a few kilometers of G-l 10. A few sherds of Chorrera- like types occur at G-110, indicating that G-110 was one of the sites occupied by the Machalilla Phase at the time of the contact which produced the Chorrera Phase; and that a short interval of intercommunication preceded abandonment of the site and integration of the two populations, at sites like G-114, leading to the emergence of the Chorrera Phase along the southern Guayas coast (fig. 89). If so, either the compromise date of 3150 ñ160 derived from averaging the two runs on the charred remains on potsherds (see p. 150), or the earlier 3320 ñ170 more accurately date the site. Although the series of carbon-14 dates for the Ecuadorian coastal complexes is in general agreement with the archeological sequence as it is now known, a problem arises when the attempt is made to correlate this set of dates with a set derived from obsidian. Since the use of obsidian is unknown during the Val- divia and Machalilla Phases, its appearance in refuse deposits is perhaps the most reliable single criterion for identifying the beginning of the Chorrera Phase. Extended research by Friedman and Smith has resulted in a method of using this material for age determina- tion, producing a long series of dates for Ecuadorian coastal complexes from the beginning of the Chorrera Phase up until European contact (Friedman, Smith, Evans and Meggers, 1960, table 1, pp. 500-508). Sub- sequent carbon-14 determinations for sites of the Bahia, Chirije, and Manteno Phases, representing the Regional Developmental and Integration Periods coincide very well with the obsidian dates (Estrada, 1962, Cuadro4). Carrying obsidian dating backward into the late Formative Period leads to estimates in the range of 3400-3600 years for the elapsed time since the in- ception of the Chorrera Phase (Friedman, Smith, Evans and Meggers, 1960, fig. 19). Samples on which this estimate is based come from three sites, one on the Santa Elena Peninsula (G-76: Libertad), one on the Rio Daule (G-D-8: Naupe) and one on the Rio Babahoyo (R-B-l: Chorrera). These dates, which are completely consistent with the rest of the obsidian series, overlap not only with the postulated termination of the Machalilla Phase, but extend back into Valdivia Period C, a situation that is in con- tradiction with stratigraphic and distributional ar- cheological evidence. Obsidian dates result from a natural process com- pletely different from that drawn upon by carbon-14 dating. Obsidian reacts with water resulting in a change of properties. When a new surface is exposed by fracture, hydration occurs, producing a more or less sharply defined layer easily observed microscopi- cally in cross section. Hydration proceeds at different rates, correlated principally with differences in mean annual temperature, so that when the appropriate temperature category has been identified, hydration rim thickness measurements can be converted into elapsed time producing a series of dates superficially comparable but very different in origin from carbon-14 determinations. A potential source of error in obsidian dating is the hydration rate selected for converting thickness measurements into elapsed time. Rates have been selected by establishing the hydration rim thickness of several specimens in contexts of different ages, dated by carbon-14 or other "absolute" methods. When several thickness-age correlations were available for a geographical area, these were connected by a straight line that represented the rate in microns squared per thousand years (Friedman, Smith, Evans and Meggers, 1960, figs. 12-17). Rates established in this way are unfortunately only as valid as the carbon-14 dates on which they are based. In the case of coastal Ecuador the rate of 11 microns squared per 1000 years was derived from "radiocarbon dating of several of the later cultures and an estimated date of 3400 years B.P. . . . for the beginning of the Chorrera Period based on the radiocarbon dating of 4000-4500 years B.P. of certain levels of the preceding occupation, the Valdivia culture" (op. cit., p. 488). The first ques- tion that must be answered in attempting to resolve the overlap between the two sequences of carbon-14 and the obsidian dates is consequently whether there is an error in the obsidian hydration rate selected for the Ecuadorian coast. At the time the coastal Ecuadorian hydration rate (Scale B) was adopted, only three carbon-14 dates were available for the corresponding time periods. Two of these represented the Bahia Phase of the Re- gional Developmental Period and were of similar magnitude: 2150ñ240 (W-833) and 2200ñ240 (W- 834). The third date, for a site of the early Manteno Phase at the beginning of the Integration Period, was 760 ñ500 (W-835). All three were derived from charcoal (Rubin and Alexander, 1960, pp. 180-181). Thirteen additional dates have subsequently been WHOLE VOLUME RELATIVE AND ABSOLUTE DATING Table H.-Carbon-14 dates for complexes of the Late Formative, Regional Developmental, and Integration Periods. 153 Provenience Date B.P. Phase Period M-6: Cerro de Hojas, Corral 560ñ200(M-736) Manteno Integration G-22: Puerto Chanduy, Cut B, Level 50-60 cm. 760 ñ500 (W-835) Chirije-Manteiio Integration M-42: Chirije, Corral A, Level 120 cm. 850ñ105 (1305) Chirije Integration M-55: Sequita (Pepa de Huso) Level 280-300 cm. 1100ñ105 (SI-42) Chirije Integration Level 260-280 cm. 2525 ñ 105 (SI-35) Bahia-Chorrera Late Formative Level 300-320 cm. 2540ñ125(SI-43) Bahia Chorrera Late Formative M-7: Esteros, No. 1 Cut 1, Level 40-80 cm. 2200ñ240 (W-834) Bahia I Regional Developmental Level 280-320 cm. 2150ñ240 (W-833) Bahia I Regional Developmental Cut A, Level 320-340 cm. 2050ñ120(M-1315) Bahia I Regional Developmental Level 320-340 cm. 2300ñ65 (SI-49) Bahia I Regional Developmental Level 340-360 cm. 2350ñ65 (SI-52) Bahia I Regional Developmental Level 380-400 cm. 2430ñ60 (SI-55) Bahia I Regional Developmental Level 400-420 cm. 2110ñ120 (M-1319) Bahia I Regional Developmental Level 400^120 cm. 2120ñ120 (M-1316) Bahia I Regional Developmental M-8: Tarqui, Level 250 cm. 2170 ñ200 (M-734) Bahia Regional Developmental M-42: Veliz, Cut B, Level 40-60 cm. 2800ñ115 (1307) Chorrera Late Formative obtained from Ecuadorian cultures using obsidian. These represent five new sites and seven more dates for the Bahia Phase site of M-7, which provided two of the original samples (table H). The results confirm the first three dates and generally substantiate the estimated time scale for the Ecuadorian coast. Four of the dates represent the Integration Period, subdivided by Estrada (1962) into an earlier Chirije and a later Manteno Phase. Three are from single period sites: M-6, M-42, and G-22. The fourth is from M-55, which has an earlier occupation cor- responding to the transition between late Chorrera and early Bahia and a later reoccupation by the Chirije Phase (Estrada, 1962, fig. 110). Since there is a large discrepancy between one of the three deter- minations for the site (SI-42) and the other two (SI-35, 43), and since the recent date falls within the expected range for the Chirije Phase it seems safe to attribute it to that portion of the sequence. These four dates give a span of 560-1100 years ago for the Integration Period on the coast. Of the nine dates available for the Bahia Phase, eight are from M-7 and one from M-8 immediately adjacent and probably part of the same large site (Estrada, 1962, fig. 5). The levels from which the samples were obtained at M-7 are correlated by Estrada with Bahia I or the first half of the Bahia Phase (op. cit., Cuadro 3, figs. 107-108). Although there is some discrepancy between the results from different laboratories, the range of 2050-2430 years ago places Bahia I in the first half of the estimated duration for the Regional Developmental Period. Two dates from M-55, from levels corresponding to the transition between late Chorrera and early Bahia, agree culturally and chronologically with the M-7 dates and place the beginning of the Bahia Phase at around 2525-2540 years ago. The remain- ing date, from M-42 and representing levels cor- responding to the late Chorrera Phase, is sufficientiy older at 2800 years ago to constitute an acceptable date for the site. One additional date, from La Tolita in Esmeraldas Province, can be equated with the Guayas-Manabi cultural sequence only tentatively because details of the local chronology have not yet been worked out. The sample was obtained from a chimney burial (Stirling and Stirling, 1963, p. 5) of the type characteristic of the Milagro Phase, which occupied the Guayas Basin during the Integration Period. Associated, however, was a figurine of La Plata type, suggesting contact with the Bahia Phase of the Manabi coast. The date of 1690 ñ200 years ago (M-735) is within the expected range for the Regional Developmental Period but slightly earlier than the estimated beginning of the Integration Period. This series of carbon-14 dates makes it possible to evaluate whether the overlap between obsidian dates for the early Chorrera Phase and carbon-14 dates for the latter part of the Valdivia and Machalilla Phases is the result of selection of a wrong hydration rate. For this purpose the group of dates spanning the transition from late Chorrera to early^Bahia is 154 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 particularly important, since it establishes conclusively that the Regional Developmental Period began 2400-2500 years ago, or about 500 B.C. Another rel- atively fixed point is the upper end of the sequence, corresponding to initiation of European contact. Archeological evidence suggests that there was a rather rapid disappearance of the aboriginal cultural pattern after A.D. 1550, so that prolongation of the Integration Period much beyond this date is dubious. The earliest acceptable beginning date remains about 3450 years ago or 1500 B.C., the end of Valdivia Period C. Scale B, the originally adopted hydration rate for coastal Ecuador, is 11 microns squared per 1000 years (Friedman, Smith, Evans and Meggers, 1960, fig. 17). By this rate 2.1 microns equate with 400 years ago (A.D. 1550), 3.9-4.0 microns with 1450 years ago (A.D. 500), 5.1-5.2 microns with 2450 years ago (500 B.C.), and 6.1-6.2 microns with 3450 years ago (1500 B.C.). This scale produces good agreement with ob- sidian samples from late Chorrera and Guangala, Tejar and Daule Phases (and subsequentiy measured samples from Bahia and Jama-Coaque Phases of north Manabi). Hydration rims on the late Chorrera Phase are typically thicker than 5.1 microns and those of the Daule Phase are typically between 4 and 5 microns. The range covered by samples from the early Guangala Phase site of G-34 and the late Guangala Phase site of G-39 is shown on figure 95. (Distributions for other Regional Developmental Phases are diagrammed in the original article; op. cit., fig. 19). Scale B presents two principal diffi- culties : 1) many of the obsidian chips from Integration Period sites (illustrated by G-D-2 on fig. 95) have rims thicker than 3.9 microns, making it necessary to attribute them to re-use of material recovered from earlier sites rather than primary manufacture; and 2) a number of samples with rim thicknesses between SCALE B II MICRONS2/ 1000 YEARS SCALE B-l 12.3 MICRONS2/ 1000 YEARS SCALE B-2 14.4 MICRONS2/ 1000 YEARS TIME SCALE 2 - Ln- 1_ v> 4 - Figure 95.-Differences in agreement between period identification of selected sites and obsidian dates derived by three different scales for conversion of hydration layer thickness into elapsed time. The black bars indicate the thickness range encompassed by each scale for each time period. The white bars indicate the thickness range at sites representing the following successive cultural periods: Late Formative (Chorrera Phase): G-76, G-D-8; Regional Developmental (Guangala Phase): G-34, G-39; Integration (Milagro Phase)- G-D-2, G-M-ll. WHOLE VOLUME RELATIVE AND ABSOLUTE DATING 155 6.2 and 7.0 microns (shown by dotted extension on fig. 95) had to be left as unexplainably older than expected. A third difficulty now arises in the form of an overlap with the upper end of the carbon-14 dated Valdivia and Machalilla Phase sequences. In order to determine whether Scale B was too slow, a somewhat faster rate (Scale B-l) was empiri- cally adopted, in which 12.3 microns squared corre- sponds to 1000 years. By this rate 2.2-2.3 microns equate with 400 years ago, 4.2 microns with 1450 years ago, 5.5 microns with 2450 years ago and 6.6 microns with 3450 years ago. Comparison of this rate with the range shown by obsidian samples in the same sites used to evaluate Scale B shows the two Chorrera Phase series (G-76 and G-D-8) to begin within the estimated duration of the Late Formative Period, but the late Chorrera Phase occupation at G-D-8 now projects into the Regional Developmental Period (fig. 95). Specimens with thickness between 6.6 and 7 still are unexplained. The early and late portions of the Guangala Phase sequence (G-34 and G-39) remain within the confines of the Regional Developmental Period but have been pushed into the upper part. Increase of the hydration layer thickness corresponding to the Integration Period allows a greater proportion of the samples to be accepted as of contemporary origin, but many must still be re- jected as too early. In short, Scale B-l does not eliminate all the problems associated with Scale B and creates new maladjustments at the point where the Formative Period ends and the Regional Develop- mental Period begins. A third scale (Scale B-2) was constructed to equate 7 microns with 3400 years ago, in order to permit acceptance of specimens with hydration layers between 6.2 and 7 microns as contemporary with the beginning of the Chorrera Phase. By this rate, 2.4 microns equates with 400 years ago, 4.6 microns with 1450 years ago, 5.9-6 microns with 2450 years ago and 7 microns with 3400 years ago. While this scale accomplishes its purpose of placing all thick hydration layers within the Chorrera Phase time range, distor- tions in the remainder of the sequence are so gross as to create total disalignment with the period divisions, and consequently with the carbon-14 dates for equivalent sites and complexes (fig. 95). In short, the attempt to bring the series of obsidian dates into alignment with the Valdivia and Machalilla Phase carbon-14 dates by increasing the rate used for equating hydration rim thickness with elapsed time destroys the excellent agreement between the obsidian dates and the more recent series of carbon-14 dates. While agreement exists in part because three of these dates were used as reference points for selecting Scale B, the validity of the time scale for the period divisions and of Scale B for conversion of hydration rim thickness into elapsed time seem confirmed by the thirteen carbon-14 dates subsequentiy received. About 1000 years farther back in time, however, the two systems of dating appear to have diverged by at least 300 years. Examination of the literature brings to light the interesting evidence that carbon-14 dates tend to deviate increasingly from calendrical scales as they go backward in time. In one experiment, a group of wood samples of known age was dated by carbon-14. Agreement was within the probable error over the past 2000 years, but samples older than 2500 years showed increasing divergence, with the carbon-14 result being more recent than the known date. Damon, Long and Sigalove (1963, pp. 284-5), who conducted the experiment, interpret the deviation as follows: The cause of the large apparent excess of C-14 beyond 2500 B.P. does not appear to be the result of an underestimation of the half life of C-14 as indicated by the fairly close agreement for younger samples. Two remaining possible causes are (1) variation in the C-14 production rate and (2) fluctuation of the total CO 2 content of the atmospheric reservoir. Although the first possibility cannot be eliminated at this time, it would seem to the authors that fluctuation of the total content of the atmospheric reservoir is not only possible but quite probable. Much more work is necessary before this phenomenon can be properly evaluated. However, such large fluctuations of the initial C-14 content of wood call for caution in the evaluation of C-14 dates as absolute sidereal years. In particular, the fixing by C-14 dating, of floating archaeologic chronologies . . .can- not be successfully accomplished without reference to the C-14 content of samples of known age. A similar kind of discrepancy has been noted between tree ring dates from Sequoia gigantea and carbon-14 dates on the same wood. As Libby (1963, p. 279) reports: It is clear that the carbon-14 dates are in good agreement with the tree-ring chronology curves for the period 800 to 2400 years ago. Between 3000 and 4000 years ago, however, the ages as determined from the number of three rings are higher than the ages shown by radiocarbon dating. In the Old World, a correlation problem has emerged when historically derived estimates are compared with carbon-14 dates for the early Dynastic Period of Egypt, although the threshold of disagree- ment is about 1000 years earlier than in the New World. The Egyptian calendrical system extends back to 2000 B.C., and for this period of some 4000 years agreement between historical and carbon-14 dates is very good (Smith, 1964, p. 33). Between 4000 and 5000 years ago, however, carbon-14 dates are consistently more recent than historical dates, which for this period are based on various kinds of documentary evidence rather than on the calendar. Although there is potential ground for error in 156 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME l compiling the historical evidence, arguments in favor of the longer estimates are impressive (ibid.). The differences increase from zero at 4000 years ago to 500 years at 5000 years ago (Libby, 1963, p. 278). The consistent pattern of sudden deviation between carbon-14 dates and dates arrived at by tree-ring and historical methods in other parts of the world, be- ginning 2500-3000 years ago in the western hemi- sphere, suggests that the disconformity between the earliest Chorrera Phase obsidian dates and the most recent Valdivia Phase carbon-14 dates may be a product of factors operating on the dating systems rather than of misinterpretation of the archeological evidence. If so, reconciliation of the two series of dates could be accomplished by pushing back of the carbon-14 dates some 300 years at 3400 years ago, so that the most recent Valdivia Phase date of 3450 ñ 50 (SI-69) would be corrected to 3750 ñ50, bringing it into better chronological alignment with the beginning of the Chorrera Phase, obsidian-dated at around 3400- 3600 years ago. If the discrepancy is a cumulative one, as suggested by the experimental data quoted above, the earliest Valdivia Phase dates might have to be increased by more than 300 years. Another cumulative error affecting carbon-14 and not obsidian is the halflife used for converting meas- urements into elapsed time. The Ecuadorian dates have been computed according to a halflife of 5568 years adopted by the various laboratories for stand- ardization of results. Recent experiments have shown, however, that 5730 is a more accurate halflife. In the most recent Valdivia Period C date (3450 ñ50), this makes a difference of 100 years, and alone eliminates one-third of the overlap with the beginning of the obsidian series. Reconciliation of the disconformity by accepting the obsidian sequence dating of 3400 years ago for the approximate inception of the Chorrera Phase, and pushing back the most recent Validvia Phase carbon- 14 date of 3450 ñ50 by about 300 years creates a temporal alignment in keeping with the archeological situation. If the inference that the inland spread of the Chorrera Phase reflects substitution of intensive agriculture for gathering (perhaps supplemented by incipient cultivation) is correct, this marks the be- ginning of a process that became crystallized at around 2500 years ago in the "florescent" or "classic" Regional Developmental complexes. Evidence is ac- cumulating in other parts of the world that realization of the cultural potential of agriculture is a slow proc- ess. Acceptance of a date appreciably less than 3400 years ago for the initiation of this process on the coast of Ecuador would require the assumption that cultural change was far more rapid than during comparable portions of Mesoamerican or Middle Eastern develop- mental sequences. Even 900 years is a short time compared to that intervening between the beginning of productive agriculture and the beginning of ur- banism in other areas. However, special circum- stances may have sped up the process, so that these speculations are subject to testing by carbon-14 dates from early sites of the Chorrera Phase. Origin and Affiliations of the Valdivia and Machalilla Phases The most interesting and at the same time most difficult task of the archeologist is to integrate his findings into the web of local sequences and horizon styles fashioned by previous investigators. Contra- dictions in interpretation must be resolved and non- conforming data must be evaluated and accepted or rejected, often on arbitrary grounds. The task may become easier when archeological sequences are better described and more exactly dated, but since it is the inevitable flaw of archeological evidence that it will always be subject to random preservation of only a minor proportion of what was once made and used, increased fieldwork cannot be relied upon to resolve all contradictions between data and hypothesis. Another consideration inherent in the process of trying to generalize from a group of unique historical events is the inability of a generalization, which expresses patterning, to account for each small detail. The unrealistic nature of such an expectation has beenVell stated by a biologist (Grant, 1963, p. 5): A generalization may be likened to a map or ground plan drawn from the original landscape. A certain pattern of land and sea, or hills and valleys, is present in the unmapped area, waiting to be discovered. The earliest maps, like the earliest generalizations in science, are crude and inaccurate, but they do represent a first step even though they are destined to be replaced by more accurate versions at a later date. One of the theoretical goals of science is to formulate valid generaliza- tions, maps that summarize the pattern in nature in an accurate way. . . . Theoretical generalizations in biology often take the form of classification systems .... It must be realized that no classification . . . can fully represent all the varied and diverse facts, just as no map can show every detail in the terrain. And this inherent limitation of generalizations gives rise to endless controversies in biology whenever facts turn up that do not fit into the existing set of concepts. The attempt to reconstruct cultural development and diffusion in Latin America can also be compared to*the piecing together of a tremendously complicated jigsaw puzzle, in which 80 percent of the pieces are missing. Many of the available pieces do not fit, and those that can be joined together reveal little of the overall pattern. However, since even a tentative arrangement may suggest fruitful areas for intensive archeological research and specific problems to be investigated, it is worthwhile from time to time, and particularly when new pieces have been found, to renew the effort to put the puzzle together. On this justification, and guided by theoretical considerations outlined in the Introduction (pp. 5-9), we offer a hypothetical reconstruction of the origin and spread of certain cultural traits in northwestern South America from the appearance of pottery to about 3000 years ago. Pottery makes an abrupt appearance on the Guayas coast of Ecuador some 5000 years ago, injected into a preceramic and preagricultural, or at best, an incipient agricultural context, and transforming a group of shellfish gatherers and fishermen into what has been labeled the Valdivia Phase. The most specific and unique features of Valdivia culture, and therefore those best suited for establishing the origin of the Phase, are exhibited by the pottery complex of Period A. This early pottery contains a large number of unusual kinds of decoration, both in terms of technique and motif, as well as several distinctive rim and vessel shapes. With the passage of time, some of these die out and others become modified, so that by Period C the ceramic complex is very different from its initial content. Elements present at the inception of the Valdivia Phase on the Ecuadorian coast include the following: Decorative Techniques Figure 55 Shell stamping (Valdivia Shell Stamped) Fine zoned hachure (Valdivia Fine-line Incised) Shell combing (Valdivia Combed) Cut and beveled rim (Valdivia Cut-and-Beveled Rim) Incision on red slip (Valdivia Red Incised) 157 158 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Finger punching from the interior (Valdivia Mo- deled) Relief faces (Valdivia Embossed) Pseudo-corrugation (Valdivia Pseudo-Corrugated) Finger grooving (Valdivia Finger Grooved) Fingernail and fingertip punctation (Valdivia Fin- gernail Decorated) Corrugation (Valdivia Corrugated) Incision on an unpolished surface (Valdivia Incised) Shell scraping or brushing (Valdivia Brushed) Broad-line incision (Valdivia Broad-line Incised) Rim Shapes Figure 54 Indented rim (Combined Form 1) Broad exteriorly thickened rim (Combined Form 2) Direct rim with flat or square lip (Combined Form 4) Folded-over rim (Combined Form 13) Folded-over, finger-pressed rim (Combined Form 14) Short necked direct jar rim (Combined Form 18) Cambered rim bowl (Combined Form 11) In addition to technique of decoration, a number of specific motifs can be isolated as representative of Period A. Among these are the use of a row of punctation at the lower edge of the decorated area and zigzag, Crosshatch and horizontal parallel lines (Motifs 1, 2, and 3) in Valdivia Incised designs; I- shaped elements (or "dog-bones") (Motif 4) and anthropomorphic faces (Motif 1) in Valdivia Excised; one or more horizontal parallel lines below the rim (Motif 1), a band of rectilinear incision (Motif 2) and zoned parallel lines (Motif 4) in Valdivia Broad- line Incised. For pottery types restricted to Periods A and B, all motifs are of course characteristic. Filling of incisions or excisions with red pigment after firing is also probably present in Period A, although noted only on sherds from Period B. In vessel shape, undulating or lobed lips are fre- quent during Period A, as are small tetrapod feet. Sharply defined concave bases are also present, but the majority of both bowls and jars have slightly flattened bottoms. Surfaces may be unpolished, stri- ated polished or well polished. Some jars and bowls are red slipped, with application of the slip varying from uneven to even, and polishing varying from striated to complete. Period A vessels tend to be thicker walled than those of later periods. When the Valdivia Phase pottery was discovered in 1956, it was recognized by Estrada (1956) as con- taining a number of decorative elements characteristic of early Formative ceramics of coastal Peru, particu- larly Guafiape and Ancon. Now that the Valdivia sequence of pottery change can be described in great detail, it is of interest to note that decorative tech- niques and motifs used for this analysis were those of Period D and to a lesser extent Period C, since the Phase was then known only from the site of G-25 and later levels of G-31. This comparison thus sheds no light on the origin of the Valdivia ceramic com- plex. When a more detailed report was published three years later (Evans, Meggers and Estrada, 1959), the situation was not much different. Tentative cor- relations were made with the earliest Formative com- plexes in Peru (Guafiape), Panama (Monagrillo), Colombia (Barlovento) and Mexico (Tlatilco), but all of these except Tlatilco are characterized by pot- tery that is simple in form and predominantly undeco- rated. Tlatilco, which approaches Valdivia in artistry of the ceramic complex, was dated 1000 years more recent than the earliest Valdivia carbon-14 date available in 1959, ruling it out even then as a possible ancestor. During extensive excavations at G-31 in December of 1960, a new door was opened in the inquiry by the finding of part of the castellated rim of a Valdivia Red Incised vessel (pi. 103 a-b) bearing a striking resemblance in both rim treatment and incised decoration to a jar of Middle Jomon date from the Japanese island of Honshu (Estrada, 1961, Lam. 5, 1-2). Examination of other characteristics of pottery from Early and Middle Jomon sites revealed a large number of additional similarities, which were sum- marized by Estrada (op. cit., Cuadro 1) in terms of their distribution in Asia and the New World. This analysis showed the early Valdivia Phase to share a larger number of traits with Early and Middle Jomon than with any New World area or complex. Attempts to pin down more specifically the distribution and antiquity of the relevant traits in Japan were hampered, however, by linguistic barriers and by emphasis in most publications on the florescent type of Jomon pottery characteristic of the Middle period on Honshu, particularly as illustrated by complete vessels. Several clues, such as the reportedly small frequency of cord marking and the lesser exuberance of modeled decoration, pointed to Kyushu as a more likely source for Valdivia pottery than Honshu, but the approach to pottery analysis employed by Japanese archeologists (see Kidder, 1957, pp. 1-5) and the relatively small number and obscure nature of publications on Kyushu sites frustrated further bibliographic research. Fortunately, it was possible for us to visit Japan and to examine and photograph pottery from a number of Early and Middle Jomon sites on Kyushu, with results that have supported the hypothesis of transpacific origin for Valdivia Phase pottery beyond all reasonable expectation. Examination of Jomon collections was directed primarily toward finding early Valdivia elements of decoration and vessel shape. It became quickly WHOLE VOLUME ORIGIN AND AFFILIATIONS 159 Islands Period Divisions Sites Shell stamped Brushed Multiple drag-and-jab punctate Combed ?*" Finger grooved Rocker stamped q Nicked rib g Broad-line incised > "Valdivia Incised"-like ^ Corrugated H Nicked broad-line incised 3! _ 2 Fingertip punctate ?o Fingernail punctate - <¯ Excision *> Pseudo-corrugated Fine-Une incised Braid impressed Red slip Lobed rim Direct rim, flat lip JS Folded-over, finger-pressed rim > Curved cambered rim  Exteriorly thickened rim & Folded-over rim w Tapered rim o Tetrapod base w Concave base Honshu Middle Mito Natsushima Oomiyama Moroiso Kitashikawa Kominatodai Tatehira Fukkirizawa Shirahama 0 El 0 0 E El H H 13  E s E E E (?) EI E El El El El El El El El El E E E E E E E E Late Early Earliest Kyushu Late Middle Nampukuji Uki Mie Ataka Izumi Sobata Kasugacho Ryuo Yoshida Todoroki Nanshu Shrine Ishizaka E E El El E El El El El E E E E EIE E E E EEEEIE1E 12 El E E E E El E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E E 12 E E E El E E E E Middle Early Ear- liest Figure 96.-Early Valdivia Phase ceramic features at Early and Middle Jomon sites on Kyushu and Honshu. Data based on the following collections: Archaeological Museum, Meiji Univ. (Mito, Natsushima); Kyoto Univ. Museum (Moroiso, Kitashikawa, Uki); Furuta Collection, Shimabara (Mie, Todoroki); Kumamoto Castle Museum (Ataka, Sobata); Kumamoto Univ. (Sobata, Nampukuji, Ataka, Todoroki); Gyokuryu High School, Kagoshima (Izumi, Yoshida, Ishizaka, Nanshu Shrine, Kasugacho); Dept. of Archeology, Keio Univ. (Shirahama, Fukkirizawa, Tatehira, Kominatodai, Sobata, Ryuo); Oomiyama City Hall (Oomiyama). 160 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 evident that Late Jomon materials could be eliminated from consideration. Among Middle and Early Jomon samples, those from sites on Kyushu contained the most similar features. In order to examine the most material in the short time at our disposal, visits were arranged to individuals and institutions known to possess such materials on Kyushu. Without their wholehearted cooperation in allowing unrestricted access to their collections, it would be impossible to present the detailed comparative data on which the case for transpacific contact depends. Jomon sites from which pottery samples were examined are shown on figure 96, arranged from west to east (south to north) by islands and within each island from early to late. Geographical location is provided on the map (fig. 97). Dating by period follows the identification supplied by Japanese archeologists, and is based on a "horizon style" type of framework. Early Valdivia decorative techniques and rim shapes are listed across the top. It is im- mediately evident that the relevant items are clustered in sites from Kyushu, but scattered and unevenly represented at sites on Honshu. Since this analysis concerns only Valdivia-like traits, it does not take into consideration other characteristics of Jomon pottery from these areas, such as cord marking. When this is done, resemblances with Honshu become more tenuous than the chart suggests. Among the Kyushu sites taken as a group, all the early Valdivia traits are represented except four: pseudo-corrugation, braid impression and fine-line incision in decoration, and tapered rim in vessel shape. Some of the traits occur throughout most of the sequence (e.g., broad-line incision, corrugation, finger grooving, shell stamping, rim lobing), while others tend to be restricted to one time period. Rocker stamping, brushing (shell scraping), and nicked ribs are Early Jomon, while fingertip and fingernail punctation, and excision correlate with Middle Jomon. A similar temporal distribution is evinced by Honshu sites where the techniques are represented. It should be emphasized that complete sherd samples were not available from any of the sites, with the result that occurrence of some elements may have been overlooked. There is no reason to believe, however, that such additions would do any more than strengthen the existing pattern. The three sites showing the largest number of Valdivia Phase traits are Sobata, Izumi, and Ataka, dated as Late Early or Early Middle Jomon. Sobata and Ataka are located along the southern margin of what is now the Kumamoto Valley of Kyushu Island, but which several thousand years ago must have been a broad shallow bay. Numerous shell middens have been reported from the area, dating back to Early Jomon, but few have been investigated even super- ficially. Izumi, from southwestern Kyushu, is in a similar environmental setting. All three sites occupy the portion of the western coast protected from the open sea by Shimo Island and an irregularly shaped peninsula straggling southward. The existence of several such large nearly enclosed bays along the Kyushu coast may explain part of its attraction for Jomon fishermen (fig. 97). Although no carbon-14 dates are yet available from Kyushu Jomon Period sites, there are several from Honshu and Hokkaido for complexes representing all major period subdivisions on those islands (fig. 98). Since the Kyushu resemblances seem to concentrate around the transition between Early and Middle Jomon, the most significant Honshu date is that from Kamo Shellmound, an Early Middle Jomon site in Chiba Prefecture, of 5102 ñ400 (M-240) years ago. This is remarkably close agreement with the earliest Valdivia carbon-14 date of 5150 ñ150 (M-1320) years ago. Middle Jomon pottery, unlike that of the early Valdivia Phase, is the product of many centuries of local evolution, in which marked changes took place in decoration and vessel shape. Although the validity of the earlier dates, in excess of 9000 years ago, has been questioned, agreement between period affili- ation and carbon-14 date of more recent sites makes them appear possible. Words do not adequately express the degree of similarity between early Valdivia and contemporary Jomon pottery, and only photographs do justice to the remarkable resemblances (pis. 160-186, 189-190). In the Valdivia Incised and Valdivia Broad-line Incised types, not only technique of incision but motifs and combinations of motifs are the same (pis. 160-167). In most categories of decorative technique, examples can be found so similar in appearance that they might almost have come from the same vessel. There are, however, certain differences in emphasis that support an inference of evolutionary relationship, particularly in finger grooved, combed, and incised kinds of decoration. In the Valdivia Phase, it is possible to recognize two distinct kinds of incised designs; 1) broad-line incision on a polished surface (Valdivia Broad-line Incised), and 2) intermediate incision on an un- polished surface (Valdivia Incised). The former usually occurs on bowls, the latter on jars. Motifs are also dissimilar. However, among incised sherds of the Early and Middle Jomon periods there is a continuum in surface treatment and technique of incision that makes the division adopted for the Valdivia Phase inapplicable. Surfaces are less fre- quently polished, and Valdivia Incised motifs may be executed in Valdivia Broad-line Incised technique t,,,, ,u~,.,;.,~ !,-.,.>?;®., ,f t 160 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 evident that Late Jomon materials could be eliminated from consideration. Among Middle and Early Jomon samples, those from sites on Kyushu contained the most similar features. In order to examine the most material in the short time at our disposal, visits were arranged to individuals and institutions known to possess such materials on Kyushu. Without their wholehearted cooperation in allowing unrestricted access to their collections, it would be impossible to present the detailed comparative data on which the case for transpacific contact depends. Jomon sites from which pottery samples were examined are shown on figure 96, arranged from west to east (south to north) by islands and within each island from early to late. Geographical location is provided on the map (fig. 97). Dating by period follows the identification supplied by Japanese archeologists, and is based on a "horizon style" type of framework. Early Valdivia decorative techniques and rim shapes are listed across the top. It is im- mediately evident that the relevant items are clustered in sites from Kyushu, but scattered and unevenly represented at sites on Honshu. Since this analysis concerns only Valdivia-like traits, it does not take into consideration other characteristics of Jomon pottery from these areas, such as cord marking. When this is done, resemblances with Honshu become more tenuous than the chart suggests. Among the Kyushu sites taken as a group, all the early Valdivia traits are represented except four: pseudo-corrugation, braid impression and fine-line incision in decoration, and tapered rim in vessel shape. Some of the traits occur throughout most of the sequence (e.g., broad-line incision, corrugation, finger grooving, shell stamping, rim lobing), while others tend to be restricted to one time period. Rocker stamping, brushing (shell scraping), and nicked ribs are Early Jomon, while fingertip and fingernail punctation, and excision correlate with Middle Jomon. A similar temporal distribution is evinced by Honshu sites where the techniques are represented. It should be emphasized that complete sherd samples were not available from any of the sites, with the result that occurrence of some elements may have been overlooked. There is no reason to believe, however, that such additions would do any more than strengthen the existing pattern. The three sites showing the largest number of Valdivia Phase traits are Sobata, Izumi, and Ataka, dated as Late Early or Early Middle Jomon. Sobata and Ataka are located along the southern margin of what is now the Kumamoto Valley of Kyushu Island, but which several thousand years ago must have been a broad shallow bay. Numerous shell middens have been reported from the area, dating back to Early Jomon, but few have been investigated even super- ficially. Izumi, from southwestern Kyushu, is in a similar environmental setting. All three sites occupy the portion of the western coast protected from the open sea by Shimo Island and an irregularly shaped peninsula straggling southward. The existence of several such large nearly enclosed bays along the Kyushu coast may explain part of its attraction for Jomon fishermen (fig. 97). Although no carbon-14 dates are yet available from Kyushu Jomon Period sites, there are several from Honshu and Hokkaido for complexes representing all major period subdivisions on those islands (fig. 98). Since the Kyushu resemblances seem to concentrate around the transition between Early and Middle Jomon, the most significant Honshu date is that from Kamo Shellmound, an Early Middle Jomon site in Chiba Prefecture, of 5102 ñ400 (M-240) years ago. This is remarkably close agreement with the earliest Valdivia carbon-14 date of 5150 ñ150 (M-1320) years ago. Middle Jomon pottery, unlike that of the early Valdivia Phase, is the product of many centuries of local evolution, in which marked changes took place in decoration and vessel shape. Although the validity of the earlier dates, in excess of 9000 years ago, has been questioned, agreement between period affili- ation and carbon-14 date of more recent sites makes them appear possible. Words do not adequately express the degree of similarity between early Valdivia and contemporary Jomon pottery, and only photographs do justice to the remarkable resemblances (pis. 160-186, 189-190). In the Valdivia Incised and Valdivia Broad-line Incised types, not only technique of incision but motifs and combinations of motifs are the same (pis. 160-167). In most categories of decorative technique, examples can be found so similar in appearance that they might almost have come from the same vessel. There are, however, certain differences in emphasis that support an inference of evolutionary relationship, particularly in finger grooved, combed, and incised kinds of decoration. In the Valdivia Phase, it is possible to recognize two distinct kinds of incised designs; 1) broad-line incision on a polished surface (Valdivia Broad-line Incised), and 2) intermediate incision on an un- polished surface (Valdivia Incised). The former usually occurs on bowls, the latter on jars. Motifs are also dissimilar. However, among incised sherds of the Early and Middle Jomon periods there is a continuum in surface treatment and technique of incision that makes the division adopted for the Valdivia Phase inapplicable. Surfaces are less fre- quently polished, and Valdivia Incised motifs may be executed in Valdivia Broad-line Incised technique /FukiriZQwa Kasugacho 8 Nanshu Shrine TANEGASHIMA ISLAND Figure 97.-Map of Japan, showing location of Jomon sites producing pottery resembling Valdivia Phase types. 166 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Japan, when we were shown stone figurines from the recentiy discovered rock shelter of Kamikuroiwa, in Ehime Prefecture, northwestern Shikoku Island. Associated pottery is of a unique type, leading to tentative identification of the site as pre-Jomon (Esaka, 1962). The unshaped, thin, flat, waterworn pebbles, 4.3-6.4 cm. long, 2.0-4.0 cm. wide and 4-5 mm. in maximum thickness, are plain on one face and decorated with fine scratchlike incisions on the other (pi. 187 a-f). On the majority, scratches run in two groups from one end toward the middle, leaving a gap between them, and the other end plain. Esaka sug- gests that this may be stylized hair. Two have a band of parallel incised lines at the other end bounded by a horizontal incision, which if this is the lower end might represent a fiber skirt (pi. 187 a, c). The material is micaceous schist of a pale green hue. Reexamination of Valdivia Phase stone figurines of Palmar Plain and Palmar Notched types in the light of these features suggests the possibility that vertical scratches covering the surface of some examples may not simply reflect method of manufacture, but may instead have symbolic significance. This impression is strengthened by the fact that a few of the figurines have incisions restricted to one half of the body (pi. 187i), or have additional short diagonal strokes at the upper end (pi. 187g). One has short parallel incisions bounded by a horizontal line at one end, similar to the Japanese "fiber skirt" treatment (cf. pis. 187j and 187e). While not conclusive evidence of connection, these resemblances suggest that further search may reveal crude stone figurines in Early Jomon sites that will serve as a link between the Kamikuroiwa examples and Valdivia stone figurines of Palmar Plain and Palmar Notched types. Turning to parallels in stone, bone and shell artifacts between early Valdivia and Early Middle Jo- mon introduces a factor of negligible significance in pot- tery and figurines, namely, functional limitation on form. While pottery is utilitarian, whether rims are direct or exteriorly thickened, whether bases are flattened or concave, whether shoulders are angular or rounded does not affect the ability of a vessel to hold liquid, nor are these differences attributable to characteristics of the raw material. However, an awl must conform to the general contour of the antler fragment from which it is carved, and a sinker must be of appropriate weight or its efficiency is reduced. Similarities must also be evaluated in terms of the antiquity of shellfish gathering and fishing as a subsistence pattern, and the possibility that survival rather than diffusion may account for common types of artifacts. Such seems to be the case with notched pebble sinkers, which occur in Early Jomon and early part of the Valdivia Phase, but also occur in preceramic sites elsewhere in the New World (e.g., Emperaire and Laming, 1961, pp. 29-30 and fig. 6). The Valdivia shell fishhook complex, including saws and reamers used for manufacture and the hooks themselves, fits into a generalized New World preceramic tradition, and the objects were apparently so well adapted to the function they served that they survived in places without major alteration until recent times (Schumacher, 1960, pp. 23-24 and fig. 5 a-j). The efficiency of these tools and techniques of manufacture for supplying subsistence requirements can be cited to account for failure of Jomon types of fishhooks and projectile points to be adopted, if the Valdivians were made aware of their characteristics. The few Valdivia Phase shell ornaments can also be duplicated in preceramic sites on the coast of Peru (Engel, 1963a, pp. 52-53) and California shell mounds (Gifford, 1947), suggesting either that they are part of the same widely diffused ancient tradition or that they are sufficientiy simple and decorative to have been independently invented a number of times. The bulk of stone artifacts are flakes and cores chipped by percussion either accidentally during use in hammering, or intentionally to produce a cutting edge. In general character, they duplicate the com- plex described by Bird at Huaca Prieta, and his comment that "if only the stone artifacts had survived one would be forced to think of these people as exceedingly primitive culturally" (1948, p. 25) is equally applicable to the Valdivia Phase situation. The principal difficulty with postulating New World antecedents for nonceramic and nonfigurine aspects of Valdivia culture is that no shell middens of unques- tionable preceramic date have been identified as yet on the Ecuadorian coast. In view of the large num- ber of such sites now known on the adjacent Peruvian coast, however, it seems probable that this is explain- able on two grounds: 1) the relatively slight amount of search for such sites in Ecuador, and 2) the wet climate, which destroys all perishable materials. Shell deposits, composed principally or entirely of oyster, are numerous along margins of salitres, particularly between the Santa Elena Peninsula and Posorja, but our spot testing of these places produced no artifacts, firecracked stones or ash indicative of human origin. In view of the sparcity of cultural remains in pre- ceramic refuse deposits in other parts of South America (e.g., Bird, 1938, p. 253), however, more excavation is required before a natural origin can be confidently assumed. A shell sample from one accumulation in El Morro Salitre was carbon-14 dated in the hope of shedding light on the problem. The resulting date of 26,900 ñ900 (1-255) years is inconclusive, since it is within the magnitude now suggested for man's occu- pation of North America although considerably older WHOLE VOLUME ORIGIN AND AFFDUATIONS 167 than any previously obtained South American dates. Consequently, at present we can only suggest that all evidence points to the existence of a preceramic popu- lation prior to the introduction of pottery on the Ecuadorian coast, while noting that confirmation in the form of sites or artifacts does not exist so far. One other parallel between Valdivia and Jomon is the position of burial, in which the legs are tightiy flexed and the arms extended at the side. This posi- tion was represented in all the Valdivia Phase skeletons from the Period C site of G-54 that were sufficientiy well preserved for observation (p. 17; pi. 12). The same position occurs in Jomon burials (Komatsu, 1962, pi. 64). However, since the only Valdivia burials found so far are from a late part of the Phase, and Jomon burials also exhibit a variety of other positions, this similarity may be coincidental. Physi- cal anthropological analysis of the Valdivia Phase skulls showed them to differ from preceramic popu- lations of similar or greater antiquity on the Peruvian coast, but comparative series are too few to indicate what significance may be attached to these differences. Consolidation of evidence relevant to the origin of the Valdivia Phase leads to a reconstruction of events somewhat as follows. Some 5000 years ago, the coasts of Japan and western America were occupied by small groups of people who subsisted by fishing and shellfish gathering, supplemented by hunting of terrestrial mammals. In addition to gathering plants, they may have taken initial steps toward cultivation. Their tools and utensils, although differing in details of construction, were of similar kinds-fishhooks, awls, choppers, flakes, sinkers, hammerstones-in- dicating both a similar level of technological com- petence and a similar ecological adaptation. In Japan, sites are found not only along the coast, but up river valleys into the interior, where salmon fishing, hunting of terrestrial fauna, and an unusually rich edible flora seem to have approximated wild food conditions on the northwest coast of North America. Bones from the refuse indicate that deep sea fishing provided part of the food supply from early times, and it is probable that dugout canoes were the vehicle from which this was done. According to Ploszajski (1963, pp. 88-89): A large number of dug-out canoes have been found, the earliest one being found in peat at Kamo, Chiba Prefecture is estimated to be from early Jomon Period before 3000 B.C. It has a shallow rounded cross-section with width-depth ratio of approximately 4:5 and tapering bow and stern which un- fortunately were seriously damaged. A comparatively well preserved paddle was also found with this boat. If a boatload of Early Middle Jomon fisherman left the sheltering bays of Kyushu and went out into the sea off the southeastern coast in October or November, they would have entered a zone with some of the strongest currents in the northern Pacific, running northeastward at 24-32 miles per day (fig. 103). U.S. Weather Bureau records for the 40-year period between 1901-1940 tabulate 802 typhoons, of which 130 were in October and 67 in November. A canoe caught too far from shore by one of these storms might easily be swept by the combined northeasterly pressure of wind and current far out to sea before control was regained. Even if the occupants retained possession of their paddles, they might have been unable to turn back. During the month of November, westerly and northerly winds predominate in the northern hemisphere, and are steadiest and of greatest force between about the 40th and 55th parallels. In addition, the percentage of gales increases during November in high latitudes, occurring at an average frequency of one every 8-10 days over the greater part of the northern Pacific except near coasts (Hydro- graphic Office, 1960). A combination of these forces would have borne a canoe eastward along the great circle route, which on a flattened map curves far north of Hawaii (fig. 103). Records during the past century demonstrate the feasibility of such a drift vessel reaching land with living passengers after a voyage of as long as 11 months (Sittig, 1896, p. 530). Indeed, possibility of survival would be stronger for people accustomed to living from the sea than for land-oriented agriculturalists of more recent times. The currents would have carried a canoe southward as it neared the Pacific coast of the Americas. De- pending on the force of the wind and the direction of the currents, the dugout might have been brought in sight of land off Mexico or Central America, some- thing that may have happened to other voyagers in later times (fig. 103). These particular Jomon fisher- men were less lucky, however, and it was some weeks more before they reached the coast of Ecuador. At this point, southeastward flowing currents meet the northwestward flowing Humboldt current, and join forces to run westward across the Pacific. The Guayas coast was thus the "end of the road" as far as the New World was concerned. The junction of these opposite currents fluctuates northward or southward during the year, causing minor changes of direction in local currents off Colombia and Panama (fig. 103, upper right). It is tempting to try to suggest a time of year that would have been more propitious for arriving at Valdivia, but the complexity of current patterns and the unknown quantity represented in wind direction and force reduce any such effort to the status of pure guess. An estimate of the number of months necessary to make a trip of some 9450 miles (8230 nautical miles) also involves so many incalculables, such as days of calm, temporary devia- WHOLE VOLUME ORIGIN AND AFFELIATIONS 161 JOMON VALDIVIA HOKKAIDO HONSHU GUAYAS LATE Uenae-3230*160 (W-322) Kusaka-3060*II0 (GaK-170) 3450*50 (SI-69) 4050*200 (W-630) 4270*60 (SI-81) 4620*140 (M-1322) D MIDDLE Nakazawa-3825*175 (1-552) Taniguchi-3950ñ200 (W-372) Tokoro-4I50*400 (GaK-188) C Ubayama-4513*300, 4526ñ220 Oomiyamo- 4 580*60 (SI-93) Kcmo-5I02ñ400 (M-240) B A EARLY Misato Cave-6800ñ225 (1-553) Isonomori -7830*350 (M-238) EARLIEST Kojohama -7680 + 200 (1-550) 77001200 (1-551) Kishima - 8400*350 (M-237) Natsushima - 9240*500 (M-770) 9450*400(M-769) Figure 98.-Carbon-14 dates for Jomon sites on Hokkaido and Honshu, with selected Valdivia Phase dates for comparison. (pi. 164 a-c, 167h). In other words, the Jomon incised style can be viewed as a common ancestor out of which the two Valdivia types have differentiated. A similar situation exists with finger tip punctation (pi. 178 c-g) and pseudo-corrugation. These two decorative techniques are readily distinguished in the Valdivia Phase, and would not necessarily be recognized as related. However, in Jomon pottery there is a steady progression from one to the other (pi. 169), making it possible to postulate a process of evolution in which intermediate examples disappeared with the resulting emergence of two seemingly un- related pottery types. Multiple drag-and-jab punctate, a relatively rare technique in early Valdivia pottery, is more common in Jomon and displays a wider variation in execution. In Valdivia examples, adjacent bands are so carefully placed that the decorative technique is difficult to analyze. Although drag-and-jab by a multitoothed instrument, probably a piece of fluted shell, seemed the most probable method of execution, this was confirmed only after examination of the Jomon sherds. Here, multiple drag-and-jab occurs both in a continuous zone and in isolated bands, so that the track of a tool is clearly defined (pis. 170, 171, 183 g-i) and the technique obvious. Combed and finger-grooved decoration are also more elaborated in Jomon. Valdivia Phase combing consists of individual vertical continuous or broken bands on a plain surface, whereas the Jomon version utilizes continuous straight or wavy lines, typically on a horizontally combed surface (pis. 172, 173). In both areas, this kind of decoration can be dis- tinguished from brushing or shell scraping as a technique of surface texturing by its more systematic patterning, brushing being random in orientation (pis. 174, 175). Finger grooving is also more elabo- rated in Jomon, the surface between grooves typically ornamented with gashes or punctates. This added embellishment is rare in Valdivia (pi. 176). On the other hand, the characteristic Valdivia arrangement of vertical finger grooves on the neck of jars seems to be uncommon in Jomon. In other instances, technique and motif of decora- tion are nearly identical. Excision makes use of the same crudely gouged out "hour-glass" or "dog-bone" elements (fig. 99; pi. 177). Red slipped vessels have interlocking rectilinear designs in broad-line incision alike in all details (pi. 178 a-b). Rocker stamping appears in the unusual "dragged" variant, as well as the more familiar form (pi. 179), and is applied 162 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 either as overall treatment or in isolated straight or meandering bands. Certain combinations of techniques occur in both areas, among them shell scraping and shell stamping (pi. 180), fingertip punctation and finger grooving, excision and broad-line incision, rocker stamping and incision, incision and punctation (pi. 168), incision and nicked or finger-pressed rib (pi. 189 b, g). Others are restricted to Jomon, such as incision and shell stamping, multiple drag-and-jab punctate and shell stamping, fingernail punctation and shell stamping-all of which reflect the greater frequency of shell stamping in Jomon as opposed to Valdivia. Undulating and lobed rim embellishments in both areas are associated principally with broad-line incised and excised decoration (pi. 181). Folded- over rims are almost exclusively correlated with undecorated jars in Valdivia, whereas they often occur on incised vessels in Jomon. In both complexes, the lip may be nicked or finger pressed (pi. 182). Three of the most popular Valdivia Period A decorated types are absent from all of the Kyushu collections examined. The principal early incised type, Valdivia Fine-line Incised, characterized by fine zoned hachure on a red-slipped surface, is represented by a few sherds from Natsushima and Mito shell mounds, on the southern coast of Honshu, both said to be of Middle Jomon date (pi. 183, a-f). Although this type has not to our knowledge been reported from Kyushu, its existence on Honshu suggests that it may be found with more careful search. No examples of the interior finger punching were observed in any of the collections, nor have we been able to discover any illustrations, in spite of the fact that several Japanese archeologists reported this kind of decoration to exist in Japan, and also on the north Asiatic mainland. Authorities and personal inspection of collections were both negative regarding a rim treatment like that characteristic of Valdivia Cut and Bevelled Rim. This could be a Valdivia Phase refinement of the lobed treatment often found on Jomon rims, in which case it would not exist in Japan; on the other hand, further search may yet turn it up. Two rare early Valdivia techniques of decoration are also represented on Jomon pottery, but not in Kyushu collections examined. Valdivia Embossed, in which a face or stylized circular element is executed in relief on the exterior adjacent to the rim, is paralleled in the U.S. National Museum collections from the Late Jomon shell middens of Omori (pi. 184c), and Okubo (pi. 184a) in Chiba Prefecture, Honshu. A similar face on the side of a jar (pi. 184b), indicates the persistence of such treatment into the Later Jomon Period on Honshu (Esaka, 1960, p. 266). The similarity of these Valdivia faces to Jomon ones is particularly interesting in view of the total discon- tinuity between this form of anthropomorphic repre- sentation and that of pottery figurine faces in Valdivia Phase. Designs incorporating impressions apparently pro- duced by pressing a braided cord onto the surface or into an incised line occur on the pottery of both areas (pi. 185 c-d, f-g). Jomon examples come from Honshu or Hokkaido. Impression with a single cord (pi. 185 a-b, e) does not give the appearance of having been added to previously incised lines charac- teristic of braided cord impressions. The uniqueness of the braided cord impression and the fact that its maximum frequency is during Valdivia Period B, suggest that this decorative technique must be part of the introduced complex, and that its apparent absence from Valdivia Period A and Kyushu Early Middle Jomon must be a reflection of the small size of the pottery sample or the rarity of the technique or both. In rim and vessel shape, the degree of similarity between Early Middle Jomon and early Valdivia is equally striking. Open or constricted bowls with direct rim and various kinds of carinated bowls are typical Jomon shapes (pis. 185 h-i, 186). Limitations Figure 99.-Jomon and Valdivia Phase sherds with similar excised decoration, a, Mie. b, Nagasaki, c, Valdivia Phase. WHOLE VOLUME ORIGIN AND AFFILIATIONS 163 of time permitted drawing rim profiles only for sherds bearing Valdivia Phase types of decoration, so that relative frequency of the shapes in the Early Middle Jomon ceramic complex is not known. However, it is interesting and probably significant that there is a similar association between rim profiles and decorative treatment in Jomon and Valdivia. When examples from the two areas are intermixed, as has been done for seven groups (fig. 100), matching extends to thick- ness, orientation, and lip form. Although variation exists within each group, examples from both areas occur throughout the range. The only consistent difference is a tendency toward rounding of the lip on Valdivia rims of Form A, and rounding or tapering of the lip on Form B, while Jomon rims typically have a flat lip. This distinction is not evident in the remaining forms. Parallels in Forms E and F are particularly interesting, since these are very rare rim types in early Valdivia. The existence of a Jomon prototype suggests they represent an introduction that failed to become popular. The folded-over rim, Form B, appears to occur with similar frequency in both areas, but the finger-pressed variant so abundant in early Valdivia is much rarer in Jomon, where embellishment was channeled more in the direction of lobing. Kyushu rim treatment never approached /Ml m\ ? ? i ? i ?' 0 I 2 3 CM Figure 100.-Similarity between rim profiles of Jomon (solid) and early Valdivia (outline) pottery vessels, a, Exteriorly thickened. b, Folded-over, c, Direct, flat lip. d, Everted, flat lip. e, Everted and angular on interior. /, Curved cambered, g, Tapered. 164 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 the baroque elaboration typical of Honshu Middle Jomon, and is limited principally to undulating, notched or rectanguloid lobed treatment of the lip on bowls and jars, approximating the early Valdivia situation (pi. 181). Two typical early Valdivia base forms-small tetra- pod feet and a well defined circular depression- might also be expected to be represented in the Jomon material. Kidder (1957, fig. 5, no. 11) illustrates a tetrapod bowl as one of the undecorated vessel shapes in Horinouchi and Kasori types of pottery of Late Jomon date on Honshu, but we were able to find only two examples from Kyushu. One is from Izumi Shellmound, a site with numerous other Valdivia-like features (fig. 96). The other is from the site of Tanaka, for which no other information is available. Both are much cruder and somewhat larger than the Valdivia examples, but could have served as proto- types. Unfortunately, no concave bases similar to the Valdivia ones were observed, but it is possible that this was in part because of the rarity of plain sherds in collections available for examination. Support for such an assumption exists in Kidder's (1957, p. 90) statement that concave bases occur in Izumi pottery. Castellated rim treatment, the feature that first called attention to the possibility of a transpacific origin, is very rare in the Valdivia Phase ceramic complex and apparently confined to the earliest part of Period A. Two variants are represented: 1) a slender vertical "prong," and 2) an outflaring point. Both of these extremes, and all degrees of variation between them occur on Jomon vessels (Kidder, 1957, figs. 12-27). One example of the first type, classified as Horinouchi A, shows not only the same rim treat- ment as the early Valdivia sherd (pi. 103 a-b), but has the rib leading to the castellation nicked hori- zontally in the same manner, and similar zigzag incised decoration on the upper body wall. It is somewhat later in date, if Jomon sequences are reli- able, implying a convergence of vessel shape, decora- tive technique and motif in the two widely separated areas. A rim sherd very similar in outline and orien- tation to the second Valdivia type of castellation comes from Todoroki Shell Mound in southern Kyushu, dating from the Early Jomon Period (fig. 101a). Broad-line incision in horizontal parallel lines occupies the exterior adjacent to the rim, leaving the remaining surface undecorated. The Valdivia exam- ple has a similar band adjacent to the rim, with zoned punctation below and nicked lip treatment (fig. 101b). Although rim profiles are slightly differ- ent, there is no doubt that the general vessel shape is the same. Several points of negative evidence also favor a Kyushu origin for Valdivia Phase pottery. One is the relative rarity of cordmarking. Cordmarking is almost nonexistent in Valdivia pottery, a fact that would be difficult to reconcile with a derivation from Honshu or Hokkaido, where this technique of decora- tion is exceedingly common from an early time. A second important consideration is the absence on Kyushu of the excessively ornate pottery, featuring fantastic castellations and high relief ornament producing a sculptural effect, associated with Middle Jomon on Honshu (Kidder, 1957, figs. 15-17). In fact, this "classic" Jomon pottery is less like contemporary material from southern Kyushu than is the pottery from across the Pacific at Valdivia (op. cit. p. 98). a ~^/ b Figure 101.-Jomon and Valdivia Phase sherds with similar castellated rim treatment, a, Todoroki. b, Valdivia Phase. WHOLE VOLUME ORIGIN AND AFFILIATIONS 165 If it can be granted that the idea of pottery making in the Valdivia Phase originated by transpacific contact from somewhere on Kyushu, it is of interest to investigate whether other elements may have been introduced at the same time, and particularly whether any Jomon antecedents can be found for the Valdivia Phase Period A stone figurines. Three types of the latter have been recognized: 1) Palmar Plain: a small, thin, elongated, plain surfaced slab (pi. 117 a-q); 2) Palmar Notched: a slightly thicker slab with a notch at one end suggesting legs (pi. 117 r-ee); and 3) Palmar Incised: a broader, thicker, larger stone with face and arms incised on one surface (pi. 118 a-j). The faces are executed in a stylized manner, often with a broad nose and thick lips. A search for parallels in Jomon stone or pottery material is at first disappointing. The Japanese pottery figurines are not only totally dissimilar in all respects, but are generally confined to the Late Jomon period, so that their only significance from a comparative point of view is to demonstrate that use of figurines was shared by the two cultures. There are in addition, however, a number of carved stone "plaques", which although principally repre- senting Late Middle and Late Jomon show several Valdivia-like features. The plaques are rectanguloid or ovoid in outline, with flat or slightly convex surfaces and rounded edges (Kidder, 1957, figs. 44-45). One or both surfaces are decorated, often elaborately, with a bilaterally symmetrical or asymmetrical incised design making liberal use of spirals and volutes. Some have a readily recognizable face at one end (e.g., Kidder, 1957, fig. 44, no. 18). More frequently, only two ovoid eyes with a horizontal incision across the center reveal the anthropomorphic nature of the artifact (fig. 102c). Many bear a purely geometric decoration. Within this wide variety, several plaques show features also represented on the Valdivia Phase stone figurines of the Palmar Incised type. One is a pair of concentric semicircles extending from each side toward the center, in a manner like that employed on Valdivia figurines for stylization of the arms (fig. 102b). Another is a T-shaped junction of horizontal eyebrows and a narrow vertical nose (Kidder, 1957, fig. 44, no. 8-9, 20-21). A third similarity lies in use of an ovoid or rectanguloid outline with a horizontal incision at the center for depicting eyes (op. cit., fig. 45, no. 9-10), a treatment otherwise exhibited in the Valdivia Phase only on anthropomorphic faces incised or excised on vessel walls (pis. 41, 58 a-m). Several of the Late Jomon plaques with geometric ornamentation duplicate the interlocking motif characteristic of Valdivia Red Incised and Valdivia Excised pottery (fig. 102a-b; pis. 59a, 102 c-d) in a striking way. Efforts to locate figurines resembling Valdivia stone Palmar Plain and Palmar Notched types in Jomon collections were unproductive at first. All scholars who saw the photographs disclaimed knowledge of anything similar until almost the end of our trip in Figure 102.-Late Jomon stone and pottery plaques from Honshu sites with decorative motifs resembling those on Valdivia Phase pottery. a, Height 26.2 cm. (after Esaka, 1960, fig. 149). b, Height 7.5 cm. (after op. cit., fig. 119). c, Height 14.5 cm. (after op. cit., fig. 141). 168 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 tions from the most direct route, etc. as to be of little value. It can only be said with certainty that the trip must have taken many months, and that one or more members of the original crew must have sur- vived, probably well tanned! Arriving on the Ecuadorian shore, the travelers were met or soon found by the local residents, who pre- sumably were living much the same kind of life as had been left behind on Kyushu-fishing, shellfish gathering, a little hunting and gathering of plants for food and fiber. The results make it apparent that the newcomers were welcomed and incorporated into the community. In the process, they introduced the art of pottery making, and very probably new religious practices that are reflected in the stone figurines. Other new ideas may also have been incorporated, but no tangible evidence has been recognized. The Valdivians quickly became skillful potters, and in fact the most striking aspect of the material is its superiority in quality and artistry to pottery not only of contemporary Kyushu, but also of all other early Formative complexes in the New World. It is an ironical fact that attempts to relate Valdivia ceramics to Guafiape on the coast of Peru, Monagrillo on the Pacific coast of Panama, or Puerto Hormiga on the Caribbean coast of Colombia involve more tenuous comparative evidence than can be brought to support the inference of Jomon-Valdivia connections. Un- doubtedly, this situation would, if properly under- stood, shed light on the kind of contact that underlay diffusion of potterymaking both north and south (see pp. 6-7). Puerto Hormiga, on the north Colombian coast (fig. 104), has produced the earliest pottery in South America aside from the earliest Valdivia date. Two carbon-14 dates are available, 4875 ñ170 (1-445) and 4515 ñ250 (1-1123) years ago, the former based on a shell sample and the latter on a charcoal sample from 30 cm. higher up in the deposit. The ceramic complex shows a number of striking similarities to early Valdivia (fig. 105) including decoration by multiple drag-and-jab punctate (pi. 188 c-e), finger grooving (pi. 188 a-c), broad-line incision, sometimes red-filled (Reichel-Dolmatoff, 1961b, p. 352), or nicked (pi. 188 c), and a ring with a dot at the center (Reichel-Dolmatoff, 1961b, Lam. II, 6). Several non-Valdivia Period A-B techniques and motifs of decoration are associated, including zoned dentate (pi. 188f), incisions terminating in a punctate (pi. 188b) and narrow bands of parallel hachure bounded by broad-line incision (pi. 188 1). Until publication of the detailed analysis of the Puerto Hormiga sequence it cannot be determined whether the latter are present from the beginning or introductions at some later time. Vessel shapes and rim profiles have not been described in sufficient detail for comparison with early Valdivia shapes. Monagrillo, on the Pacific coast of Panama, with a carbon-14 date of 4090ñ70 years (Y-585), is approximately contemporary with the beginning of Valdivia Period C, and it is interesting to note the occurrence of nicked broad-line incision (Willey and McGimsey, 1954, fig. 47c), a decorative technique diagnostic of this part of the Valdivia Phase sequence (fig. 105). Other Monagrillo decorative techniques are excision (op. cit., figs. 12 a-d, 46 c, g, k), broad- line incision (op. cit., figs. 46f, 47b), and incisions terminating in a punctate (op. cit., figs. 12z, 46 g-i, 1, 47a). Some sherds have scoring in the excised zones that may be related to the zoned hachure of other early complexes (op. cit., fig. 46 c, k-1). A Valdivia-like feature of vessel shape is the folded-over rim (op. cit., fig. 45h). The most recent of the shell midden complexes reported on the north Colombian coast, that of Barlovento (fig. 108), incorporates a number of ele- ments aligning it with the Valdivia-Puerto Hormiga- Monagrillo tradition (fig. 105), such as broad-line incision (Reichel-Dolmatoff, 1955, Lam. Ill, 6, 9), incision terminating in a punctate (op. cit., Lam. IV, 7), and post-firing application of red pigment to incisions (op. cit., p. 256). Certain examples of zoned hachure (op. cit., Lam. Ill, 3) can be seen as evolving from the Monagrillo practice of scoring excised areas. Circles or ovals frequently have a central punctate (op. cit., Lam IV, 3-6) reminiscent of the Puerto Hormiga dot-in-ring. Small rings stamped with a tubular instrument are relatively frequent (op. cit., Lam. Ill, 3, 8). Zoned punctation is the most common kind of decoration, produced with a variety of tools some of which created a circular mark (op. cit., Lam. IV, 5, 9). The abun- dance and variety of punctation makes it possible to interpret the presence of ring and zoned circular punctate elements in Barlovento and Valdivia Phase pottery as coincidental. On the other hand, they form part of a complex with wide temporal and spacial distribution and in this context their possible diagnostic value cannot be overlooked. With carbon- 14 dates ranging between 3470 ñ120 (W-739) and 2980 ñ120 (W-741) years ago, Barlovento is ap- proximately contemporary with late Period C and Period D of the Valdivia Phase, and if the analysis of its origin is correct represents the product of some 1300 years of divergent evolution. Turning toward the south, the earliest pottery is from the Guafiape complex on the north Peruvian coast (fig. 104), with a carbon-14 date of 4300ñ200 years (L-122D) for a shell sample from upper Early Guafiape and 3100ñ200 years (L-122C) for a 4ARCH- MAY ASIA sept.-nov, *. ( |ao AFTER PILOT CHART OF THE NORTH PACIFIC OCEAN NO. 1401 - NOVEMBER I960 INSERTS AFTER ATLAS OF PILOT CHARTS: SOUTH PACIFIC ft INDIAN OCEANS. PUB. NO. 107 HYDROGRAPHIC OFFICE, NAVY DEPT. WASHINGTON, D.C. Figure 103.-The northern Pacific Ocean, showing circle route between Kyushu, Japan dnd the Guayas 168 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 tions from the most direct route, etc. as to be of little value. It can only be said with certainty that the trip must have taken many months, and that one or more members of the original crew must have sur- vived, probably well tanned! Arriving on the Ecuadorian shore, the travelers were met or soon found by the local residents, who pre- sumably were living much the same kind of life as had been left behind on Kyushu-fishing, shellfish gathering, a little hunting and gathering of plants for food and fiber. The results make it apparent that the newcomers were welcomed and incorporated into the community. In the process, they introduced the art of pottery making, and very probably new religious practices that are reflected in the stone figurines. Other new ideas may also have been incorporated, but no tangible evidence has been recognized. The Valdivians quickly became skillful potters, and in fact the most striking aspect of the material is its superiority in quality and artistry to pottery not only of contemporary Kyushu, but also of all other early Formative complexes in the New World. It is an ironical fact that attempts to relate Valdivia ceramics to Guafiape on the coast of Peru, Monagrillo on the Pacific coast of Panama, or Puerto Hormiga on the Caribbean coast of Colombia involve more tenuous comparative evidence than can be brought to support the inference of Jomon-Valdivia connections. Un- doubtedly, this situation would, if properly under- stood, shed light on the kind of contact that underlay diffusion of potterymaking both north and south (see pp. 6-7). Puerto Hormiga, on the north Colombian coast (fig. 104), has produced the earliest pottery in South America aside from the earliest Valdivia date. Two carbon-14 dates are available, 4875 ñ170 (1-445) and 4515 ñ250 (1-1123) years ago, the former based on a shell sample and the latter on a charcoal sample from 30 cm. higher up in the deposit. The ceramic complex shows a number of striking similarities to early Valdivia (fig. 105) including decoration by multiple drag-and-jab punctate (pi. 188 c-e), finger grooving (pi. 188 a-c), broad-line incision, sometimes red-filled (Reichel-Dolmatoff, 1961b, p. 352), or nicked (pi. 188 c), and a ring with a dot at the center (Reichel-Dolmatoff, 1961b, Lam. II, 6). Several non-Valdivia Period A-B techniques and motifs of decoration are associated, including zoned dentate (pi. 188f), incisions terminating in a punctate (pi. 188b) and narrow bands of parallel hachure bounded by broad-line incision (pi. 188 1). Until publication of the detailed analysis of the Puerto Hormiga sequence it cannot be determined whether the latter are present from the beginning or introductions at some later time. Vessel shapes and rim profiles have not been described in sufficient detail for comparison with early Valdivia shapes. Monagrillo, on the Pacific coast of Panama, with a carbon-14 date of 4090 ñ70 years (Y-585), is approximately contemporary with the beginning of Valdivia Period C, and it is interesting to note the occurrence of nicked broad-line incision (Willey and McGimsey, 1954, fig. 47c), a decorative technique diagnostic of this part of the Valdivia Phase sequence (fig. 105). Other Monagrillo decorative techniques are excision (op. cit., figs. 12 a-d, 46 c, g, k), broad- line incision (op. cit., figs. 46f, 47b), and incisions terminating in a punctate (op. cit., figs. 12z, 46 g-i, 1, 47a). Some sherds have scoring in the excised zones that may be related to the zoned hachure of other early complexes (op. cit., fig. 46 c, k-1). A Valdivia-like feature of vessel shape is the folded-over rim (op. cit., fig. 45h). The most recent of the shell midden complexes reported on the north Colombian coast, that of Barlovento (fig. 108), incorporates a number of ele- ments aligning it with the Valdivia-Puerto Hormiga- Monagrillo tradition (fig. 105), such as broad-line incision (Reichel-Dolmatoff, 1955, Lam. Ill, 6, 9), incision terminating in a punctate (op. cit., Lam. IV, 7), and post-firing application of red pigment to incisions (op. cit., p. 256). Certain examples of zoned hachure (op. cit., Lam. Ill, 3) can be seen as evolving from the Monagrillo practice of scoring excised areas. Circles or ovals frequently have a central punctate (op. cit., Lam IV, 3-6) reminiscent of the Puerto Hormiga dot-in-ring. Small rings stamped with a tubular instrument are relatively frequent (op. cit., Lam. Ill, 3, 8). Zoned punctation is the most common kind of decoration, produced with a variety of tools some of which created a circular mark (op. cit., La"m. IV, 5, 9). The abun- dance and variety of punctation makes it possible to interpret the presence of ring and zoned circular punctate elements in Barlovento and Valdivia Phase pottery as coincidental. On the other hand, they form part of a complex with wide temporal and spacial distribution and in this context their possible diagnostic value cannot be overlooked. With carbon- 14 dates ranging between 3470 ñ120 (W-739) and 2980 ñ120 (W-741) years ago, Barlovento is ap- proximately contemporary with late Period C and Period D of the Valdivia Phase, and if the analysis of its origin is correct represents the product of some 1300 years of divergent evolution. Turning toward the south, the earliest pottery is from the Guafiape complex on the north Peruvian coast (fig. 104), with a carbon-14 date of 4300 ñ200 years (L-122D) for a shell sample from upper Early Guafiape and 3100ñ200 years (L-122C) for a -¯ r~ ^xS^v t IS J _®~ f j - 4 JUNE AUG. ( 180ø MARCH-MAY AFTER PILOT CHART OF THE NORTH PACIFIC OCEAN NO. 1401 - NOVEMBER I960 INSERTS AFTER ATLAS OF PILOT CHARTS: SOUTH PACIFIC ft INDIAN OCEANS. PUB. NO. 107 HYDROGRAPHIC OFFICE, NAVY DEPT. WASHINGTON, O.C. Figure 103.-The northern Pacific Ocean, showing direction and speed of principal currents, paths of cyclonic storms and the great circle route between Kyushu, Japan dnd the Guayas coast of Ecuador. WHOLE VOLUME ORIGIN AND AFFILIATIONS 169 CARIBBEAN charcoal sample from the same level. In terms of the Ecuadorian sequence, the shell date seems a more accurate reflection of the age of Early Guafiape. Although this would place Early Guafiape con- temporary with the end of Valdivia Period A, the pottery is so simple that few similarities can be observed. Decoration is by punctation (Strong and Evans, 1952, fig. 48) or nicked and finger pressed rib (op. cit., figs. 45-46), both techniques represented in Valdivia Period A-B pottery. However, execution and motif are so different that considerable allowance must be made for readaptation and simplification during the process of diffusion if a connection is inferred. Similarities are somewhat greater in vessel shape, particularly in jar Form 18, with a rounded body and short insloping to vertical neck (Evans, Meggers and Estrada, 1959, fig. 81 top). Although these resemblances may seem to be a tenuous basis for inferring derivation of the earliest pottery of coastal Peru from Valdivia, the probability of a correlation is increased by failure to find anything earlier or more primitive in Peru in spite of a great deal of search. As Engel (translation, 1963b, p. 23) has recently noted: Up to the present time, we have been unsuccessful in efforts to find underneath the levels with well developed pottery, pottery of beginners. This does not of course prove that such pottery does not exist on the Peruvian coast, but the present state of knowledge finds us at the same point as Ford, Willey and other classic authors who twenty years ago reported a stage called "Cupisnique" in the Chicama Valley and "Guafiape" in the Viru Valley superimposed on the preceramic remains. An additional hint of possible Valdivia Phase influence on the Peruvian coast can be seen in similar- ities in face treatment between a carved gourd from preceramic levels at Huaca Prieta (Bird, 1962, figs. 7-9) and Valdivia Phase Palmar Incised stone figurines (pi. 118 a-b, g), and some incised and ex- cised designs on pottery (pi. 58 f, h-i). Dated at about 4000 years ago, this is a little later than the Valdivia face motifs, but as an isolated find it may not represent the oldest Peruvian occurrence. Returning to the evolving Valdivia culture, we find that Period C, estimated as beginning about 4000 years ago, is characterized by the introduction of several new and striking ceramic features, including Valdivia Nicked Broad-line Incised, Valdivia Rocker Figure 104.-Northwestern South America, showing location of pottery complexes carbon-14 dated between 5000-4000 years ago and their possible derivation from the early Valdivia Phase. Complexes with names in parentheses are of later date but incorporate early Valdivia decorated treatment suggesting they represent survivals. Routes are highly stylized and not intended to r eflectexact or exclusive paths of communication. 82ø 170 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Phases and Periods Pottery Techniques Coastal Puerto Mona- Barlovento Kotosh Ecuador Hormiga grillo 3500- 3800- Momil Ancon Tutish- 5000- 4800- 4000 BP 3000 BP 3000 BP 2900 BP cainyo 3000 BP 4500 BP Machalilla Phase 4000-3000 BP Tiered jar neck Stirrup spout Cylindrical spout Zoned hachure (fine-line border) 13 Bl El 3 El El 3 3 3 Bl El Valdivia Phase Period D: 3400-3000 BP Zoned hachure (broad-line border) Concave-walled carinated bowl B E) (?) H El El 3 El El El Period G: 4000-3400 BP Nicked broad-line incision Zoned circular punctate Ring Flange 12 El El 12 Bl El 12 El 3 3 El El 3 3 El Period A-B: 5000-4000 BP Multiple drag-and-jab punctate Finger grooving Excision Red-filled incision Broad-line incision Dot-ended incision Zoned dentate stamping Dot in ring 13 3 3 3 3 3 3 3 3 (?) 3 3 3 33333333 3 3 3 3 3 3 3 3 3 3 3 (?) 3 El 3 Figure 105.-Occurence of selected Valdivia and Machalilla Phase decorative techniques and vessel form elements in other Formative complexes of Colombia and Peru. Information is derived from the following sources: Puerto Hormiga, Reichel-Dolmatoff, 1961b and pers. commun.; Monagrillo, Willey and McGimsey, 1954; Barlovento, Reichel-Dolmatoff, 1955; Kotosh, Izumi and Sono, 1963; Momil, Reichel-Dolmatoff, G. and A., 1956; Ancon, Willey and Corbett, 1954; Tutishcainyo, Lathrap, 1962. Stamped and Valdivia Applique Fillet in decoration (fig. 55), and the angular form of cambered rim and the spout in vessel shape (fig. 56). Attempts to explain these by evolution from previously existing types are inconclusive. Valdivia Nicked Rib or Nubbin, which has Jomon parallels (pi. 189), may be earlier than Valdivia Applique Fillet, and if so the latter could represent a modification of rib size and treatment. It is a logical step from broad-line incision to the additional embellishment of this incision by punctation or scoring, and this innovation may have been made independently by the Valdivia potters. In vessel shape, an evolutionary sequence can be made from the Period A gently curved rim to increasing angularity, resulting in the Period C cambered form (fig. 56). However, the common variety of rocker stamping, in which the instrument is dragged producing a "negative" impression in a broad groove (pis. 107-111) and the spout at the base of the neck (pi. 69p) are without apparent antecedents in the Valdivia Phase. On the other hand, they and most of the other new Period C elements have parallels in Japan. The chronological and geographical position of these Valdivia Period C traits in Japanese material is more scattered than was the case with Valdivia Period A traits. Rocker stamping, identical in technique, occurs on Kyushu and Honshu and is dated on both islands as Early Jomon (fig. 96). An interesting detail of difference between the Jomon and Valdivia versions is that they tend to be mirror images. Jomon is more consistent than Valdivia in respect to direction of curvature, and examples of the Jomon variant appear among Valdivia examples more frequently than the Valdivia variant does in Jomon (pi. 179). Whether this decorative technique is restricted temporally to Early Jomon, or whether it survives sporadically into later times cannot be determined on the basis of existing information. Nicked incision closely resembling several Valdivia varieties (pi. 190 a-f) occurs sporadically in Early and Middle Jomon complexes from Kyushu (fig. 96) and is represented in the Katsusaka and Atamadai types from Middle Jomon of Honshu (Groot and Sinotd, 1952, pi. IV, XV). Frequently, an elongated horizontal zone delimited by nicked incision has a WHOLE VOLUME ORIGIN AND AFFILIATIONS 171 zigzag incised line running from end to end (op. cit., pis. Ill, 13; IV, 3, 6, 8; VIII, 2), exactly like that characterizing Valdivia Broad-line Incised, Motif 3, which is limited to Period C. One variant of Valdivia Nicked Broad-line Incised, in which the lower line is discontinuous or converges from two directions onto a low nubbin, is very similar to Honshu Middle Jomon style of execution (pi. 190a). Vessels with short slightly upslanting spouts are present in Japan from Early Jomon times, although they become common only in Late Jomon of Honshu (Kidder, 1957). In the earliest period, the spout is just below the rim of a wide-mouthed, deep jar, while the typical Late Jomon Horinouchi or Kasori form resembles a teapot, with the spout on the rounded shoulder (ibid.). A small patch of red slip on the upper interior edge and a few incisions on the corresponding portion of the exterior identify the Valdivia spout as from just below the junction of the decorated neck with the rounded shoulder on a Valdivia Incised jar, a position that coincides better with the later than with the earliest Jomon form of spouted vessel. In view of these Valdivia Period C innovations of Jomon-like affiliation, it is of interest to note a stone figurine head showing Buena Vista type characteristics from the Late Jomon Sakaizaki Shellmound, Kumamoto Prefecture, Kyushu (fig. 106; Esaka, 1960, p. 192). Eyes and nose are shown as depressions rather than slits, but the general treatment is much more like that of Valdivia figurines than of Jomon ones. The fact that most Valdivia Period C innovations can be explained by local evolution in the Valdivia Phase ceramic complex makes it seem prudent to account for rocker stamping and the spouted jar form as earlier introductions only becoming sufficiently common to be represented in the refuse at this later time. Rocker stamping, which in more recent times has often been used for decorating gourds, might have been initially employed for this perishable medium. Since only one spout was identified in the entire pottery sample from all Valdivia Phase sites, it cannot be confidently concluded that the form was restricted to Period C. About the beginning of Valdivia Period C, trade sherds of Machalilla Phase origin make their appear- ance in the refuse of Valdivia Phase sites, attesting to the arrival of this new complex on the central Ecua- dorian coast. Strong evidence of interrelations in the form of trade sherds of the other Phase in both kinds of sites continues throughout Valdivia Period C and traces persist into Period D (fig. 93). Detailed anal- ysis of the significance of this material for correlating the two Phases was given earlier (pp. 147-148) and need 767-841-65 13 not be repeated here. It is of interest, however, to comment on the lack of evidence that this apparently intensive intercourse had any acculturative effect on the ceramics of either Phase for some 600 years. Although large quantities of Machalilla Red Banded have been recovered from G-54, and although red slipping of the entire vessel surface was a common Figure 106.-Late Jomon stone figurine head from Sakaizaki Shell Mound, Kyushu bearing a slight resemblance to some examples of the Buena Vista type of the Valdivia Phase. (After Esaka, 1960, fig. 137.) Valdivia Phase practice, there is no indication that Valdivia Phase potters made any effort toward dupli- cating Machalilla Phase wide or narrow banding. Nor does there appear to be any modification in vessel shape, although Machalilla Phase stirrup-spout jars for example were, to judge from their frequency, much appreciated by the Valdivians. The fact that a considerably larger sample of Machalilla Double-line Incised was recorded from the Valdivia Phase site of G-54 than from all of the Machalilla Phase sites combined suggests that this may have been a type made principally for "the market" rather than for home consumption, although the alternative possi- bility that it reflects developments during Machalilla Period B, not well illustrated by the present seriated sequence, has been given preferential consideration. Some light is shed on this absence of acculturation by observations of the lack of receptivity on the part of modern Papago potters to potentially acculturative influences. Several women were shown films of pottery making by New Mexico Pueblo Indians, as well as a variety of archeological materials in the museum, and although they expressed great interest and asked numerous questions, they made no effort 172 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 to incorporate any of the techniques or motifs of decoration into their own work. The investigators (Fontana, Robinson, Cormack and Leavitt 1962, pp. 82-83) concluded: What is most interesting is that Papago potters who have been "trained" in traditional Papago pottery techniques adhere to those techniques even when shown others . . . Papago pottery changes, to be sure, and there is an extensive allowable range in forms, clays, tempers, firing fuels, and so on. But the changes, like the range in techniques and materials, all occur within a framework that is distinctively Papago and beyond whose bounds no true Papago pottery will go. That this kind of attitude is characteristic of small groups in general is suggested by the frequency with which situations paralleling the Valdivia-Machalilla exchange without acculturation have been reported by archeologists. If this is the normal attitude, the question is raised as to what factors bring about ac- culturation on the part of one or both groups between which trade relations exist. About the end of Valdivia Period C, the seemingly harmonious lives of the people of the Valdivia and Machalilla Phases were disrupted. A geographical dislocation apparently occurred, in which the Machalilla Phase expanded southward to the Santa Elena Peninsula, and the Valdivia Phase retreated or was pushed southeastward into the Posorja area (fig. 2). That some of the Valdivians may have continued down the coast is suggested by Lanning's report (1963, pp. 153 and 207) of sherds with Valdivia Applique Fillet decoration (not illustrated by Lanning) from the coast of Peru near the mouth of the Chira River. On the other hand, illustrated sherds said to be associated are not similar to other Valdivia Period D pottery types, so that if his identification is correct it seems most likely to reflect trade relations. No evidence of the Valdivia Phase was encountered during survey of the coast of El Oro Province in southern Ecuador (Estrada, Meggers and Evans, 1964), but in adjacent Peru Izumi and Terada (1965) report a few sherds similar to Valdivia Zoned Incised, a Period D decorated type. Whether the shift in geographical distribution of Machalilla and Valdivia Phase sites took place before or as a result of the arrival of immigrants from Mesoamerica will not be known until the transition between the Machalilla and Chorrera Phases is worked out in sufficient detail to make it clear where the earliest contact took place. In any case, the pottery of Valdivia Period D is strikingly different from that of Period C in spite of the continuation of several decorative types (fig. 55). Body wall thickness is markedly less, typically from 3-7 mm. in contrast to 6-12 mm. during Periods A-C. Several distinctive decorated types become popular, such as Punta Arenas Incised, Valdivia Carved and Valdivia Zoned Incised, while Valdivia Applique Fillet reaches maximum frequency. Vessels tend to be smaller and simpler in form, although most Period C shapes continue to be made (figs. 54, 56). When the attempt is made to account for this seemingly rapid transformation, most clues point toward the Machalilla Phase. Punta Arenas Incised is similar to Machalilla Incised in both motif and vessel shape (cf. figs. 21 and 80; pis. 26 and 144 a-v). The marked decline in body wall thickness brings Valdivia Period D pottery into the range characteristic of Machalilla Phase pottery, so that sherds of Machalilla Polished Plain are more difficult to distinguish from those of Valdivia Polished Plain from Period D than the latter are to separate from sherds of the same type from Period C. The near dis- appearance of figurines in Period D parallels more closely the slight emphasis on this form of religious expression in the Machalilla Phase than the situation during Valdivia Periods B and C. If these changes imply acculturation, the question arises as to what brought this about after several centuries of seeming resistance. A certain amount of ceramic deculturation might be accounted for by the relatively impoverished subsistence resources of the new environment, reflected in smaller size and lesser concentration of refuse deposits. Even if such a pos- sibility is accepted, however, it fails to explain the tendency for Valdivia Phase Period D pottery to diverge toward Machalilla Phase norms rather than to develop into a simpler version of the preceding Period C complex. The most probable explanation would seem to be intermarriage, which would permit Machalilla Phase women to introduce some of their ceramic traditions into a Valdivia Phase community. Over a period of time, this kind of situation could produce an amalgamation of the two complexes with results similar to that illustrated by Valdivia Period D ceramics. Identification of an adult female with the Machalilla Phase variety of skull deformation (see Appendix 2, p. 226) among the burials at the Valdi- via Period C site of G-l 15 (San Pablo) suggests that intermarriage did take place between the two popula- tions, since it is doubtful that a war victim would be buried in the village even if cultural indications existed of hostility between the two groups, which they do not. During the time that the Valdivia Phase was under- going transformation, the Machalilla Phase was changing in a different way, with results that laid the foundation for the whole future development of ab- original culture on the Ecuadorian coast. Before the Machalilla Phase was discovered, a number of Mesoamerican-like features were noted in the Chor- WHOLE VOLUME ORIGIN AND AFFILIATIONS 173 rera Phase and influence from that region was postu- lated to account for them (Evans and Meggers, 1957, pp. 243-245). Shortly thereafter, a site producing most of the relevant ceramic and nonceramic traits was described by Coe (1960) on the coast of Guate- mala. Sherds of Ocos and Conchas period types of Guatemala were matched with Chorrera and Tejar Phase sherds, illustrating the near identity of decora- tive techniques and motifs (op. cit., figs. 2-7). The keystone in the reconstruction was set in place with recognition of the Machalilla Phase. Adding Mach- alilla Phase elements to those present in Ocos and Conchas made it possible to account for nearly all the characteristic features of the Chorrera Phase (Meggers and Evans, 1962, table 1). There remains, however, the problem of the origin of the Machalilla Phase, and here we run into an almost blank wall. There is little possibility of deriving it from Peru, where work on early ceramic cultures has failed to turn up anything even remotely similar. The attempt to suggest linkages with Mesoamerica runs afoul of the relatively late date for the first pottery complexes recognized so far in that area. The arrival of the Machalilla Phase is estimated by correlation with Valdivia Phase carbon- 14 dates to have taken place about 4000 years ago. Nearly all early Formative ceramic complexes so far identified in Mesoamerica are 500 or more years more recent (Meggers and Evans, 1963, figs. 4-5), and are too sophisticated in form and decoration to represent the beginning of pottery making. An earlier and much simpler ceramic complex has recently been discovered in the Tehuacan Valley by Mac- Neish (1962, p. 36 and fig. 9), but in spite of its crudity and early date, between 2500 and 1500 B.C., he suggests that it "may not be the first modeled in Mexico but only an imitation of still earlier pottery (as yet unfound) in some other area" (MacNeish, 1964, p. 536). By analogy with the seafood subsistence pattern associated with Machalilla Phase pottery in Ecuador, the most likely place to search for such an early complex would seem to be along the coast. Some of the earliest Mesoamerican styles incorporate elements of decoration and vessel shape so similar to some of the Machalilla Phase pottery as to suggest a relationship. Decoration resembling Ayangue In- cised in technique and motif occurs in the Escalera and Francesca Phases in the region of Frailesca, southwest of Tuxtla Gutierrez, Chiapas (Navarrete, fig. 29), correlated with Chiapa III and IV; and in the Chiapilla Phase at Santa Cruz, on the Rio Grijalva east of Tuxtla Gutierrez (Sanders, fig. 23). At La Victoria, on the Guatemalan coast, Machalilla-like elements appear in the Conchas ceramic complex, including Ayangue Incised designs (Coe, 1961, fig. 35-i) and vessel shapes (op. cit., figs. 35i, 361), and Machalilla Embellished Shoulder bumps (op. cit., fig. 36n). An even larger number of similarities occurs in Tlatilco pottery, including Ayangue Incised-like designs (Pina Chan, 1958, vol. I, figs. 16f, 35j, 36 r, s, 47), red banding (op. cit., figs. 14, 15, 28 b-k, 29 d-f), tiered jar (op. cit., fig. 37n), angular shouldered jar (op. cit., fig. 37p), cylin- drical spouted jar (op. cit., fig. 39a'), stirrup spout jar (op. cit., figs. 43m, 44n), and carinated bowl forms (op. cit., figs. 21 i-1, p, 23 n-o; lam. 20). Unfortunately, carbon-14 dates so far obtained place all these complexes nearly 1500 years more recent than the inception of the Machalilla Phase, making it difficult to argue for a derivation from Mesoamerica. This relatively late position is supported by the stratigraphic occurrence of the Machalilla Phase-like elements, which are clearly introduced in the Chiapas and La Victoria sequences into earlier ceramic traditions of different origins (fig. 107). Farther north, the Tlatilco complex in the Valley of Mexico is not well pinned down either stratigraphically or in terms of absolute dates. Existing carbon-14 determi- nations place it earlier than the Chiapas complexes, and radical guess dates extend it backward to about 1400 B.C. (Pina Chan, 1963, fig. 4). This is still at least 500 years too recent, however, to account for the origin of the Machalilla Phase, and would argue for derivation of Tlatilco from the Machalilla Phase rather than the reverse. A similar interpretation might fit the occurrence of Machalilla Phase-like features in Conchas, since these survive into the Chor- rera Phase on the Ecuadorian coast, for which Ocos- Conchas connections have been established. Such possibilities, while fitting existing distributional and chronological evidence, fail to account for the appear- ance of the Machalilla Phase on the coast of Ecuador. Since the seafood subsistence pattern implies deriva- tion from a coastal environment, and since the coast of central and southern Mexico and of Central America is still incompletely known, it can be hoped that an ancestral complex may yet be found in Mesoamerica. Otherwise, it may be necessary to conclude that the Lord was not resting all the time on the seventh day. As a preliminary to suggesting possible interrelations between late Valdivia and Machalilla Phase pottery and other early ceramic complexes in the Andean Area, it is necessary to backtrack a few centuries in time and to examine the earliest materials from the central Peruvian highlands, represented by the Kotosh sequence with an initial carbon-14 date of between 3420ñ220 and 3800ñ110 years ago (Izumi, pers. commun.). As a result of intensive excavation at this site, the University of Tokyo archeologists have produced a sequence of six periods, defined in terms 174 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Figure 107.-Chronological position of certain Mesoamerican and Colombian complexes incorporating decorative techniques and motifs resembling Val- divia and Machalilla Phase (Ayangue Incised) types. Alignment is based on the following sources: Ocos, Coe, pers. commun.; Frailesca, Navarette, 1960; Santa Cruz, Sanders, 1961; Colombia, Reichel-Dolmatoff, 1955, 1956 and Angulo, 1963. of pottery types and architectural constructions. Un- fortunately, the attempt to correlate this sequence with other early South American sequences is frus- trated by their unique approach to pottery classifica- tion, which groups under a single ware a number of the decorative techniques isolated by other archeol- ogists, and by the assumption that complete vessels are the only reliable indicators of ceramic chronology, which causes rigid periodization of pottery types (Izumi and Sono, 1963, table 6). Tabulations of sherds by ware (op. cit., table 7), decorative technique (op. cit., table 9) and motif (op. cit., table 10) suggest a greater degree of continuity between periods and a longer time range for certain kinds of decoration than the complete vessels indicate. That continuity is a more accurate inference than discontinuity is also implied by the absence of significant innovations in architectural features during occupation of the site (op. cit., p. 65). Examination of elements diagnostic of the two ear- liest periods, Waira-jirca and Kotosh, leads to the interesting observation that many are also represented in the "Valdivia-derived" complexes of northern Co- lombia (fig. 105). Similarities include traits that can be traced to Valdivia Periods A-B, a few diagnostic of Period C, and others originating in Colombia, such as dot-ended incision (pi. 191 1-m), zoned dentate, dot-in-ring (pi. 191 1-m), and zoned parallel hachure (pi. 191 n-s). Along with these are a number of traits missing in the early Valdivia complex, but diagnostic of Valdivia Period D and Machalilla Phase pottery, including tiered (pi. 191 a-b) and stirrup- spout jar forms. The latter imply communication between the early Machalilla Phase and a Kotosh- like ceramic complex, an implication further sup- ported by the presence of three sherds at Machalilla Phase sites (pi. 158 g-i) reminiscent of zoned punc- tated pottery from Kotosh in vessel shape and decora- tive motif (pi. 191 f-k). Association in the central Peruvian highlands of these two groups of elements clearly separated chrono- logically on the Ecuadorian coast can be accounted for by postulating convergence of two movements, one southward along the intermontane valleys from the Caribbean coast of Colombia and the other inland from the coast of Ecuador (fig. 108). Although pot- tery complexes of sufficient antiquity to serve as "stepping stones" have so far not been reported in the intervening highland area (an area literally un- touched from the standpoint of modern archeological research), certain theoretical considerations tend to support such a reconstruction. Among the most significant contributions of the Tehuacan Archaeological-Botanical Project has been WHOLE VOLUME ORIGIN AND AFFILIATIONS 175  /1.3600IIIO A S clear definition of the steps from plant gathering to plant cultivation and increasingly settled life, and the establishment of a preceramic context for this transi- tion in Mesoamerica. Pottery making was added to the material culture inventory only after agriculture became the principal subsistence resource (MacNeish, 1962, fig. 9), and since Tehuacan ceramics are the earliest yet reported in Mesoamerica, it seems prob- able that pottery postdates rather well developed agriculture in other inland regions. Whether the same is true along the Mesoamerican coast cannot be determined on the basis of existing evidence. While all known coastal complexes are more recent, the relatively sedentary community pattern permitted by seafood subsistence appears to be a more congenial context for adoption of pottery than the wandering life of incipient agriculturalists or food gatherers. In South America, present evidence indicates that pottery making appeared first among seafood exploiters on the coast of Ecuador, and was trans- mitted in the beginning to groups in similar circum- stances along the coasts of Panama, Colombia, and possibly Peru (cf. Willey, 1958, p. 372). By analogy with Tehuacan evidence, inland diffusion of pottery making was dependent on prior acquisition of a staple crop permitting sedentary life. This being the case, it follows that the appearance of pottery in a highland situation implies previous (or simultaneous) acquisi- tion of domesticated plants of sufficient productivity to constitute a dependable food resource. The exist- ence of a preceramic building stage at Kotosh would fit such a theoretical model in two ways: 1) by attest- ing to the existence of sedentary populations in the central Peruvian highlands before the introduction of pottery; 2) by permitting projection of a similar situation of "subsistence readiness" toward the north, until a source for certain elements of the ceramic complex is reached. The feasibility of the intermontane route for com- munication is well attested from Inca times to the present. Within Colombia, the Magdalena-Cauca river system, navigable for 900 miles, has been the major route of transportation for centuries (Herring, 1964, p. 8), taking second place only in the last few years to road and railroad, which follow most of its course. The Pan American Highway continues south- ward through the intermontane valleys of Ecuador Figure 108.-Northwestern South America, showing location of pottery complexes carbon-14 dated between 4000-3400 years ago and possible routes by which they were spread. Dotted line passes through highland zones where no complexes of sufficient antiquity have been reported to provide corroboration for the hypothesis of cultural exchange. Routes are highly stylized and not intended to reflect exact or exclusive paths of communication. PACIFIC 176 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME and Peru, which have served as the principal route of communications since Colonial times (Morales y Eloy, 1942, pp. 34-5). Not only does this zone present no major geographical obstacles, but it is characterized by botanical continuity in the form of tropical montane forest (Eyre, 1963, map 6). Lack of pronounced regional diversity in soil or climate, reflected in the similarity of wild vegetation, would have facilitated dispersal of cultivated plants, and perhaps also ideas or experimental methods of incipient cultivation, from one part of this zone to another, in much the same manner as in Mesoamerica. The suggestion that Colombia may have played a major role in New World plant domestication comes not only from its environmental similarity to the region definable as the Mesoamerican hearth (Meggers and Evans, 1963, fig. 18), but also from evidence (admittedly meager) that manioc utilization decreases in antiquity the farther one moves from this center toward the east, southeast or south. Neither of the principal South American staples-manioc and po- tato-have been found in early archeological sites in Mesoamerica, suggesting that they are of South American origin. Sweet manioc appears on the coast of Peru only after 1000 B.C., with pottery associations that can be traced to a highland ancestor resembling Kotosh Chavm, the third subdivision in the Kotosh sequence (figs. 105, 109). A further kind of information bearing on this hypothesis is the variety of plants cultivated in the Colombian highland valleys at the time of the Con- quest. According to Reichel-Dolmatoff (1961a, p. 85): In practically all regions maize was the most important food plant. Next in importance followed the root crops such as sweet manioc, racacha, sweet-potato, and yam. The highland dwellers grew at least two local types of potato and, besides, quinoa, ulluco, oca, topinambur, and cubios, a tuberous nasturtium. Other root crops were achira, yautia, and groundnuts; kidney and lima beans were widely cultivatad. A large number of more or less cultivated fruits included avocado pears, papaya, guava, custard apple, pineapple, soursop and others. The peach palm provided an important and very common source of food, and squashes, pepino, tomato, caimito, hogplums, and peppers were widely used. The greatest number of species was grown on the fertile temperate slopes of the Cordilleras. While a number of these plants undoubtedly reached Colombia after being domesticated elsewhere, others are highland or temperate varieties for which an Andean origin seems probable. In view of the argument presented by Lathrap (1963) for derivation of the Machalilla Phase from Tutishcainyo, the earliest phase at Yarinacocha in eastern Peru, examination of the possibility of an Amazonian origin is necessary. Unfortunately, such an hypothesis can be sustained only by ignoring the chronological position of Machalilla-like traits in the Yarinacocha sequence, and by attributing missing Machalilla Phase elements to local invention on the Ecuadorian coast. Two important elements shared by Early Tutishcainyo and Kotosh Waira-jirca-broad zoned parallel hachure and dot-ended incision-are unknown in the Machalilla Phase, while the early Machalilla Phase diagnostics-red banded decoration and stirrup spout jar form-have never been reported in Tutishcainyo ceramics. A further clue to relative chronological position of the two complexes comes from the appearance in Late Tutishcainyo of the tiered jar form introduced into the Kotosh sequence during the Kotosh period along with stirrup spouts, both characteristic of the beginning of the Machalilla Phase on the Ecuadorian coast. Numerous illustra- tions of decoration and vessel shape characteristic of Early and Late Tutishcainyo provided by Lathrap (1962, figs. 20-52) show no similarities in either category that might suggest any direct connection with the Machalilla Phase, but instead numerous features indicating derivation from a complex like that represented at Kotosh (figs. 105, 108). Several other bits of evidence can be fitted into this hypothetical reconstruction of the spread of pottery making in northwestern South America. One is the appearance in Cupica Phase I on the north Pacific coast of Colombia (fig. 104) of several early Forma- tive Ecuadorian ceramic features, such as jars with Valdivia Modeled decoration around the shoulder (Reichel-Dolmatoff, G. and A., 1961, Lam. IV, 1-2), and Machalilla Embellished Shoulder gashes (op. cit., Lam. IV, 5). The context suggests a considerably later date for the Cupica I complex, but the geograph- ical location of the site along the route by which an early movement could have passed permits inter- pretation of these resemblances as survivals. A simi- lar explanation may apply to jars with Valdivia Modeled decoration from Paita Phases C and D of the north Peruvian coast (Lanning, 1963, figs. 3 and 8), correlated by Lanning with Machalilla and Chorrera Phases (op. cit., table 22), both subsequent to dis- appearance of this type of decoration in the Valdivia Phase. On the eastern side of the Andes, pottery from the Yasuni site on the Rio Napo shares broad zoned hachure and the flanged bowl form with Kotosh and Tutishcainyo, suggesting that it may represent an offshoot from a member of this tradition located in highland Ecuador, along the postulated route of movement from north to south (fig. 108). In the highlands of Colombia, the presence of dentate stamp- ing on San Agustin pottery can be interpreted as another survival that may serve as a clue to the past. WHOLE VOLUME ORIGIN AND AFFILIATIONS 177 If this general picture is valid, the complex of ceramic features presented by Momil on the north coast of Colombia (fig. 105) is of considerable in- terest. Unfortunately, no carbon-14 dates have been obtained for the Momil site, but dates from earlier complexes such as Malambo in the nearby area sug- gest that an estimate of 1000 B.C. for its inception may be too early. In addition to elements that align it with the earlier north Colombian complexes like Puerto Hormiga, Barlovento, and Malambo (Angulo, 1963, p. 58), Momil includes a style of incision closely resembling Ayangue Incised (Reichel-Dolmatoff, G. and A., 1956, Lam. XVIII, 2, 4, 10) and duplicating the motif noted earlier as shared by the Machalilla Phase, Conchas of coastal Guatemala, and Escalera- Francesca and Chiapilla of southern Mexico. Typical Machalilla Phase bowl rim shapes occur in Momil I, where they are associated with incised decoration (op. cit., p. 146 and fig. 8 LL, MM, NN) or embel- lished shoulder treatment (op. cit., Lam. VI-8). However, in Momil I, these traits are associated with polypod and annular bases, polychrome painting, negative painting, and other elements suggesting a later chronological placement for the entire complex. Momil II marks the first reported occurrence in northern Colombia of the so-called "basal flange" (op. cit., fig. 6), also represented at Kotosh and Yarinacocha. To complete the confusion, small tet- rapod feet similar to the Valdivia form occur in Momil I (op. cit., p. 225 and fig. 12-7), along with zoned dentate, in which zones may or may not be outlined by incision (op. cit., Lam. VIII-IX). Such a combination of traits might be expected if Momil is the consequence of developments of the kind postulated to have occurred in northern Colom- bia. Although too late to account for the origin of the Kotosh assemblage, it suggests that several earlier traits missing in Barlovento (such as zoned dentate) may have survived elsewhere in the region. The stirrup spout vessels in the late coastal Colombian complex of Santa Marta (Bennett, 1944, p. 106 and pi. 12-1) may constitute local persistence of an- other early trait. Stirrup spouts are also represented in the Playa de los Muertos complex, which in other respects Figure 109.-Northwestern South America, showing location of selected pottery complexes carbon-14 dated between 3400-3000 years ago and possible routes of communication between them. San Agustin, although of more recent date, incorporates decora- tive techniques aligning it with earlier complexes at Momil and Kotosh, suggesting the possibility of earlier occurrence of these traits in the Colombian highlands. Routes are highly stylized and not intended to reflect exact or exclusive paths of communi- cation. 62ø 178 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 appears to be closely related to Tlatilco (Porter, 1953, pp. 66, 89). It is unfortunate that the Playa de los Muertos sequence is not well defined, and that chronological placement is based on tentative cor- relations with other early styles. However, its geographical location intermediate between central Mexico and northern South America points once again to Central America as a fruitful area for more intensive search for Machalilla Phase origins. The marginality of this region to centers of incipient agriculturism is an additional factor in its favor, in view of the probable preagricultural subsistence pattern of the Machalilla Phase. Although most of the ceramic evidence points to a diffusion route into the highlands of southern Ecuador and Peru toward the end of the Ecuadorian early Formative Period, the introduction of fronto-vertico- occipital or tabular erecta flattening of the skull on the Peruvian coast in late preceramic times suggests that communication with that area also occurred. Munizaga (pp. 230-232) interprets the distributional and chronological contexts as reflecting differential diffusion of some of the constituent elements south- ward from the coast of Ecuador. This places skull deformation in a class with such other cultural traits as cotton and maize, which are apparently derived from areas where pottery was being made. If so, the failure of pottery making to be adopted on the coast of Peru at any of these successive opportu- nities is a situation for which no satisfactory explana- tion seems to exist at the present time. The end of the early Formative Period on the coast of Ecuador is signaled by abandonment of the sea- oriented subsistence economy for a new way of life that permitted sedentary occupation along the river banks. The existence of early Chorrera Phase sites on the Rio Daule and Rio Babahoyo suggests that this inland flow was a relatively rapid one. The inference that it reflects the introduction of maize is based on several kinds of evidence. First, maize is associated with the coastal Guatemalan Ocos and Conchas Phases presenting the closest ceramic similarities to Chorrera. Second, maize becomes a major subsistence plant on the Peruvian coast about 800 B.C. in associa- tion with a ceramic complex that can be linked with the Chorrera Phase. Third, inland expansion of a sedentary community pattern depends on substitution of a reliable terrestrial staple for seafood, a role that can only have been played by agriculture. If infer- ences regarding the existence of early sedentary, nonceramic communities in the highland valleys are correct, failure of domesticates from this region to be adopted on the coast might be explained by their poor adaptation to lowland conditions and consequent inability to compete with seafood as the primary source of subsistence. Reconstruction of the origin, interrelations and ramifications of the Valdivia and Machalilla Phases has moved us beyond the boundaries of available evidence into the realm of hypothesis. Although we hope that some of the interpretations will be confirmed by future work, others will undoubtedly be altered. If calling attention to the potential importance of Colombia and Central America in the origin and dissemination of certain South American cultural complexes serves to stimulate fieldwork in these areas, our efforts will have been well spent regardless of whether or not the hypotheses are upheld. LITERATURE CITED Angulo Valdes, Carlos 1963. Cultural development in Colombia. In Aboriginal cultural development in Latin America: An interpretative review, edited by Betty J. Meggers and Clifford Evans. Smithsonian Misc. Coll., vol. 146, no. 1, pp. 55-66. Bennett, Wendell C. 1944. Archeological regions of Colombia: a ceramic survey. Yale Univ. Publ. in Anthropology, no. 30. Bird, Junius 1938. Antiquity and migrations of the early inhabitants of Patagonia. The Geographical Review, vol. 28, pp. 250-275. 1948. Preceramic cultures in Chicama and Viru. In A reappraisal of Peruvian archaeology, assembled by Wendell C. Bennett. Society for American Archaeology Memoir 4, pp. 21-28. 1962. Art and life in old Peru: an exhibition. Curator, vol. 2, pp. 147-210. Bushnell, G. H. S. 1951. The archaeology of the Santa Elena Peninsula in South-west Ecuador. Cambridge (England). Coe, Michael D. 1960. Archeological linkages with North and South America at La Victoria, Guatemala. American Anthropologist, vol. 62, pp. 363-393. 1961. La Victoria, an early site on the Pacific coast of Guatemala. Papers of the Peabody Museum of Archaeology and Ethnology, Harvard Univ., vol. 53. Damon, Paul E.; Long, Austin; and Sigalove, Joel J. 1963. Arizona radiocarbon dates IV. Radiocarbon, vol. 5, pp. 283-301. Disselhoff, H. D. 1949. Grabungen und Funde im Canton Sta. Elena (Ecuador). El Mexico Antiguo, vol. 7, pp. 343-410. Emperaire, Jose, and Laming, Annette 1961. Les gisements des lies Englefield et Vivian dans la mer d'Otway, Patagonie Australe. Journ. de la Societe des Americanistes, new series, vol. 50, pp. 7-75. Engel, Frederic 1963a. A preceramic settlement on the central coast of Peru: Asia, Unit 1. Transactions of the American Philosophical Society, new series, vol. 53, part 3, Philadelphia. 1963b. Notes relatives a des explorations archeologiques a Paracas et sur la cote sud du Perou. Travaux de l'lnstitut Frangais d'Etudes Andines, tome 9, pp. 1-72, Paris. Esaka, Teruya 1960. Figurines [in Japanese]. Tokyo. 1962. Female figurines from prehistoric remains at Kamikuroiwa, Ehime Pref., Shikoku [in Japanese]. Scientific Asahi, vol. 12, pp. 102-105. Estrada, Emilio 1956. Valdivia, un sitio arqueologico formativo en la costa de la Provincia del Guayas, Ecua- dor. Publ. del Museo Victor Emilio Estrada, no. 1, Guayaquil. 1958. Las culturas pre-clasicas, formativas o arcaicas del Ecuador. Publ. del Museo Victor Emilio Estrada, no. 5, Guayaquil. 767-841-65 14 179 180 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Estrada Emilio-Continued 1961. Nuevos elementos en la cultura Valdivia: sus posibles contactos transpacificos. Publ. del Sub-Comite Ecuatoriano de Antropologia, Guayaquil. 1962. Arqueologia de Manabi central. Publ. del Museo Victor Emilio Estrada, no. 7, Guayaquil. Estrada, Emilio; Meggers, Betty J.; and Evans, Clifford. 1964. The Jambeli culture of south coastal Ecuador. Proc. of the U.S. Nat. Mus., vol. 115, no. 3492. Evans, Clifford, and Meggers, Betty J. 1957. Formative period cultures in the Guayas Basin, coastal Ecuador. American Antiquity, vol. 22, pp. 235-247. Evans, Clifford; Meggers, Betty J.; and Estrada, Emilio. 1959. Cultura Valdivia. Publ. del Museo Victor Emilio Estrada, no. 6, Guayaquil. Eyre, S. R. 1963. Vegetation and soils. London. Ferdon, Edwin N., Jr. 1950. Studies in Ecuadorian geography. Monographs of the School of American Research, no. 15, Santa Fe. Fontana, Bernard L.; Robinson, William J.; Cormack, Charles W.; and Leavitt, Ernest E., Jr. 1962. Papago Indian pottery. American Ethnol. Soc. Monograph, Seattle. Ford, James A. 1954. On the concept of types. American Anthropologist, vol. 56, pp. 42-54. 1962. A quantitative method for deriving cultural chronology. Pan American Union Tech- nical Manual 1, Washington, D.C. Ford, James A., and Webb, Clarence H. 1956. Poverty Point, a late archaic site in Louisiana. American Mus. of Nat. Hist., An- thropological Papers, vol. 46, part 1. Friedman, Irving; Smith, Robert L.; Evans, Clifford; and Meggers, Betty J. 1960. A new dating method using obsidian. American Antiquity, vol. 25, pp. 476-537. Gifford, E. W. 1947. Californian shell artifacts. Anthropological Records 9:1, Berkeley and Los Angeles. Grant, Verne 1963. The origin of adaptations. New York. Groot, Gerard J., and Sinoto, Yosihiko H. 1952. The shell mound of Ubayama. Nipponica, First Series; Archaeologia Nipponica, vol. II. Archaeological Institute of Japan, Konodai, Itikawa City. Herring, Hubert 1964. A history of Latin America. Second edition, revised, New York. Hydrographig Office, United States Government 1960. Pilot chart of the north Pacific Ocean, no. 1401. Izumi, Seiichi, and Sono, Toshihiko 1963. Andes 2: Excavations at Kotosh, Peru, 1960. Tokyo. Izumi, Seiichi, and Terada, Kazuo 1965. Excavations at Pechiche and Garbanzal, Tumbes Valley, Peru. Univ. of Tokyo Scientific Expeditions to the Andes, Report no. 3, Tokyo. Kidder II, Alfred; Lumbreras, Luis G.; and Smith, David B. 1963. Cultural development in the central Andes-Peru and Bolivia. In Aboriginal Cultural Development in Latin America: An Interpretative Review, edited by Betty J. Meggers and Clifford Evans. Smithsonian Misc. Coll., vol. 146, no. 1, pp. 89-101. Kidder, J. Edward, Jr. 1957. The Jomon pottery of Japan. Artibus Asiae, Supplementum XVII, Institute of Fine Arts, New York. Komatsu, Isao 1962. The Japanese people: origins of the people and the language. Series of Japanese Life and Culture, vol. 1. Soc. for International Cultural Relations, Tokyo. Kroeber, Alfred L. 1948. Anthropology. New York. whole volume LITERATURE CITED 181 Kroeber, Alfred L., and Kluckhohn, Clyde 1952. Culture: a critical review of concepts and definitions. Papers Peabody Mus. American Archaeology and Ethnology, vol. 47. no. 1, Cambridge, Mass. Lanning, Edward P. 1963. A ceramic sequence for the Piura and Chira coast, north Peru. Univ. of California Publ. in American Archaeology and Ethnology, vol. 46, no. 2, pp. 135-284. Lathrap, Donald W. 1962. Yarinacocha: stratigraphic excavations in the Peruvian montafia. Unpublished Ph. D. dissertation, Harvard Univ. 1963. Possible affiliations of the Machalilla complex of coastal Ecuador. American Antiquity, vol. 29, pp. 239-241. Libby, W. F. )1963. Accuracy of radiocarbon dates. Science, vol. 140, pp. 278-280. Linton, Ralph 1955. The tree of culture. New York. MagNeish, Richard S. 1962. Second annual report of the Tehuacan Archaeological-Botanical Project. Robert S. Peabody Foundation for Archaeology, Tehuacan Archaeological-Botanical Project Reports, no. 2, Andover, Mass. 1964. Ancient Mesoamerican civilization. Science, vol. 143, no. 3606, pp. 531-537. Meggers, Betty J. 1964. North and South American cultural connections and convergences. In Prehistoric Man in the New World, edited by Jesse D. Jennings and Edward Norbeck, pp. 511-526. Meggers, Betty J., and Evans, Clifford 1958. Identificagao das areas culturais e dos tipos de cultura na base da ceramica das jazidas arqueologicas. Archivos do Museu Nacional, vol. 46, pp. 9-32, Rio de Janeiro. 1962. The Machalilla culture: an early Formative complex on the Ecuadorian coast. American Antiquity, vol. 28, pp. 186-192. Meggers, Betty J., and Evans, Clifford, editors 1963. Aboriginal cultural development in Latin America: An interpretative review. Smith- sonian Misc. Coll., vol. 146, no. 1. Miller, E. V. 1959. Agricultural Ecuador. The Geographical Review, vol. 49, pp. 183-207. Morales y Eloy, Juan |1942. Ecuador: Atlas historico-geografico. Ministerio de Relaciones Exteriores, Quito. Murphy, Robert Cushman 1939. The littoral of Pacific Colombia and Ecuador. The Geographical Review, vol. 29, pp. 1-33. Navarrete, Carlos. 1960. Archeological explorations in the region of the Frailesca, Chiapas, Mexico. Papers of the New World Archaeological Foundation, no. 7. Oschinsky, L.; Gall, P.; MacDonald, J.; Niemann, L.; Spence, M.; and Wilson, S. 1964. Parallelism, homology and homoplasy in relation to hominid taxonomy and the origin of Homo sapiens. Anthropologica, new series, vol. 6, pp. 105-117. Pina Chan, Roman 1958. Tlatilco. 2 vols. Instituto Nacional de Antropologia e Historia, Serie Investigaciones 1, Mexico. 1963. Cultural development in central Mesoamerica. In Aboriginal Cultural Development Latin America: An Interpretative Review, edited by Betty J. Meggers and Clifford Evans. Smithsonian Misc. Coll., vol. 146, no. 1, pp. 17-26. Ploszajski, J. A. 1963. A history of ships and boats of Japan. Kultuurpatronen, Bulletin van het Etnografisch Museum Delft (Nederland), Deel 5-6, pp. 85-120. Porter, Muriel Noœ 1953. Tlatilco and the pre-classic cultures of the New World. Viking Fund Publications in Anthropology, no. 19. Reichel-Dolmatoff, Gerardo 1955. Excavaciones en los conchales de la costa de Barlovento. Revista Colombiana de Antropologia, vol. 4, pp. 247-272. 182 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME l Reichel-Dolmatoff, Gerardo-Continued 1961a. The agricultural basis of the sub-Andean chiefdoms of Colombia. In The Evolution of Horticultural Systems in Native South America: Causes and Consequences, edited by Johannes Wilbert, pp. 83-100. Sociedad de Ciencias Naturales La Salle, Caracas. 1961b. Puerto Hormiga: un complejo prehistorico marginal de Colombia (nota preliminar). Revista Colombiana de Antropologia, vol. 10, pp. 347-354. PvEichel-Dolmatoff, Gerardo and Alicia 1956. Momil: excavaciones en el Sinu. Revista Colombiana de Antropologia, vol. 5, pp. 109-333. 1961. Investigaciones arqueologicas en la costa pacifica de Colombia. I. El sitio de Cupica. Revista Colombiana de Antropologia, vol. 10, pp. 237-330. Rubin, Meyer, and Alexander, Corrinne 1960. U.S. Geological Survey radiocarbon dates V. American Journ. of Science Radio- carbon Supplement, vol. 2, pp. 129-185. Sanders, William T. 1961. Ceramic stratigraphy at Santa Cruz, Chiapas, Mexico. Papers of the New World Archaeological Foundation, no. 13. Schumacher, Paul 1960. The manufacture of shell fish-hooks by the early inhabitants of the Santa Barbara Channel Islands. Reports of the Univ. of California Archaeological Survey, no. 50, Papers on California Archaeology, no. 83. Silva, Fernando Altenfelder, and Meggers, Betty J. 1963. Cultural development in Brazil. In Aboriginal Cultural Development in Latin America: An Interpretive Review, edited by Betty J. Meggers and Clifford Evans, Smithsonian Misc. Coll., vol. 146, no. 1, pp. 119-129. Simpson, George Gaylord 1961. Principles of animal taxonomy. Columbia Biological Series, no. 20, New York. 1964. This view of life: the world of an evolutionist. New York. Sittig, Otto 1896. Compulsory migrations in the Pacific Ocean. Smithsonian Institution Annual Report to July 1895, pp. 519-535. Smith, H. S. 1964. Egypt and C-14 dating. Antiquity, vol. 38, pp. 32-37. Stirling, Matthew W., and Stirling, Marion 1963. Tarqui, and early site in Manabi Province, Ecuador. Bur. of American Ethnology Bull. 186, pp. 1-28. Anthropological Papers, no. 63. Strong, William Duncan, and Evans, Clifford 1952. Cultural stratigraphy in the Viru Valley, northern Peru: The Formative and Florescent epochs. Columbia Studies in Archeology and Ethnology, vol. IV, New York. Vogt, Evon Z. 1960. On the concepts of structure and process in cultural anthropology. American Anthro- pologist, vol. 62, pp. 18-33. 1964. The genetic model and Maya cultural development. In Desarrollo Cultural de los Mayas, edited by Evon Z. Vogt and Alberto Ruz L., pp. 9-48. Publication Especial del Seminario de Cultura Maya, Mexico. Willey, Gordon R. 1958. Estimated correlations and dating of South and Central American culture sequences. American Antiquity, vol. 23, pp. 353-378. Willey, Gordon R., and Corbett, John M. 1954. Early Ancon and Early Supe culture. Columbia Series in Archeology and Ethnology, vol. 3, New York. Willey, Gordon R., and McGimsey, Charles R. 1954. The Monagrillo culture of Panama. Papers of the Peabody Museum of Archaeology and Ethnology, Harvard Univ., vol. 49, no. 2, Cambridge, Mass. Zevallos Menendez, Carlos, and Holm, Olaf 1960. Excavaciones arqueologicas en San Pablo: Informe preliminar. Guayaquil. Appendix 1 TABLES 1-17 Table 1.-Frequency of species of mollusks in levels in G-31, Cuts A, F and H, of the Valdivia Phase {fig. 9) a a 5 Levels in seriated ocardi gosa ocardi brica i costat i ulosa '5. a cten aris sJ-O sequence "8 e a ~3 -~ >. "a 3 -ø ??2 ?& ae tsch hili by. *gœ ?3 "8> j S fi gc*  œc/D 1 3® C -a Ph is ?®,<*> &. S C/3 ~ 3 < Periods tj "3 TJ ^ TJ ti T| cq Cut Level (cm.) No. % No. % No. % No. % No. % No. % No. % No. % H 0-20 248 43.8 1 0.2 _ _ 10 1.7 8 1.4 _ _ 6 1.0 _ 20-40 521 47.3 14 1.2 - - 15 1.3 16 1.4 - - 2 0.2 1 0.1 C 40-60 1,101 51.3 20 0.9 1 0.1 29 1.3 24 1. 1 4 0.2 2 0.1 1 0.1 60-80 776 49.3 10 0.6 1 0. 1 36 2.2 24 1. 5 1 0. 1 1 0.1 - - 80-100 1, 196 57.5 15 0.7 3 0. 1 40 2.0 14 0.6 1 0. 1 1 0. 1 2 0.1 100-120 1,286 65.0 7 0.3 6 0.3 72 3.6 7 0. 3 1 0. 1 4 0.2 1 0.1 F 0-20 59 67.9 - - 1 1.1 1 1. 1 1 1. 1 - - - - - - B 20-40 203 61.8 1 0.3 1 0.3 2 0.6 - - 1 0.3 - - - - H 120-130 1,176 64.7 4 0.2 5 0.3 64 3.5 15 0.8 - - 5 0.3 - - F 40-60 126 77.0 1 0.6 - - 2 1.2 H 130-140 1,307 50.8 7 0.3 9 0.4 116 4.5 22 0.9 1 0.1 12 0.5 - - A 160-180 405 62.2 1 0.2 - _ 76 11.6 6 0.9 - - 39 6.0 - _ A 180-200 484 64.7 4 0.5 2 0.3 77 10.4 5 0.7 - - 42 5.6 - - 200-220 214 53.8 5 1.2 3 0.8 37 9.3 1 0.2 - - 1 0.2 - - 220-240 62 51.7 2 1.7 - - 9 7.6 1 0.8 " " 2 1.7 - - 183 184 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Table 1.-Frequency of species of mollusks in levels in G-31, Cuts A, F and H, of the Valdivia Phase {fig. 9)-Continued Levels in seriated sequence Cerethidea valida C.B. Adams Columbella major Sby. Conus purpurascens Brod. Cypraea robertsi Hidalgo C. arabicula Lamarck Chama echinata Brod. Cerithium adjustum Kiener Dosinia dunkeri Philippi Fissurella oirescens Sby. "O Peric Cut Level (cm.) No. % No. % No. % No. % No. % No. % No. % No. % H 0-20 95 16.8 11 1.9 7 1.2 2 0.4 _ _ 1 0.2 38 6.7 4 0.7 20-40 160 14.5 23 2.1 4 0.4 8 0.7 - - - - 62 5.6 9 0.8 C 40-60 344 16.1 28 1.3 6 0.3 3 0.1 15 0.7 - - 98 4.5 9 0.4 60-80 308 19.7 26 1.6 3 0.2 2 0.1 6 0.4 - - 55 3.5 8 0.5 80-100 336 16.6 30 1.5 8 0.4 3 0.1 5 0.2 - - 64 3.2 7 0.3 100-120 249 12.6 21 1.1 2 0.1 7 0.3 1 0. 1 2 0.1 40 2.0 3 0. 1 F 0-20 2 2.3 3 3.5 1 1. 1 - - - - - - 2 2.3 1 1. 1 20-40 20 6.1 15 4.5 12 3.7 7 2.1 - - 2 0.6 10 3.0 - - B H 120-130 300 16.5 15 0.8 1 0.1 6 0.3 1 0. 1 - - 33 1.8 1 0.1 F 40-60 2 1.2 6 3.7 11 6.7 1 0.6 - - - - - - - - H 130-140 748 29. 1 18 0.7 5 0.2 4 0.2 - - 2 0. 1 57 2.2 4 0.2 A 160-180 34 5.2 1 0.2 _ - - - - - - - 42 6.4 - - A 180-200 37 4.9 2 0.3 - - 2 0.3 - - - - 54 7.2 - - 200-220 59 14.8 - - 1 0.2 - - - - - - 28 7.0 2 0.5 220-240 22 18.3 6 5.0 - - - " 3 2.5 ?3 Levels in seriated sequence Glycymeris inaequalis Sby. Mactra velata Philippi Natica chemnitzil unifasciata Lamarck Nodipecten subnodosus Sby. Oliva peruviana Lamarck Ostrea iridescens Cpr. Ostrea fischeri Dall Panamicorbula inflata C. B. Adams Pliyllonatus regius Swainson, P. radix Gemlin Peric Cut Level (cm.) No. % No. % No. % No. % No. % No. % No. % No. % No. % C H 0-20 20-40 40-60 60-80 80-100 1 3 1 0. 1 0.2 0. 1 2 1 0.2 0.1 9 3 4 0.8 0.1 0.3 1 3 1 1 0.1 0. 1 0.1 0.1 2 5 6 0.1 0. 3 0.2 17 49 94 59 48 3.0 4.4 4.3 3.7 2.3 8 12 4 2 2 1.4 1. 1 0.2 0. 1 0.1 5 3 3 4 0.5 0. 1 0.2 0.2 2 6 16 3 3 0.4 0.5 0.7 0.2 0.1 B F H F H 100-120 0-20 20-40 120-130 40-60 130-140 1 2 0. 1 0.6 3 2 0.1 0.1 5 3 2 8 0.2 0.9 0.1 0.3 2 3 0.1 0.2 2 3 0.1 0. 1 29 2 11 31 1 40 1.4 2.3 3.4 1.7 0.6 1.5 1 1 2 4 0. 1 0. 1 1.2 0.2 1 4 1 0. 1 0.2 0.6 4 1 2 3 0.2 0.3 0. 1 0.1 A A 160-180 180-200 200-220 220-240 - - - - - - 5 0.7 - - 12 16 16 1 1.8 2. 1 4.0 0.8 2 0.5 - - 2 2 0.3 0.5 WHOLE VOLUME APPENDIX 1 185 Table 1.-Frequency of sp zcies of mollusks in levels in G-31, Cuts A, F and H, of the Valdivia Phase {fig. 9)- -Continued 3 Levels in seriated sequence Pinctada mazatlanica Hanley Pilar {Lamelliconcha) concinnus Sby. II o 00 <*> Strombus granulatus Wood, 6\ galeatus Swainson 4 J & J? s t-< Peri< Cut Level (cm.) No. % No. % No. % No. % No. % No. % No. % No. % C H 0-20 20-40 40-60 60-80 80-100 2 4 9 4 11 0.4 0.4 0.4 0.3 0.5 2 11 2 2 0.2 0.5 0.1 0.1 15 28 28 10 19 2.7 2.5 1.3 0.6 0.9 26 47 76 62 53 4.5 4.2 3.5 3.9 2.6 17 26 74 39 79 3.0 2.3 3.4 2.4 3.9 6 11 5 6 15 1.0 1.0 0.2 0.4 0.7 1 3 6 5 3 0.2 0.3 0.3 0.3 0.1 6 14 61 39 48 1.0 1.2 2.8 2.6 2.3 B F H F H 100-120 0-20 20-40 120-130 40-60 130-140 7 2 2 13 2 3 0.3 2.3 0.6 0.7 1.2 0.1 1 2 0.1 0.1 3 2 4 6 7 0.1 2.3 1.2 0.3 0.3 48 3 10 35 6 66 2.4 3.5 3.0 1.9 3.6 2.5 50 5 16 7 3 16 2.5 5.8 4.9 0.4 1.8 0.6 31 2 31 4 1.5 0.6 1.7 0.2 5 8 0.2 0.3 51 3 27 34 2.5 0.9 1.4 1.3 A A 160-180 180-200 200-220 220-240 1 1 4 0.2 0.1 1.0 1 0.2 1 0.2 3 3 6 7 0.5 0.4 1.5 5.8 7 1 1.0 0. 1 6 3 1 0.9 0.8 0.8 2 1.7 11 4 5 1.7 0.5 1.2 3 iger S | Levels in seriated sequence Thais crassa Blainu Trivia radia Lamarck Turbo squarr Reeve Triumphis distorta Wood Trachycardiu senticosum Sby. Vasum muricatum caestum Brod. Unclassifiec Totals 'u Hi Cut Level (cm.) No. % No. % No. % No. % No. % No. % No. % No. % H 0-20 13 2.3 1 0.2 _ _ 13 2.3 _ _ 7 1.2 2 0.4 567 100.0 20-40 21 1.9 1 0.1 5 0.5 15 1.3 1 0.1 4 0.4 4 0.4 1,105 100.0 C 40-60 25 1.1 - - 12 0.5 18 0.8 3 0.1 1 0. 1 17 0.8 2,157 100.0 60-80 23 1.5 - - 8 0.5 26 1.6 2 0. 1 5 0.3 7 0.4 1,575 100.0 80-100 16 0.7 2 0.1 12 0.5 12 0.5 1 0.1 - - 6 0.3 2,070 100.0 100-120 14 0.7 - - 1 0.1 4 0.2 4 0.2 - - 6 0.3 1,980 100.0 F 0-20 2 2.3 - - - - - - - - - - - - 87 100.0 20-40 1 0.3 - - - - - - 329 100.0 B H 120-130 10 0.6 - - - - 5 0.3 2 0.1 - - 8 0.4 1,819 100.0 F 40-60 - - - - - - - - - - - - - - 164 100.0 H 130-140 16 0.6 - - 11 0.4 15 0.6 6 0.2 3 0.1 11 0.4 2,570 100.0 A 160-180 1 0.2 3 0.5 651 100.0 180-200 2 0.3 - - - - - - - - 4 0.5 3 0.4 748 100.0 A 200-220 - - - - 3 0.8 2 0.5 1 0.2 - - 3 0.8 399 100.0 220-240 2 1.7 120 100.0 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY Table 2.-Shell artifacts from the Valdivia Phase {fig. 20) VOLUME l Some levels not found in the seriated pottery sequences are interdigitated into this sequence based on their approximate position in time from all the evidence available. No samples were available from Period D Levels. Levels in seriated sequence Abraders & Polishers Beads Bowl or Cup Disks Fishhook Blanks Fishhooks Small- Medium Drilled Clam Shell Pendants Drilled & Shaped Pendants Drilled Pecten Pendant Scoops, Spoons or Spatulas D O O -o Periods Site Cut: Sec- tion Level (cm.) Miscella ', Work Shell C G-54 G-31 1 A H 20-40 40-60 60-80 80-100 100-120 120-130 20-40 0-20 20-40 40-60 60-80 80-100 1 3 1 1 1 - 1 2 1 2 1 2 4 3 2 1 2 1 1 2 1 - - B G-31 G-88 G-31 G-88 G-31 A F A 2 H F H 2 A J:E J:E J:E J:E 100-120 40-60 0-20 20-40 60-80 80-100 0-30 30-60 120-130 40-60 130-140 60-90 100-120 120-140 140-160 90-120 120-150 150-180 180-210 1 1 1 1 3 1 1 1 6 4 1 4 1 1 4 2 2 1 1 - 2 1 - WHOLE VOLUME APPENDIX 1 Table 2.-Shell artifacts from the Valdivia Phase {fig. 20)-Continued 187 Levels in seriated sequence 1 1 Abraders & Polishers Beads Bowl or Cup Disks 1 Fishhook Blanks Fishhooks Small- Medium Drilled Clam Shell Pendants Drilled & Shaped Pendants 1 Drilled , Pecten Pendant Scoops, Spoons or Spatulas ieous Period Site Cut: Sec- tion: Level (cm.) Miscellar Workec SheU A 160-180 B 180-200 - _ - - 1 1 - - - 3 1 A 180-200 - - - - - - - - - - - B 200-220 - _ - - - - 1 - - 1 - 220-240 - - - - - - - - - - - A 200-220 1 - - - - - - - - - - B 240-260 - - - - 1 - - - - 1 - 260-280 - - - - - - - - - - - J:E 240-270 1 - - 1 1 2 - - - - - J:D 150-180 - - - - - - - - - - - J:E 270-300 - - - - - - - - - - 2 B 280-300 - - - - - - - - - - - 300-320 1 - - - - - - - - - - 320-340 - - - - - - - - - - 340-360 1 - - - - - - - - 1 - A 360-380 - - - - - - - - - - - A 220-240 - - - - - - - - - - - J:E 300-330 - 1 - 1 2 1 - 1 - - - A 240-260 - - - - 1 - - - - - - 260-280 - - - - - - - - - - - 280-300 - - - - - - - - - - - J:E 360-390 - 1 - - 1 1 - 1 - - 1 J:E 390-420 - - - - 1 - - - - - - G-88 3 0-20 - - - - - 1 - - - - - G-31 A 320-340 - - - - - - - - - - - J:D 240-270 - - - - 1 - - - - - - 270-300 - - - - - - - - - - - 300-330 - 2 - - 1 2 - 2 - - - A 340-360 - - - - - - - - - - - 360-380 - - - - - - - - - - - 380-400 - - - - - - - - - - - 400-420 - - - - - 1 - - - - Totals : 12 11 1 9 32 17 10 11 1 9 4 188 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Table 3.-Stone artifacts and natural stone from Valdivia Phase {fig. 18) Some levels not found in the seriated pottery sequences are interdigitated into this sequence based on their approximate position in time from all the evidence available. No samples were available from Period D levels. Levels in seriated sequence Abraders Blades Choppers Cores Gravers Grinding Stones Hammerstones "Jaketown" Perforators Paint Stones Pebble Polishers Reamers Saws Scrapers Sinkers Burnt Sandstone Fragments Coquina Fragments Mudstone Fragments Flakes Natural Waterworn Pebbles Periods V 3 U > <-> i) ?>- C G-54 G-31 1 A H 20-40 40-60 60-80 80-100 100-120 120-130 20-40 0-20 20-40 40-60 60-80 80-100 1 1 1 2 4 6 3 2 3 1 1 2 3 1 1 5 3 2 3 2 5 4 7 1 1 1 1 4 1 1 1 5 4 2 1 10 3 2 3 6 5 1 2 1 2 2 1 - 3 8 5 3 3 1 1 3 1 3 2 5 3 3 2 1 1 1 1 1 204 130 87 72 74 53 28 190 365 363 318 509 16 21 28 21 26 226 245 180 284 268 71 59 380 220 276 186 250 51 70 60 39 61 21 34 55 87 118 62 58 2 9 12 19 19 6 15 26 21 34 27 43 B G-88 G-31 G-88 G-31 A F A 2 H F H 2 A 100-120 40-60 0-20 20-40 60-80 80-100 0-30 30-60 120-130 40-60 130-140 60-90 100-120 120-140 140-160 1 3 1 9 3 2 4 3 7 1 4 3 2 1 1 1 2 1 1 3 3 9 4 4 2 2 1 3 2 6 1 3 3 1 1 2 3 2 2 1 1 1 2 1 - 1 1 1 1 4 3 2 1 626 162 196 239 210 55 28 9 108 54 77 1 58 37 126 18 7 11 2 12 1 5 3 1 8 157 128 131 266 163 50 21 23 687 44 1,136 31 32 23 48 36 20 15 21 27 11 36 13 23 6 26 16 1 2 8 22 12 8 17 19 2 2 4 12 1 12 4 4 3 A G-88 G-31 B A B A B A 3 A 160-180 180-200 180-200 200-220 220-240 200-220 240-260 260-280 280-300 300-320 320-340 340-360 360-380 220-240 240-260 260-280 280-300 320-340 340-360 360-380 380-400 400-420 1 1 1 1 5 1 38 13 1 1 2 1 2 1 1 1 1 16 3 2 1 91 1 1 3 1 4 2 1 23 - 1 1 1 3 1 1 1 1 - 2 1 3 1 3 1 104 60 143 69 25 29 5 9 1 19 23 3 6 5 2 1 2 1 5 44 24 63 45 44 28 14 7 6 22 16 8 2 2 3 3 2 9 1 3 1 1 2 4 5 6 6 6 2 4 2 2 1 12 8 Totals : 9 60 12 5 41 3 8 40 4 4,869 202 5,906 1,015 404 WHOLE VOLUME APPENDIX 1 189 Table 4. Chronological distribution of fauna! remains from Valdivia Phase excavations Levels in seriated "3 sequence Mamm 1 Periods Site 1 Cut Level (cm.) Fish Crab Bird 1 Sea Turtle Deer 1 Small Land 1 Reptile Cameloid O P G-54 1 20-40 45 (a) _ 4 22 4 40-60 52 0) - 3 21 3 1 - - 60-80 110 () - - 49 - - - - 80-100 126 () 3 14 153 4 - 4 3 100-120 109 (a) 1 4 311 2 2 9 2 C 120-130 13 () 6 1 80 3 2 2 - G-31 A 20-40 9 (tt) - - _ - _ _ _ G-31 H 0-20 7 2 - - 3 - _ - _ 20-40 16 10 - - 5 - _ - _ 40-60 27 19 - - - - - - _ 60-80 25 22 - 3 2 - - - - 80-100 72 74 - 4 - - - - - 100-120 62 84 _ 4 _ - _ _ _ G-31 A 40-60 16 2 - _ _ - _ _ - G-31 F 10-20 1 - - - 1 - _ _ - 20-40 6 - - - - - _ _ - G-31 A 60-80 19 1 - - 1 - - - - 80-100 4 1 - - - - _ - - B G-88 2 0-30 20 - - 1 14 - 1 - - 30-60 17 - - 1 9 - - - - G-31 H 120-130 47 50 - 3 - - - - - G-31 F 40-60 - 1 - - - - - - - G-31 H 130-140 22 39 - 3 1 - - - - G-88 2 60-90 9 - - 2 11 - 1 - - G-31 A 140-160 5 7 - - - - - - - 160-180 3 8 _ _ _ - _ _ _ 180-200 4 1 - - - - - - - 200-220 4 - - - - - - - - A 220-240 10 1 - - - - - - - 240-260 - - - - - - - - - 260-280 - - - - - - - - - 280-300 - - - - - - - - Table 5.-Bone and teeth artifacts from Valdivia Phase. Only those portions of the sequence having specimens are shown; hence, levels from Period D, the lower part of Pe- riod B and all of Period A are omitted. Levels in seriated sequence Worked Fish Vertebrae Fish Bone Awl Deer Antler Awl Saw Fish Teeth Awl oint Period Site Cut Level (cm.) BoneP G-54 1 20^10 1 1 40-60 - - - - - 1 60-80 - 1 1 1 - Bank Excava- - - 5 - - tion Test Trench - - 1 - - on Hill 1 | 80-100 - - 1 2 - C 1 (N.E. ext.) 85-100 - - - - 1 1 100-120 - 2 1 1 - 1 120-130 - - - - - G-31 A 20-40 - - - - - H 0-20 1 - - - - H 20-40 - - - - - H 40-60 - - - - - H 60-80 - - - - - H 80-100 - - - - - H 100-120 _ _ - _ _ A 40-60 - - - - - F 0-20 - - - - - F 20-40 - - - - - B A 60-80 - - - - - A 80-100 - - - - - G-88 2 0-30 - - - - - 2 30-60 - - - - - G-31 H 120-130 - 2 - - - B Count unavailable for these levels. 190 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Table 6.-Frequency of pottery types in surface collections and stratigraphic and test excavations at sites of the Valdivia Phase {figs. 49-53) G-31: Valdivia Pottery types G-25: Punta Arenas de Posorja Tests 20-40 cm. 40-60 cm. Cut A 60-80 cm. 80-100 cm. 100-120 cm. No. % No. % No. No. % No. % No. % Punta Arenas Polished Plain . San Pablo Plain Valdivia Plain Valdivia Polished Plain . . . Valdivia Polished Red .... Valdivia Striated Polished Plain 422(*) 79.5 2.3 1. 1 20.4 18.2 8.0 5.7 6 17 128 86 66 49 1.3 3. 8 28.6 19.2 14.7 10. 9 13 117 76 90 23 3.6 32.6 21. 1 25.0 6.4 4 18 117 82 74 32 1.0 4.5 29.8 20.8 18.8 8. 1 7 11 141 84 114 34 1.5 2.4 30.4 18.1 24.6 7.3 Validiva Applique Fillet Valdivia Broad-line Incised Valdivia Brushed Valdivia Carved Valdivia Combed Valdivia Cord Impressed Valdivia Corrugated Valdivia Cut-and-Beveled Rim Valdivia Embossed Valdivia Excised Valdivia Fine-line Incised Valdivia Fingernail Decorated Valdivia Finger Grooved Valdivia Incised Valdivia Modeled Valdivia Multiple Drag-and-Jab Punctate Valdivia Nicked Broad-line Incised . . . Valdivia Nicked Rib or Nubbin Valdivia Pebble Polished Valdivia Pseudo-Corrugated Valdivia Punctate Valdivia Red Incised Valdivia Red Zoned Punctate Valdivia Rocker Stamped Valdivia Shell Stamped Valdivia Zoned Incised Unclassified 25 0.7 8.3 0.6 2. 1 4.7 1.7 0.7 1.7 24 3.4 9. 1 27. 3 4.5 1.8 5.6 7. 1 0.7 0.7 . 5 0.3 2 0. 22 5. 16 4. 0.7 0. 5 2.8 3. 9 0.8 6. 3 1. 3 0.4 0.2 Machalilla Double-line Incised . Machalilla Embellished Shoulder Machalilla Incised Machalilla Red Banded .... Unclassified Decorated (Machalilla Complex) .... Ayangue Incised Plain Types (Machalilla Complex) Grand Totals . . Subtotals : Plain types . . Decorated types Trade pottery . 531 422 109 0 100.0 100.0 448 352 96 0 100.0 360 319 41 0 100.0 394 327 47 0 100.0 464 391 73 0 100.0 ¯ Eroded surfaces prevented typing. All plain types lumped into a total count. WHOLE VOLUME APPENDIX 1 191 Table 6.-Frequency of pottery types in surface collections and stratigraphic and test excavations at sites of the Valdivia Phase {figs. 49-53)-Continued G-31: Valdivia-Continued Pottery types Cut A-Continued 120-140 cm. 140-160 cm. 160-180 cm. 180-200 cm. 200-220 cm. 220-240 cm. No. % No. % No. % No. % No. % No. % Punta Arenas Polished Plain San Pablo Plain Valdivia Plain Valdivia Polished Plain Valdivia Polished Red Valdivia Striated Polished Plain Valdivia Applique Fillet Valdivia Broad-line Incised Valdivia Brushed Valdivia Carved Valdivia Combed Valdivia Cord Impressed Valdivia Corrugated Valdivia Cut-and-Beveled Rim Valdivia Embossed Valdivia Excised Valdivia Fine-line Incised Valdivia Fingernail Decorated Valdivia Finger Grooved Valdivia Incised Valdivia Modeled Valdivia Multiple Drag-and-Jab Punctate Valdivia Nicked Broad-line Incised . . . Valdivia Nicked Rib or Nubbin Valdivia Pebble Polished Valdivia Pseudo-Corrugated Valdivia Punctate Valdivia Red Incised Valdivia Red Zoned Punctate Valdivia Rocker Stamped Valdivia Shell Stamped Valdivia Zoned Incised Unclassified Machalilla Double-line Incised Machalilla Embellished Shoulder .... Machalilla Incised Machalilla Red Banded Unclassified Decorated (Machalilla Complex) Ayangue Incised Plain Types (Machalilla Complex) . . . Grand Totals Subtotals: Plain types Decorated types Trade pottery 2 13 60 36 54 32 238 197 41 0 0. 8 5.5 25.2 15.1 22.8 13.5 0.8 5. 1 0. 8 8.4 0.8 0.4 0.8 100.0 26 377 267 179 98 3 40 4 13 106 2 1, 119 947 172 0 2.3 33.6 23.9 16.0 8.7 0.2 3. 6 0.4 1.2 9. 5 0.2 0. 1 0.2 0. 1 100.0 16 178 189 174 57 21 674 614 60 0 2.4 26.5 28. 1 25.8 8.5 3. 1 1.8 0. 1 3.4 0. 1 0. 1 0. 1 100.0 15 190 155 184 36 17 11 617 580 37 0 2.4 30.8 25.1 29.8 5.8 2. 8 1.0 0.2 0. 3 100.0 21 119 54 96 9 12 333 299 34 0 6.3 35.8 16.2 28.8 2.7 0.6 100.0 37 130 75 126 18 2 40 438 386 52 0 8.4 29.7 17. 1 28.8 4. 1 1.4 0.7 0.5 9. 1 0.2 100.0 192 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Table 6.-Frequency of pottery types in surf ace collections and stratigraphic and test excavations at sites of the Valdivia Phase {figs. 49-53)--Continued G-31 : Valdivia-Continued Pottery types Cut A-Continued 240- 260 cm. 260- 280 cm. 280- 300 cm. 300- 340 cm. 340- 360 cm. 360- 380 cm. No. % No. % No. % No. % No. % No. % Z Punta Arenas Polished Plain San Pablo Plain 41 110 38 136 21 10.6 28.7 9.9 35.4 5.5 23 81 16 85 4 10.0 35.2 6.9 37.1 1.7 10 34 14 47 8 8.6 29. 1 11.9 40. 3 6.8 14 59 13 28 9.3 39. 1 8.6 18.6 10 50 11 26 9.3 46.4 10.2 24. 1 10 72 4 14 8.8 < Valdivia Plain 64.2 Ph Valdivia Polished Plain 3.6 Valdivia Polished Red 12.4 Valdivia Striated Polished Plain Valdivia Applique Fillet 1 1 2 2 31 1 0.3 0.3 0.5 0.5 8.0 0.3 1 2 2 16 0.4 0.9 0.9 6.9 1 1 2 0.8 0.8 1.7 3 21 11 1 1 2.0 13.9 7.3 0.6 0.6 1 1 7 2 0.9 0.9 6.4 1.8 1 1 11 Valdivia Broad-line Incised 0.7 Valdivia Carved Valdivia Combed Valdivia Cord Impressed Valdivia Corrugated Valdivia Cut-and-Beveled Rim Valdivia Embossed - Valdivia Excised Valdivia Fine-line Incised RATEE Valdivia Fingernail Decorated Valdivia Finger Grooved 0.7 DECO Valdivia Modeled 9.6 Valdivia Multiple Drag-and-Jab Punctate . . Valdivia Nicked Broad-line Incised Valdivia Nicked Rib or Nubbin Valdivia Pebble Polished Valdivia Pseudo-Corrugated Valdivia Punctate Valdivia Red Incised Valdivia Red Zoned Punctate Valdivia Rocker Stamped Valdivia Shell Stamped - Valdivia Zoned Incised - p Machalilla Double-line Incised Machalilla Embellished Shoulder Machalilla Incised Machalilla Red Banded - - - - - - - - - - - - < Pi H Unclassified Decorated (Machalilla Complex) _ Plain Types (Machalilla Complex) - Grand Totals 384 346 38 0 100.0 230 209 21 0 100.0 117 113 4 0 100.0 151 114 37 0 100.0 108 97 11 0 100.0 113 100 13 0 Subtotals: Plain types 100. 0 Decorated types - WHOLE VOLUME APPENDIX 1 193 Table 6.-Frequency of pottery types in surface collections and stratigraphic and test excavations at sites of the Valdivia Phase {hgs. 49-53)-Continued G-31: Valdivia-Continued Pottery types Cut A-Continued CutF 380-400 cm. 400-420 cm. 420-440 cm. 0-20 cm. 20-40 cm. 40-60 cm. No. % No. % No. No. No. % No. % Punta Arenas Plain San Pablo Plain Valdivia Plain Valdivia Polished Plain iValdivia Polished Red Valdivia Striated Polished Plain .... Valdivia Applique Fillet Valdivia Broad-line Incised Valdivia Brushed Valdivia Carved Valdivia Combed Valdivia Cord Impressed Valdivia Corrugated Valdivia Cut-and-Beveled Rim Valdivia Embossed Valdivia Excised Valdivia Fine-line Incised ' Valdivia Fingernail Decorated Valdivia Finger Grooved Valdivia Incised Valdivia Modeled Valdivia Multiple Drag-and-Jab Punctate Valdivia Nicked Broad-line Incised ' . . Valdivia Nicked Rib or Nubbin .... Valdivia Pebble Polished Valdivia Pseudo-Corrugated Valdivia Punctate Valdivia Red Incised Valdivia Red Zoned Punctate Valdivia Rocker Stamped Valdivia Shell Stamped Valdivia Zoned Incised Unclassified Machalilla Double-line Incised Machalilla Embellished Shoulder .... Machalilla Incised Machalilla Red Banded Unclassified Decorated (Machalilla Complex) Ayangue Incised Plain Types (Machalilla Complex) . . . Grand Totals Subtotals: Plain types Decorated types Trade pottery 10 73 9 11 114 103 11 0 64.2 7.9 9.7 0.8 0.8 0.8 7.0 100.0 23 170 13 22 19 249 228 21 0 9.2 68.4 5.2 0.4 0.4 7.6 100.0 11 92 4 7 124 114 10 0 8.9 74.2 3.2 5.7 1.6 0.8 4.0 1.6 100.0 103 747 372 181 91 9 49 50 501 8 1 7 16 1 12 2,157 1,494 663 0 34.5 17.3 8.4 4.2 0.4 2.3 2.3 0.4 23.3 0.3 0. 1 0.3 0.8 0. 1 0.5 100.0 49 273 276 157 138 286 1,262 893 369 0 3.9 21.6 21.8 12.4 10.9 0.2 2. 1 2. 6 0.5 22.6 0.3 0.2 0.2 0.6 0. 1 100.0 10 40 72 33 43 245 198 47 0 4. 1 16.4 29.4 13.5 17.6 2.4 1.2 0.4 11.4 0.4 0.4 0.8 2.0 100.0 194 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Table 6.-Frequency of pottery types in surface collections and stratigraphic and test excavations at sites of the Valdivia Phase {figs. 49-53)-Continued G-31: Valdivia-Continued Pottery types Cut H 0-20 cm. 20-40 cm. 40-60 cm. 60-80 cm. 80-100 cm. 100-120 cm. No. % No. % No. % No. % No. % No. % Punta Arenas Plain San Pablo Plain Valdivia Plain Valdivia Polished Plain Valdivia Polished Red Valdivia Striated Polished Plain Veldivia Applique Fillet Valdivia Broad-line Incised Valdivia Brushed Valdivia Carved Valdivia Combed Valdivia Cord Impressed Valdivia Corrugated Valdivia Cut-and-Beveled Rim Valdivia Embossed Valdivia Excised Valdivia Fine-line Incised Valdivia Fingernail Decorated Valdivia Finger Grooved Valdivia Incised Valdivia Modeled Valdivia Multiple Drag-and-Jab Punctate Valdivia Nicked Broad-line Incised . . . Valdivia Nicked Rib or Nubbin . . . . Valdivia Pebble Polished Valdivia Pseudo-Corrugated Valdivia Punctate Valdivia Red Incised Valdivia Red Zoned Punctate Valdivia Rocker Stamped Valdivia Shell Stamped Valdivia Zoned Incised Unclassified Machalilla Double-line Incised Machalilla Embellished Shoulder . . . . Machalilla Incised Machalilla Red Banded Unclassified Decorated (Machalilla Complex) Ayangue Incised Plain Types (Machalilla Complex) . . . Grand Totals Subtotals: Plain types Decorated types Trade pottery 9 119 141 109 101 19 42 124 93 780 479 301 0 1.1 15.4 18.1 14.0 13.0 2.4 5. 3 15.9 0.9 11.9 0.1 1.5 0.4 100.0 7 243 224 208 161 21 71 442 13 117 1 1 6 1 19 1,535 843 692 0 0.5 15.8 14.6 13.6 10.5 1. 3 4.6 28.8 0.8 7.6 0. 1 0. 1 0.4 1 2 100.0 26 243 227 142 186 37 65 392 145 4 15 7 1,497 824 673 0 1.7 16.3 15.2 9.5 12.4 2.5 4. 3 26.1 0.5 9.7 0.3 1. 0 0.5 100.0 12 162 137 105 82 22 50 168 64 1 5 6 1 823 496 325 2 1.5 19.7 16.7 12.8 9.9 2.7 6.1 20.4 0. 6 7.8 0. 1 0. 6 0.7 0. 1 0. 1 0. 1 0. 1 100.0 9 303 156 95 103 15 42 128 10 125 994 666 328 0 0.9 30.4 15.7 9.6 10.4 1.5 4.2 12.9 1.0 12.6 0.1 0. 1 0.3 0.2 0. 1 100.0 22 256 165 109 72 15 26 93 1 74 1 1 847 624 223 0 2.6 30.2 19.5 12.9 8.5 1.8 3. 1 11.0 0.1 0. 1 0.5 8.7 0.1 0.1 0.1 0.5 0. 2 100.0 WHOLE VOLUME APPENDLX 1 195 Table 6.-Frequency of pottery types in surface collections and stratigraphic and test excavations at sites of the Valdivia Phase {figs. 49-53)-Continued G-31: Valdivia-Continued Pottery types Cut H-Continued Cut J: Section D 120- 130 cm. 130-140 cm. 60-90 cm. 90-120 cm. 120-150 cm. 150-180 cm. No. % No. % No. % No. % No. % No. % Punta Arenas Plain 9 236 147 141 49 1.2 33.5 21.0 20.0 6.9 26 211 230 127 47 3.5 28.2 30.8 17.0 6.3 U>617 500 522 68 31.3 25.3 26.4 3.4 1,495 1,238 1,111 243 33.00 27. 35 24.70 5.38 3,729 3,559 3,326 681 30.30 28.94 27.10 5. 55 5,645 2,201 3,010 653 z San Pablo Plain . . < Valdivia Plain 46. 70 ?J Ph Valdivia Polished Plain Valdivia Polished Red Valdivia Striated Polished Plain. . . . 18.30 25.00 5.40 Valdivia Applique Fillet Valdivia Broad-line Incised Valdivia Brushed 2 33 22 8 55 2 1 0.3 4.7 3.1 1.1 7.8 0.3 0.1 1 36 13 5 1 43 4 1 1 2 0.1 4.8 1.7 0.7 0. 1 5.8 0.5 0.1 0. 1 0.3 11 63 37 2 4 4 2 83 12 42 5 2 0.6 3.2 1.9 0.1 0.2 0.2 0.1 4.2 0.6 2. 1 0.3 0.1 24 70 189 1 2 1 3 7 88 29 1 28 4 2 2 0. 50 1.50 4.05 0.02 0.03 0.02 0.05 0.15 1.90 0.60 0.02 0.60 0.07 0.03 0.03 74 204 390 3 2 1 15 1 103 151 1 48 8 4 0.60 1.66 3.17 0.02 0.01 0.01 0. 10 0.01 0.80 1.23 0.01 0.40 0.06 0.03 21 108 25 4 3 8 25 4 4 19 330 10 0.20 0.90 0.21 Valdivia Carved _ Valdivia Combed 0.03 Valdivia Cord Impressed Valdivia Cut-and-Beveled Rim .... 0.01 0.06 ATED Valdivia Fingernail Decorated .... 0.20 0.03 0.03 0.15 DECOR 2.70 Valdivia Multiple Drag-and-Jab Punc- tate Valdivia Nicked Broad-line Incised . . Valdivia Nicked Rib or Nubbin . . . Valdivia Pebble Polished Valdivia Pseudo-Corrugated 0.08 Valdivia Red Incised Valdivia Red Zoned Punctate .... Valdivia Rocker Stamped Valdivia Shell Stamped Valdivia Zoned Incised - Machalilla Double-line Incised .... Machalilla Embellished Shoulder . . . - - - - - - - - - - - - TRADE Machalilla Red Banded Unclassified Decorated (Machalilla ; _ Plain Types (Machalilla Complex) . . - Subtotals: 705 582 123 0 100.0 748 641 107 0 100.0 1,974 1,707 267 0 100.0 4,538 4,087 451 0 100.0 12, 300 11,295 1,005 0 100.0 12,070 11,509 561 0 100.0 b In Cut J, Sections D and E, unpolished plain sherds were not separated into two types (see text p. 42). 196 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Table 6.-Frequency of pottery types in surface collections and stratigraphic and test excavations at sites of the Valdivia Phase {figs. 49-53) L,on G-31: Valdivia-Continued Pottery types Cut J: Section D-Continued 180-210 cm. 210-240 cm. 240-270 cm. 270-300 cm. 300-330 cm. 330-360 cm. No. % No. % No. % No. % No. % No. % Punta Arenas Plain San Pablo Plain Valdivia Plain Valdivia Polished Plain Valdivia Polished Red Valdivia Striated Polished Plain. . . Valdivia Applique Fillet Valdivia Broad-line Incised .... Valdivia Brushed Valdivia Carved Valdivia Combed Valdivia Cord Impressed Valdivia Corrugated Valdivia Cut-and-Beveled Rim . . . Valdivia Embossed Valdivia Excised Valdivia Fine-line Incised Valdivia Fingernail Decorated . . . Valdivia Finger Grooved Valdivia Incised Valdivia Modeled Valdivia Multiple Drag-and-Jab Punc tate Valdivia Nicked Broad-line Incised . Valdivia Nicked Rib or Nubbin . . Valdivia Pebble Polished Valdivia Pseudo-Corrugated .... Valdivia Punctate Valdivia Red Incised Valdivia Red Zoned Punctate . . . Valdivia Rocker Stamped Valdivia Shell Stamped Valdivia Zoned Incised Unclassified Machalilla Double-line Incised . . . Machalilla Embellished Shoulder . . Machalilla Incised Machalilla Red Banded Unclassified Decorated (Machalilla Complex) Ayangue Incised Plain Types (Machililla Complex) . Grand Totals Subtotals: Plain types Decorated types .... Trade pottery 3,080 918 942 170 72 4 3 47 10 1 315 10 5,587 5,110 477 0 55.20 16.40 16.85 3.03 1.29 0.07 0.07 0.05 0.05 0.84 0.18 0.01 5.64 0.18 0.14 100.0 2,084 887 638 118 28 3 23 4 1 1 3 42 3 7 174 4,030 3,727 303 0 51.80 22.00 15.80 2.93 0.68 0.07 0.56 0.10 0.03 0.03 0.07 1.04 0.07 0.17 4.30 0.10 0.10 0.15 100.0 3,122 796 689 162 36 5 25 6 54 1 12 6 200 5,145 4,769 376 0 60.69 15.50 13.40 3.16 0.70 0.09 0.48 0.05 0.11 1.05 0.02 0.23 0.11 3.90 0.03 0.07 0.41 100.0 5,197 1,512 604 117 29 10 41 4 75 8 9 4 231 7,887 7,430 457 0 65.90 19.21 7.66 1.48 0.36 0.12 0. 52 0.10 0.02 0.05 0.95 0.10 0.11 0.05 2.93 0.07 0.11 0.23 0.03 100.0 3,577 905 636 33 50 1 24 2 4 7 181 19 5,458 5,151 307 0 1,268 275 146 65.60 16.54 11.64 0.61 15 16 2 0.09 0.92 0.05 0.02 53 4 0.05 100. 0 1, 872 1,697 175 0 67.7 14.7 7.8 0.4 0.8 0.9 0.1 2.9 0.2 0.3 3.3 0. 1 0.5 0.3 0.1 100.0 WHOLE VOLUME APPENDIX 1 197 Table 6.-Frequency of pottery types in surface collections and stratigraphic and test excavations at sites of the Valdivia Phase {figs. 49-53)-Continued G-31: Valdivia-Continued Pottery types Section D-Con. Cut J: Section E 360-390 cm. 90-120 cm. 120-150 cm. 150-180 cm. 180-210 cm No. No. % No. % No. % No. % Punta Arenas Plain San Pablo Plain 1 Valdivia Plain Valdivia Polished Plain Valdivia Polished Red Valdivia Striated Polished Plain . . 1,136 78 71 4 83.9 5.7 5.2 0.3 7,009 6,679 3,658 732 28.10 26.68 14.63 2.93 3,409 3,814 1,881 253 25.80 29.06 14.32 1.92 3,333 3,214 2,102 409 28.29 27.28 17.88 3.48 4,187 2,816 2,080 398 37.00 24.90 18.35 3.55 Valdivia Applique Fillet Valdivia Broad-line Incised Valdivia Brushed Valdivia Carved Valdivia Combed Valdivia Cord Impressed Valdivia Corrugated Valdivia Cut-and-Beveled Rim . . . Valdivia Embossed Valdivia Excised Valdivia Fine-line Incised Valdivia Fingernail Decorated .... Valdivia Finger Grooved Valdivia Incised Valdivia Modeled Valdivia Multiple Drag-and-Jab Punc- tate Valdivia Nicked Broad-line Incised. . Valdivia Nicked Rib or Nubbin . . . Valdivia Pebble Polished Valdivia Pseudo-Corrugated Valdivia Punctate Valdivia Red Incised Valdivia Red Zoned Punctate .... Valdivia Rocker Stamped Valdivia Shell Stamped Valdivia Zoned Incised Unclassified 14 1 3 36 0.9 0.1 1.0 0.1 0.2 2.6 490 1,360 2,993 3 12 1 23 8 13 1,659 9 28 269 1 24 3 1 1.97 5.45 12.00 0.01 0.04 0.01 0.9 0.03 0.05 6.65 0.03 0.11 1.07 0.01 0.09 0.01 0.01 0.01 322 699 1,635 5 11 43 9 1 873 9 14 146 2.44 5.34 12.41 0.04 0.08 0.32 0.07 0.01 6.64 0.07 0.10 1.12 0.21 0.01 221 554 1,043 1 1 4 1 28 6 3 672 22 142 1.88 4.70 8.87 0.01 0.01 0.03 0.01 0.23 0.06 0.02 5.66 0.18 0.05 1.21 0.10 0.02 0.02 162 413 633 428 63 60 1.43 3.64 5.59 0.02 0.02 0.01 0.36 0.07 0.02 3.78 0.55 0.01 0.52 0.14 0.01 0.03 Machalilla Double-line Incised . . . Machalilla Embellished Shoulder . . Machalilla Incised Machalilla Red Banded Unclassified Decorated (Machalilla Complex) Ayangue Incised Plain Types (Machalilla Complex). . , 0.02 0.04 0.01 Grand Totals . . Subtotals : Plain types . . Decorated types Trade pottery 1,356 1,289 67 0 100.0 24, 981 18, 078 6,899 4 100.0 13,159 9,357 3,797 5 100.0 11,780 9,058 2,721 1 100.0 11,317 9,481 1,836 0 100.0 198 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Table 6.-Frequency of pottery types in surface collections and stratigraphic and test excavations at sites of the Valdivia Phase [figs. 4y-0o) on G-31: Valdivia-Continued Pottery types Cut J: Section E-Continued 210-240 cm. 240-270 cm. 270-300 cm. 300-330 cm. 330-360 cm. 360-390 cm. No. No. % No. % No. % No. % No. 7o Punta Arenas Plain San Pablo Plain Valdivia Plain Valdivia Polished Plain . . . Valdivia Polished Red .... Valdivia Striated Polished Plain 2,428 1,436 1,256 221 38.00 22.53 19.60 3.46 3,627 2,283 222 336 40.40 25.62 25.00 3.77 2,483 1,195 1,101 164 48.25 23.20 21.40 3.19 1,116 356 511 65 47.50 15.10 21.80 2.70 1,975 650 1,270 41 47.35 15.60 30.50 0.98 2,128 372 997 6 57.85 10.15 27.05 0.16 Valdivia Applique Fillet Valdivia Broad-line Incised Valdivia Brushed Valdivia Carved Valdivia Combed Valdivia Cord Impressed Valdivia Corrugated Valdivia Cut-and-Beveled Rim . . . Valdivia Embossed Valdivia Excised Valdivia Fine-line Incised Valdivia Fingernail Decorated . . . Valdivia Finger Grooved Valdivia Incised Valdivia Modeled Valdivia Multiple Drag-and-Jab Punc- tate Valdivia Nicked Broad-line Incised. . Valdivia Nicked Rib or Nubbin . . . Valdivia Pebble Polished Valdivia Pseudo-Corrugated Valdivia Punctate Valdivia Red Incised Valdivia Red Zoned Punctate .... Valdivia Rocker Stamped Valdivia Shell Stamped Valdivia Zoned Incised Unclassified 155 365 4 3 1 1 1 15 13 5 3 273 59 1.38 2.42 5.72 0.06 0.04 0.01 0.01 0.01 0.23 0.20 0.08 0.04 4.26 0.92 0.11 0.69 0.01 0.14 0.04 34 178 74 24 15 5 4 43 52 13 0.38 2.00 0.82 0.05 0.07 0.27 0. 17 0.05 0.04 0.48 0.58 0.03 0. 14 0.11 0.02 41 0.80 0.02 0.05 0. 11 0.46 0.04 0.64 1.72 0.02 0.02 0.04 0.04 46 1 45 3 37 152 1.90 0.20 0.20 0. 10 1.90 0. 10 1.60 6.50 0.20 0.20 19 5 6 1 28 5 3 20 121 0.46 0. 16 0.12 0. 14 0.02 0.67 0.12 0.07 0.48 2.90 0.16 0.07 0.20 42 1 2 2 114 0.13 0.24 0.02 1.14 0.02 0.04 0.04 3. 10 0.02 0.02 0.02 Machalilla Double-line Incised ... Machalilla Embellished Shoulder . . Machalilla Incised Machalilla Red Banded Unclassified Decorated (Machalilla Complex) Ayangue Incised Plain Types (Machalilla Complex). . 0.04 Grand Totals . . Subtotals : Plain types . . Decorated types Trade pottery 6,394 5,341 1,050 3 100.0 8,936 8,468 468 0 100.0 5,147 4,943 204 0 100.0 2,352 2,048 304 0 100.0 4,171 3,936 235 0 100.0 3,682 3,503 179 0 100.0 WHOLE VOLUME APPENDLX 1 199 Table 6.-Frequency of pottery types in surface collections and stratigraphic and test excavations at sites of the Valdivia Phase {figs. 49-53)-Continued G-31: Valdivia-Continued G-54: Buena Vista Pottery types Cut J: Section E-Continued Cut 1 390-420 cm. 420-450 cm. 0-20 cm. 20-40 cm. 40-60 cm. 60-80 cm. No. % No. % No. % No. % No. % No. % Punta Arenas Plain San Pablo Plain Valdivia Plain Valdivia Polished Plain Valdivia Polished Red Valdivia Striated Polished Plain. . Valdivia Applique Fillet Valdivia Broad-line Incised.... Valdivia Brushed Valdivia Carved Valdivia Combed Valdivia Cord Impressed Valdivia Corrugated Valdivia Cut-and-Beveled Rim . . Valdivia Embossed Valdivia Excised Valdivia Fine-line Incised .... Valdivia Fingernail Decorated . . Valdivia Finger Grooved Valdivia Incised Valdivia Modeled Valdivia Multiple Drag-and-Jab Punctate Valdivia Nicked Broad-line Incised Valdivia Nicked Rib or Nubbin. . Valdivia Pebble Polished Valdivia Pseudo-Corrugated . . . Valdivia Punctate Valdivia Red Incised Valdivia Red Zoned Punctate . . Valdivia Rocker Stamped .... Valdivia Shell Stamped Valdivia Zoned Incised Unclassified Machalilla Double-line Incised. . Machalilla Embellished Shoulder . Machalilla Incised Machalilla Red Banded Unclassified Decorated (Machalilla Complex) Ayangue Incised Plain Types (Machalilla Complex) Grand Totals Subtotals : Plain types Decorated types . . . Trade pottery .... 1,580 530 195 2,454 2,313 141 0 64.40 21.62 7.95 0.32 0. 12 0. 16 0.20 0. 12 0.49 0.12 0.04 0.12 4.02 0.08 0.08 0. 16 100.0 702 143 172 16 1,133 1,033 100 0 61.95 12.64 15.18 1.41 0.09 0.97 0.09 1.06 0.17 0.09 6.00 0. 35 100.0 10 36.6 16.2 11.8 3.2 6.5 2.1 7.5 10.8 2. 1 1. 1 2. 1 100.0 53 433 130 198 114 97 1 266 80 1,603 1,025 480 98 3.3 27. 1 8.1 12.3 7.1 6.0 0.2 3.2 6.1 0. 1 0.2 0. 1 0. 1 16.6 0.9 0. 5 0.6 0. 1 0.4 0.3 0. 5 1 1 5.0 100.0 77 536 127 295 160 147 10 76 222 393 67 109 2,329 1,342 780 207 3. 3 23. 1 5.5 12.7 6.9 6.3 0.4 3.2 9.5 0.2 0.2 16.9 0.3 0. 5 0. 9 0. 1 0. 1 0.5 0. 1 0.7 0.8 0. 1 2. 8 0.3 4.6 100.0 22 822 209 464 154 225 149 321 502 1 24 69 46 46 3,172 1,896 1,165 111 0.70 26.00 6. 60 14.68 4.90 7. 10 0.20 4.70 10. 10 0.03 0.03 15.9 0.03 0.7 2.2 1.4 0. 1 0. 1 0.5 0.03 0.6 0.2 0. 1 1.7 1.4 100.0 200 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Table 6.-Frequency of pottery types in surface collections and stratigraphic and test excavations at sites of the Valdivia Phase {figs. 49-53) Continue G-54: Buena Vista-Continued Cut 1-Continued Cut 2 Pottery types 3-100 cm. 100-120 cm. 120-130 cm. 20-30 cm. 30-40 cm. 40-50 cm. No. % No. % No. % No. % No. No. % Punta Arenas Plain San Pablo Plain Valdivia Plain Valdivia Polished Plain Valdivia Polished Red Valdivia Striated Polished Plain . . . Valdivia Applique Fillet Valdivia Broad-line Incised Valdivia Brushed Valdivia Carved Valdivia Combed Valdivia Cord Impressed Valdivia Corrugated Valdivia Cut-and-Beveled Rim . . . Valdivia Embossed Valdivia Excised Valdivia Fine-line Incised Valdivia Fingernail Decorated . . . Valdivia Finger Grooved Valdivia Incised Valdivia Modeled Valdivia Multiple Drag-and-Jab Punc- tate Valdivia Nicked Broad-line Incised. . Valdivia Nicked Rib or Nubbin . . . Valdivia Pebble Polished Valdivia Pseudo-Corrugated Valdivia Punctate Valdivia Red Incised Valdivia Red Zoned Punctate .... Valdivia Rocker Stamped Valdivia Shell Stamped Valdivia Zoned Incised Unclassified Machalilla Double-line Incised . . . Machalilla Embellished Shoulder . . Machalilla Incised Machalilla Red Banded Unclassified Decorated (Machalilla Complex) Ayangue Incised Plain Types (Machalilla Complex). . Grand Totals Subtotals : Plain types Decorated types Trade pottery 4 948 419 1,218 445 716 11 249 535 951 1 28 61 72 48 5,781 3,750 2,017 14 0.06 16.40 7.20 21.10 7.70 12.40 0.20 4.30 9.30 0.10 0.03 16.50 0.01 0.50 1.00 1.20 0.20 0.01 0.70 0.80 0.02 0.20 0.01 0.06 100.0 28 609 295 799 346 595 6 179 273 1 4 2 692 31 47 43 32 4,036 2,672 1,357 7 0.7 15.1 7.3 19.7 8.6 14.8 0. 1 4.3 6.7 0. 1 0.2 0. 1 0. 1 17.2 0.8 1.2 1.0 0. 1 0.2 0.7 0.8 0.2 100.0 14 296 173 559 214 344 70 122 1 394 37 14 15 12 2,290 1,600 690 0 0.6 12.9 7.5 24.4 9.4 15.1 3. 1 5.3 0. 1 17.2 1.6 0.6 0.6 0.1 0.2 0.8 0.5 100.0 311 76 53 47 17 43 84 16 181 857 504 349 4 36.2 8.9 6.2 5.5 1.9 5.0 9.7 1.8 21.5 0.5 0.6 1.1 0. 1 0. 1 0.5 0. 1 0.3 100.0 645 207 132 148 87 50 149 306 1,772 1,219 550 3 36.4 11.7 7.4 8.4 4.9 2.8 8.4 0.3 17.3 0. 2 0.4 0. 6 0.2 0.3 0.5 0.2 100.0 770 276 125 112 112 72 91 10 1 1 8 477 2,101 1,395 700 6 36.7 13.2 5.9 5.4 5.4 3.4 4.3 0.5 0.1 0. 1 0.3 22.6 0.2 0.3 0.4 0. 1 0.1 0.2 0.1 0.4 0.2 0.1 100.0 Level 0-20 cm. sterile. WHOLE VOLUME APPENDIX 1 201 Table 6.-Frequency of pottery types in surface collections and stratigraphic and test excavations at sites of the Valdivia Phase {figs. 49-53)-Continued G-54: Buena Vista-Continued Pottery types Cut 2-Continued 50-60 cm. 60-70 cm. 70-80 cm. 80-90 cm. 90-100 cm. 100-110 cm. No. % No. No. % No. % No. % No. % Punta Arenas Plain San Pablo Plain Valdivia Plain Valdivia Polished Plain Valdivia Polished Red Valdivia Striated Polished Plain . . . Valdivia Applique Fillet Valdivia Broad-line Incised Valdivia Brushed Valdivia Carved Valdivia Combed Valdivia Cord Impressed Valdivia Corrugated Valdivia Cut-and-Beveled Rim . . . Valdivia Embossed Valdivia Excised Valdivia Fine-line Incised Valdivia Fingernail Decorated . . . Valdivia Finger Grooved Valdivia Incised Valdivia Modeled Valdivia Multiple Drag-and-Jab Punc- tate Valdivia Nicked Broad-line Incised. . Valdivia Nicked Rib or Nubbin . . . Valdivia Pebble Polished Valdivia Pseudo-Corrugated Valdivia Punctate Valdivia Red Incised Valdivia Red Zoned Punctate .... Valdivia Rocker Stamped Valdivia Shell Stamped Valdivia Zoned Incised Unclassified Machalilla Double-line Incised . . . Machalilla Embellished Shoulder . . Machalilla Incised Machalilla Red Banded Unclassified Decorated (Machalilla Complex) Ayangue Incised Plain Types (Machalilla Complex). . Grand Totals Subtotals : Plain types Decorated types Trade pottery 501 208 133 50 116 62 107 289 1,502 1,008 494 0 33.4 13.8 8.9 3.3 7.7 4.1 7.1 0. 1 19.3 0.4 0.5 0.6 0.2 0. 1 0.5 100.0 508 182 148 94 112 49 30 1 371 11 1,546 1,044 501 1 32.9 11.8 9.6 6.1 7.3 3.1 1.9 0. 1 24.0 0.5 1.0 0.4 0.2 0. 1 0.2 0.7 0. 1 100.0 381 205 216 120 85 12 309 9 3 2 4 6 1,425 1,007 418 0 26.8 14.4 15.1 8.5 5.9 2.5 2.7 0.8 21.7 0.6 0.2 0.1 0.3 0.4 100.0 243 161 118 76 79 202 930 677 253 0 26.6 17.3 12.7 8.3 8.2 1.1 1.7 0.4 21.7 0.6 0.3 0.7 0.4 100.0 470 264 207 127 108 34 16 352 1,624 1,176 448 0 29.1 16.3 12.7 7.8 6.7 2. 1 0.9 0. 1 21.7 0.7 0.5 0.6 0.1 0.3 0.4 100.0 568 260 254 110 185 479 7 20 18 1,967 1,377 590 0 28.9 13.2 12.9 5.6 9.4 2.0 0.5 0. 1 24.5 0.3 1.0 0.9 0.1 0.1 0.1 0. 3 100.0 202 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Table 6.-Frequency of pottery types in surface collections and stratigraphic and test excavations at sites of the Valdivia Phase {figs. 49-53)-Continued G-54: Buena Vista-Continued Pottery types Cut 2-Continued Test Trench Cut 1: NE. Extension 110-120 cm. 120-130 cm. Ctr. of site E. of burial area 0-85 cm. 85-120 cm. No. % No. % No. % No. % No. % No. % Punta Arenas Plain San Pablo Plain Valdivia Plain Valdivia Polished Plain .... Valdivia Polished Red .... Valdivia Striated Polished Plain 277 222 203 135 195 19.6 15.7 14.3 9.5 13.8 92 50 106 31 48 20.9 11.4 24.2 7.1 10.9 269 35 111 115 135 13.6 1.7 5.6 5.8 6.8 431 167 327 229 304 19.8 7.7 15.6 10.6 14.0 27 6 34 60 53 3.4 0.7 4.3 7.5 6.6 33 6 34 38 45 6.7 1.2 6.9 7.6 9.1 Valdivia Applique Fillet Valdivia Broad-line Incised Valdivia Brushed Valdivia Carved Valdivia Combed Valdivia Cord Impressed Valdivia Corrugated Valdivia Cut-and-Beveled Rim Valdivia Embossed Valdivia Excised Valdivia Fine-line Incised Valdivia Fingernail Decorated Valdivia Finger Grooved Valdivia Incised Valdivia Modeled Valdivia Multiple Drag-and-Jab Punctate Valdivia Nicked Broad-line Incised . . . Valdivia Nicked Rib or Nubbin .... Valdivia Pebble Polished Valdivia Pseudo-Corrugated Valdivia Punctate Valdivia Red Incised Valdivia Red Zoned Punctate Valdivia Rocker Stamped Valdivia Shell Stamped Valdivia Zoned Incised Unclassified 273 5 18 27 32 1.7 0.2 19.3 0.3 1.2 1.9 0. 1 0.1 2.3 12 82 2.7 18.8 0.6 1.3 1. 1 0.4 0.2 0.4 1 171 143 787 4 37 52 34 16 32 28 0. 1 8.7 7.2 0. 3 0. 1 0.2 39.8 0.2 1.8 2.6 1.7 0.8 1.6 1. 4 10 402 5 16 48 11 10 0. 1 4.0 4. 1 0. 4 18.5 0.2 0.7 2.2 0. 3 0.5 0.4 344 1.1 11.2 12.4 0.4 0. 3 43.2 1.4 1.5 1. 6 0. 1 0.7 2.0 0.6 50 23 233 10.1 4.6 0.4 0.2 46.8 1.4 0.4 1.6 0.6 1.0 1.4 Machalilla Double-line Incised Machalilla Embellished Shoulder Machalilla Incised Machalilla Red Banded Unclassified Decorated (Machalilla Complex) Ayangue Incised Plain Types (Machalilla Complex) 0. 1 0.7 0. 1 0. 3 0. 7 Grand Totals . . Subtotals : Plain types . . Decorated types Trade pottery . 1,416 1,032 384 0 100.0 440 327 113 0 100.0 1,983 (d) (d) (d) 100.0 2,163 (d) (d) (d) 100.0 798 (d) (d) (d) 100.0 497 (d) (d) (d) 100.0 d Selected samples: totals not meaningful. WHOLE VOLUME APPENDIX 1 203 1 able 6. Frequency of pottery types in surface collections and stratigraphic and test excavations at sites of the Valdivia Phase {figs. 49-53)-Continued G-84: Posorja Pottery types Cut 1 40-50 cm. 50-60 cm. 60-70 cm. 70-80 cm. 80-90 cm. 90-100 cm. No. % No. % No. % No. % No. % No. % Punta Arenas Plain San Pablo Plain Valdivia Plain Valdivia Polished Plain . . . Valdivia Polished Red .... Valdivia Striated Polished Plain 234 5 37 9 7 3 61.3 1.3 9. 7 2. 3 1. 8 0. 8 464 5 40 18 16 10 61.3 0.6 5.4 2.4 2. 1 1. 3 301 11 50 14 23 17 51.5 1.8 8. 6 2.4 3. 9 2.9 572 25 62 28 18 19 53.2 2.4 5.7 2.6 1.7 1.8 204 11 30 19 1 7 53.5 2.9 7.9 5.0 0.3 1.8 206 8 20 17 5 7 53.7 2.0 5.2 4.4 1.3 1.8 Punta Arenas Incised Valdivia Applique Fillet Valdivia Broad-line Incised Valdivia Brushed Valdivia Carved Valdivia Combed Valdivia Cord Impressed Valdivia Corrugated Valdivia Cut-and-Beveled Rim Valdivia Embossed Valdivia Excised Valdivia Fine-line Incised Valdivia Fingernail Decorated Valdivia Finger Grooved Valdivia Incised Valdivia Modeled Valdivia Multiple Drag-and-Jab Punctate. Valdivia Nicked Broad-line Incised. . . . Valdivia Nicked Rib or Nubbin Valdivia Pebble Polished Valdivia Pseudo-Corrugated Valdivia Punctate Valdivia Red Incised Valdivia Red Zoned Punctate Valdivia Rocker Stamped Valdivia Shell Stamped Valdivia Zoned Incised Unclassified 1.0 0.2 14.4 4.2 0. 5 0. 2 0.8 0. 5 0.8 0.2 6 23 109 33 1.0 0.8 3.0 14.4 4.3 1.4 0.5 0. 6 0. 1 0.8 3 14 16 78 30 10 0. 5 2.4 2.7 13.4 5.1 1.7 0.7 0.6 1.2 0.2 7 14 29 208 56 0.6 1. 3 2.7 19.4 5.2 1.7 0. 1 1.3 0.1 0. 1 0.1 1.3 1.8 5.3 13.7 3. 1 1.3 0.3 1. 3 0.5 0. 2 2.0 6.5 18.0 2.0 1.5 1.3 0.2 0.2 Machalilla Double-line Incised Machalilla Embellished Shoulder .... Machalilla Incised Machalilla Red Banded Unclassified Decorated (Machalilla Complex) Ayangue Incised Plain Types (Machalilla Complex) . . . . 0.2 0.2 Grand Totals . . Subtotals: Plain types . . Decorated types Trade pottery 383 295 88 0 100.0 759 553 206 0 100.0 583 416 165 2 100.0 1,077 724 353 0 100.0 381 272 109 0 100.0 384 263 121 0 100.0 767-841-65- 204 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME l Table 6.-Frequency of pottery types in surf ace collections and stratigraphic and test excavations at sites of the Valdivia Phase {figs. 49-53)-Continued G-84: Posorja-Continued G-88: Palmar Norte Pottery types Cut 1-Continued Cut 1: 100-110 cm. 110-120 cm. 120-130 cm. 30-50 cm. 50-70 cm. 70-90 cm. No. % No. % No. % No. % No. % No. % Punta Arenas Plain San Pablo Plain Valdivia Plain Valdivia Polished Plain Valdivia Polished Red Valdivia Striated Polished Plain Punta Arenas Incised Validiva Applique Fillet Valdivia Broad-line Incised Valdivia Brushed Valdivia Carved Valdivia Combed Valdivia Cord Impressed Valdivia Corrugated Valdivia Cut-and-Beveled Rim Valdivia Embossed Valdivia Excised Valdivia Fine-line Incised Valdivia Fingernail Decorated Valdivia Finger Grooved Valdivia Incised Valdivia Modeled Valdivia Multiple Drag-and-Jab Punctate. . . Valdivia Nicked Broad-line Incised Valdivia Nicked Rib or Nubbin Valdivia Pebble Polished Valdivia Pseudo-Corrugated Valdivia Punctate Valdivia Red Incised Valdivia Red Zoned Punctate Valdivia Rocker Stamped Valdivia Shell Stamped Valdivia Zoned Incised Unclassified Machalilla Double-line Incised Machalilla Embellished Shoulder Machalilla Incised Machalilla Red Banded Unclassified Decorated (Machalilla Complex) . Ayangue Incised Plain Types (Machalilla Complex) Grand Totals Subtotals: Plain types Decorated types Trade pottery 165 10 22 11 331 224 107 0 49.9 3.0 6.7 3.3 2.4 2.4 1.5 4.6 5. 1 13.6 1.8 2.7 1.2 0.6 0.9 0.3 100.0 80 4 9 5 1 2 147 101 46 0 54.5 2.7 6.1 3.4 0.7 1.3 4.0 100.0 189 10 32 21 3 4 11 40 42 13 396 264 131 1 47.9 2.5 8.1 5.3 0.7 2.0 1.0 2.8 10. 1 10.6 3.3 2.3 0.2 2.8 0.2 0. 2 100.0 4 21 11 33 6 115 75 40 0 3.5 18.3 9.7 28.8 5.2 1.7 0.8 0.8 27.0 0. 8 2.6 0.8 100.0 58 232 75 126 37 93 2 654 528 126 0 8.9 35.6 11.5 19.3 5.7 0.3 0. 1 100.0 40 121 43 110 34 47 408 348 60 0 9.8 29.8 10.5 27.0 8.4 1.5 0.2 11.5 0.2 0.9 0.2 loo. n WHOLE VOLUME APPENDK 1 205 Table 6.-Frequency of pottery types in surface collections and stratigraphic and test excavations at sites of the Valdivia Phase {figs. 49-53)-Continued G-88: Palmar Norte-Continued Pottery types Cut 1-Cont. Cut 2 Cut 3 Test 90-110 cm. 0-30 cm. 30-60 cm. 60-90 cm. 0-20 cm. No. % No. % No. % No. % No. % No. % Punta Arenas Plain San Pablo Plain Valdivia Plain Valdivia Polished Plain . . . . Valdivia Polished Red .... Valdivia Striated Polished Plain 18 59 34 68 26 7.5 24.2 14.0 28.0 10.7 77 255 226 287 153 6.3 21.0 18.6 23.7 12.6 46 119 101 124 105 7.7 20.2 17.0 20.9 17.7 23 108 65 88 64 6.0 28.4 17.0 23.5 16.8 11 245 39 158 9 2.2 49.8 7.9 32. 1 1.8 13 21 4 1.3 10.8 17.6 28.4 5.4 Valdivia Applique Fillet Valdivia Broad-line Incised Valdivia Brushed Valdivia Carved Valdivia Combed Valdivia Cord Impressed Valdivia Corrugated Valdivia Cut-and-Beveled Rim Valdivia Embossed Valdivia Excised Valdivia Fine-line Incised Valdivia Fingernail Decorated Valdivia Finger Grooved Valdivia Incised Valdivia Modeled Valdivia Multiple Drag-and-Jab Punctate Valdivia Nicked Broad-line Incised . . . Valdivia Nicked Rib or Nubbin .... Valdivia Pebble Polished Valdivia Pseudo-Corrugated Valdivia Punctate Valdivia Red Incised Valdivia Red Zoned Punctate Valdivia Rocker Stamped Valdivia Shell Stamped Valdivia Zoned Incised Unclassified 2.1 1.6 0.4 9.9 0.4 1.2 177 1 2 1.3 0.3 0.1 0.3 14.6 0.1 0. 1 0.4 0.2 0.4 2 1 70 2 2 1.7 0.3 0.2 0.3 0.2 11.8 0.3 0.3 1.0 0.2 0.2 0.8 0.2 0.7 0.2 4.2 1.8 0.4 0.2 0.2 0.4 1.0 0.2 2.0 0. 4 16 2.7 4. 1 2.7 21.7 1.3 2.7 1.3 Machalilla Double-line Incised . . Machalilla Embellished Shoulder . Machalilla Incised Machalilla Red Banded Unclassified Decorated (Machalilla Complex) Ayangue Incised Plain Types (Machalilla Complex) Grand Totals . Subtotals: Plain types. . . Decorated types. Trade pottery. . 243 205 38 0 100.0 1,217 998 219 0 100.0 593 495 98 0 100.0 381 348 33 0 100.0 493 462 31 0 100.0 74 47 27 0 100.0 206 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Table 7.-Chronological distribution of decorative motifs of Valdivia Broad-line Incised Table 8.-Shell artifacts from Machalilla Phase Levels in seriated sequence Motifs Periods Site Cut Level (cm.) 1 2 3 4 5 6 7 8 D G-84 1 40-60 60-80 80-100 100-130 8 18 9 11 3 7 5 7 - 29 39 27 30 10 8 3 6 - - 1 1 C f G-54 G-31 1 H 20-40 40-60 60-80 80-100 100-120 120-130 0-20 20-40 40-60 60-80 80-100 3 10 20 45 29 14 9 11 11 13 16 9 11 31 64 19 10 10 7 2 1 6 11 12 32 69 60 26 3 14 26 17 20 1 17 55 45 29 18 4 2 2 1 2(?) 1 1 | 1 2 1 B G-31 G-88 G-31 G-31 G-31 G-88 F 2 H F H 2 100-120 0-20 20-40 0-30 30-60 120-130 40-60 130-140 60-90 6 26 13 6 3 15 4 19 2 7 8 5 2 2 10 1 8 1 - 11 13 4 2 6 8 4(?) 3(?) 1 - - - A G-31 G-88 G-31 A 3 A 160-180 180-200 200-220 220-240 240-260 260-280 280-300 0-20 300-340 340-360 360-380 380-400 400-420 12 15 9 5 7 2 1 2 - 4 1 1 - - - - Levels in seriated position S u based upon pottery types ers 1 "3 t ti -3 u J _o ;3 P5 .2 Ph Site Cut Level raders and Pc ads acelet sks ,hhook shhooi Small- shhook (cm.) < V bi PP P tn fe M G-l 10 1 10-20 _ _ _ 1 1 - _ 2 0-20 1 - - - 7 - 6 3 0-20 1 - 1 - - 2 1 1 20-30 1 - - - 1 - 1 c 1 30-50 1 - - - 3 - - Trench 1 - 1 - - - 3 4 3 20-30 - - - - - - 3 1 50-70 - - - - - - - 2 20-40 1 - - - 4 2 7 B G-112 Surface A M-28 1 0-20 _ 1 20-40 - - - - - - - G-110: Surface and Tests 3 - - - 8 - - Totals 8 1 1 1 24 7 22 WHOLE VOLUME APPENDIX 1 207 Table 9.-Chronological distribution of decorative motifs of Valdivia Fine-line Incised Table 10.-Chronological distribution of decorative motifs of Valdivia Incised T3 Levels in seriated sequence Motifs | Peric Site Cut: Sec- tion Level (cm.) 1 2 3 4 5 6 Unci. B G-31 J:E J:D J:E J:D J:E 90-120 90-120 120-150 150-180 180-210 120-150 210-240 240-270 150-180 270-300 2 3 2 1 2 2 3 6 1 13 2 1 2 2 _ 1 1 1 1 1 1 5 2 A J:E J:D J:E 300-330 330-360 180-210 360-390 390-420 210-240 240-270 270-300 300-330 330-360 420-450 360-390 11 10 2 5 4 5 13 12 16 5 3 2 5 1 1 3 5 2 1 3 5 4 1 2 1 2 2 1 1 1 1 1 1 3 2 2 1 1 ?iods Levels in seriated sequence Motifs Ph Site Cut Level (cm.) 1 2 3 4 5 6 7 Nicked row D G-84 1 40-60 60-80 80-100 100-130 1 1 2 3 1 3 1 2 - C G-54 G-31 1 H 20-40 40-60 60-80 80-100 100-120 120-130 0-20 20-40 40-60 60-80 80-100 171 254 345 180 275 27 46 58 24 51 12 9 12 1 12 14 19 12 13 9 12 14 35 18 32 38 29 39 20 47 (a) 23 24 31 2 40 2 5 3 3 2 4 2 4 2 19 21 26 3 16 3 2 4 1 5 B G-88 G-31 G-88 F 2 H F H 2 100-120 0-20 20-40 0-30 30-60 120-130 40-60 130-140 60-90 18 169 137 110 28 26 14 10 6 11 40 43 7 7 5 2 8 2 25 85 65 51 26 19 7 16 8 - 3 5 2 1 3 3 - 1 2 1 1 4 9 13 2 2 1 2 1 A G-31 G-88 G-31 A 2 A 160-180 180-200 200-220 220-240 240-260 260-280 280-300 0-20 300-340 340-360 360-380 380-400 400-420 10 2 1 1 2 1 2 1 2 5 3 1 1 - 2 - - 1 Restricted to Period C: separate count for each level unavailable. 208 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Table 11.-Chronological distribution of tetrapod and concave bases and lobed rims in the Valdivia Phase ?a Levels in seriated sequence Bases Lobed rim 'C Ph Site Cut Level (cm.) Tetra- pod Con- cave Bowl Jar G-84 1 40-60 D 60-80 - - - - 80-100 - - - - 100-130 - - - - G-54 1 20-40 _ 3 _ _ 40-60 - - - - 60-80 - 7 - - 80-100 - 13 - - 100-120 - 4 - - C 120-130 - 5 - - G-31 H 0-20 3 1 2 - 20-40 2 14 1 - 40-60 2 2 2 1 60-80 4 4 3 - 80-100 4 - 2 - 100-120 1 2 3 - F 0-20 3 1 4 - 20-40 3 4 1 - G-88 2 0-30 1 3 2 - B 30-60 6 1 1 1 G-31 H 120-130 5 2 4 - F 40-60 1 2 1 - H 130-140 5 1 1 - G-88 2 60-90 1 1 - - G-31 A 160-180 5 2 1 - 180-200 6 3 4 1 200-220 5 1 4 - 220-240 1 1 4 - 240-260 8 - - - 260-280 - 1 - - A 280-300 1 - - - G-88 3 0-20 - - - - G-31 A 300-340 - - - - 340-360 1 - - - 360-380 - - - - 380-400 - - - - 400-420 - - 1 Table 12.-Chronological distribution of decorative motifs of Valdivia Excised Since this type does not occur in Period D levels, this portion of the sequence is omitted from the table. -3 Levels in seriated sequence Motifs Peri Site Cut Level (cm.) 1 2 3 4 5 6 7 C G-54 G-31 1 H 20-40 40-60 60-80 80-100 100-120 120-130 0-20 20-40 40-60 60-80 80-100 1 2 1 2 2 1 2 3 3 1 1 1 2 1 2 1 3 1 1 1 1(?) 1 1 1 2 2 1 1 B G-31 G-88 G-31 G-31 G-31 G-88 F 2 H F H 2 100-120 0-20 20-40 0-30 30-60 120-130 40-60 130-140 60-90 1 2 1 1 1 1 3 3 2 2 1 3 4 - 1 1 3 1 1 A G-31 G-88 G-31 A 2 A 160-180 180-200 200-220 220-240 240-260 260-280 280-300 0-20 300-340 340-360 360-380 380-400 400-420 420-440 1 1 1(?) 3 2 2 - 2 1 - - 1 3 1 1 WHOLE VOLUME APPENDIX 1 Table 13.-Provenience of Valdivia Phase figurine types and alternative kinds of body and leg treatment {fig. 64) 209 Provenience Stone Figurine Types Pottery Figurine Types Body Treatment Leg Treatment Site Cut: Level Palmar Palmar Palmar Valdivia San Buena Punta Red Plain Bulge Tapered Bent Section (cm.) Plain Notched Incised Pablo Vista Arenas G-31 A 40-60 1 160-180 - - 1 _ - _ 2 _ _ 1 _ 180-200 - _ _ _ _ _ _ 1 _ _ _ _ F 0-20 - _ _ _ _ _ _ _ _ _ _ _ 20-40 - - - 2 - - - 3 - 2 1 - 40-60 - - - - - - - - - - - _ H 0-20 - - - 2 - - - 1 1 1 - 1 20-40 - - - 3 - - - - 1 - - - 40-60 - - - - - - - 1 - - 1 - 60-80 - - - 2 - - - 1 2 2 - - 80-100 - - - 1 - - - 1 - - - - 100-120 - - - 4 - - - 2 - 1 1 - 120-130 - - - 4 - - - 5 1 2 2 - 130-140 - - - 3 - - - 8 - 4 - - J:D 90-120 - - - - - - - - - - - - 120-150 - - - 11 3 1 - - - - - - 150-180 - - 2 4 - - - - - - - - 180-210 3 1 - - - - - - - - - - 210-240 1 1 - - - - - - - - - - 240-270 - 3 1 - - - - - - - - - 270-300 3 2 - - - - - - - - - - 300-330 1 - - - - - - - - - - - 330-360 - - - - - - - - - - - - 360-390 5 - - - - - - - - - - - J:E 90-120 - - - - - - - - - - - - 120-150 - - - - - - - - - - - - 150-180 - - - 10 - - - - - - - - 180-210 - - - 7 - - - - - - - - 210-240 - - - 4 - - - - - - - - 240-270 - - 1 8 3 - - - - - - - 270-300 - - - 2 - - - - - - - - 300-330 - - - 1 - - - - - - - - 330-360 1 - - 2 - - - - - - - - 360-390 2 3 1 1 - - - - - - - - 390-420 - - - - - - - - - - - - 420-450 - - - - - - - - - - - - G-54 1 0-20 _ _ _ _ _ - - - - - - _ 20-40 - - - - 1 1 - 1 1 - - - 40-60 - - - - 1 1 - 2 5 1 1 1 60-80 - - - 1(?) 4 3 - 4 9 - 11 1 80-100 - - - - 1 7 - 2 12 - 14 - 100-120 - - - - 1 4 - - - - 2 1 120-130 - - - - 1 - - - 4 - 2 1 1 (NE Ext.): 0-85 - - - - - - - - 5 - 6 1 85-120 - - - - - - - 1 1 - 2 - G-84 1 60-70 - - - - - - 1 - - - - _ 90-100 - - - - - - 1 - 1 - 1 - 100-110 - - - - - - - - - - - - G-88 2 0-30 - - - 1 - - - 2 - 1 - - 30-60 - - - 1 - - - 1 - - - - 60-90 " " " ~ - - - 210 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY Table 14.-Combined vessel forms of the pottery types of the Valdivia Phase {see text table A and fig. 54) VOLUME 1 G-31: Valdivia Com- Cut A bined Forms 160-180 cm. 180-200 cm. 200-220 cm. 220-240 cm. 240-260 cm. 260-280 cm. 280-300 cm. 300-340 cm. 340-360 cm. No. % No. % No. % No. % No. % No. % No. % No. % No. % 1 2 1.8 _ 1 _ _ _ _ - - - 2 2 1.8 4 3.8 - - - - - - - - - - - - - - 3 13 11.9 23 21.7 10 17.2 1 1.2 1 1.3 4 1 0.9 - - - - - - - - - - 1 - 1 - - - 5 19 17. 3 13 12.2 15 25.8 16 19.5 8 10.4 1 - 1 - 1 - 4 - 6 1 0.9 2 1.9 - - 1 1.2 7 9 8.2 14 13.2 7 12.2 9 11.0 7 9.1 3 - 2 - - - - - 8 - - 4 3.8 - - - - - - - - - - - - - - 9 - - - - - - - - - - - - - - - - - - 10 - - - - - - - - - - - - - - - - - - 11 12 13 3 2.7 - - - - 1 1.2 1 1.3 2 1.8 3 2.8 4 6.9 28 34.2 29 37.6 5 - 1 - 4 - 5 - 14 47 42.8 39 36.9 19 32.8 20 24.4 21 27.3 10 - 1 - 7 - 7 - 15 16 17 2 1 4 1.8 0.9 3.6 - - - - - - 1 1.3 _ _ _ _ _ _ 3 3.9 18 19 3 2.7 3 2.8 3 5.1 6 7.3 6 7.8 5 - 1 - 3 - - - 20 21 22 1 0.9 1 0.9 - - - - - - - - - - - - - - 23 - - - - - - - - - - - - - - - - - - Totals 110 100.0 106 100.0 58 100.0 82 100.0 77 100.0 25 U) 7 (a) 16 (a) 16 (a> a Sample too small for valid percentage calculations. WHOLE VOLUME APPENDIX 1 Table 14.-Combined vessel forms of the pottery types of the Valdivia Phase {see text table A and fig. 54)-Continued 211 G-31: Valdivia- -Continued Combined Cut A-Continued CutF CutH Forms 360-380 380-400 400- -420 0-20 cm. 20-40 cm. 40-60 cm. 0-20 cm. 20-40 cm. 40-60 cm. cm. cm. cm. No. % No. % No. % No. % No. % No. % No. % No. % No. % 1 - - _ _ 3 0.9 2 0.9 1 0.4 2 - - - - - - 1 0.3 - - - - - - - - 1 0.4 3 - - - - - - 7 2.3 5 2.6 - - 9 5.7 13 6. 1 22 9.7 4 - - - - 1 - - - - - - - - _ - - - - 5 1 - 1 - - - 54 17.4 20 10.7 3 11. 1 13 8. 3 27 12.7 16 7.0 6 - - - - - - 10 3.2 9 4.8 - - 7 4. 5 6 2.8 11 4. 8 7 - - 2 - 1 - 42 13.5 19 10. 1 2 7.4 39 24.8 39 18.4 34 15.0 8 - - - - - - 5 1.6 - - - - 2 1.2 14 6.6 9 3.9 9 - - - - - - 3 0.9 - - 1 3.7 3 1.9 4 1.8 5 2.2 10 - - - - - _ - - - - - - _ _ _ - - - 11 - - - - - - 20 6.4 8 4.2 - - 8 5.1 12 5.6 8 3.5 12 - - - - - - 3 0.9 - - - - - - 1 0.5 - - 13 3 - - - - - 12 3.8 17 9.0 2 7.4 4 2.6 1 0.5 4 1.7 14 4 - - - - - 23 7.4 18 9.6 12 44.5 39 25.1 45 21.2 55 24.3 15 - - - - - - 63 20.4 47 25.0 5 18.5 7 4.5 9 4.2 17 7.4 16 - - - - - - 5 1.6 1 0.5 - - 1 0.6 1 0.5 3 1.3 17 1 - - - - - 49 15.9 34 18.2 2 7.4 5 3.2 9 4.2 11 4.8 18 19 20 - - - - 1 - 8 2.6 9 4.8 - - 7 4.5 8 3.7 9 3.9 - - _ _ - _ 3 0.9 _ _ _ _ 5 3.1 8 3.7 8 3.5 21 - - - - - - - - - - - - 3 1.9 5 2. 4 2 1.8 22 - - - - - - - - 1 0.5 - - 5 3.1 8 3.7 10 4. 4 23 - - - - - - - - - - - - - - 1 0.5 - - Totals 9 (a) 3 (a) 3 (a) 311 100.0 188 100.0 27 100.0 157 100.0 213 100.0 226 100.0 Sample too small for valid percentage calculations. 842-039^6E -16 212 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY Table 14.-Combined vessel forms of the pottery types of the Valdivia Phase {see text table A and fig. 54)-Continued VOLUME 1 G-31: Valdivia-Continued G-54: Buena Vista Combined CutH- ?Continued Cutl Forms 60-80 cm. 80-100 cm. 100- 120 cm. 120- 130 cm. 130- 140 cm. 20- ?0 cm. 40- SO cm. 60-80 cm. 80-100 cm. No. % No. % No. % No. % No. % No. % No. % No. % No. % 1 3 2.8 _ _ _ 2 - - - - 1 0.8 1 0.9 - - - - - - - - - - 3 9 8.2 13 8.4 8 6.4 29 27.9 24 12.7 9 8.3 5 3.3 17 5.3 20 3.3 4 - - - - - - - - - - - - - - - - - - 5 19 17.4 27 17.6 18 14.5 12 11.7 31 16.3 28 25.9 32 21.2 48 15.0 62 10.3 6 3 2.8 4 2.6 2 1.6 2 1.9 - - - - 2 1.3 2 0.6 2 0.3 7 20 18.4 25 16.2 14 11.2 21 20.3 27 14.2 3 2.8 3 2.0 5 1.5 2 0.3 8 2 1.8 1 0.6 - - - - 1 0.5 7 6.4 10 6.6 27 8.4 25 4. 1 9 10 11 12 13 2 1.8 1 0.6 - - - - - - 12 11.2 19 12.5 46 14.4 140 23.3 2 1.8 8 5.1 3 1 7 2.4 0.8 5.5 1 0.9 - - - - 2 1.3 4 1.2 6 1.0 3 2.8 5 3.2 _ _ 16 8.5 2 1.8 1 0.6 1 0.3 - - 14 16 14.7 31 20.0 43 34.4 9 8.7 68 35.8 - - - - - - - - 15 6 5.5 13 8.4 2 1.6 6 5.8 1 0.5 3 2.8 1 0.6 12 3.7 2 0.3 16 - - 1 0.6 - - 1 0.9 1 0.5 4 3.7 3 2.0 6 1.9 10 1.6 17 13 11.9 13 8.4 14 11.2 9 8.7 9 4.7 29 26.9 54 36.0 103 32.2 236 39.4 18 3 2.8 6 3.9 9 7.2 8 7.7 11 5.8 4 3.7 5 3.3 14 4.3 16 2.7 19 - - - - - - - - - - 6 5.6 2 1.3 12 3.7 31 5.2 20 3 2.8 3 1.9 3 2.4 1 0.9 1 0.5 - - - - 1 0.3 3 0.5 21 1 0.9 1 0.6 - - - - - - 1 0.9 12 8.0 22 6.9 43 7.2 22 7 6.4 2 1.3 - - 1 0.9 - - - - - - - - - - 23 - - 1 0.6 - - - - - - - - - - 1 0.3 3 0.5 Totals 109 100.0 155 100.0 125 100.0 104 100.0 190 100.0 108 100.0 151 100.0 321 100.0 601 100.0 WHOLE VOLUME APPENDIX 1 213 Table 14.-Combined vessel forms of the pottery types of the Valdivia Phase {see text table A and fig. 54)-Continued G-54: Buena Vista-Cont. G-84: Posorja Site G-88: Palmar Norte Combined Cut 1-Continued Cut 1 Cut 2 Forms 100-120 cm. 120-130 cm. 40- -60 cm. 60- 80 cm. 80-100 cm. 100- 130 cm. 0- 30 cm. 30-60 cm. 60- -90 cm. No. % No. % No. % No. % No. % No. % No. % No. % No. % 1 - - _ _ _ _ 15 10. 1 3 3.3 2 - - - - - - - - - - - - - - 2 2. 1 2 4.4 3 4 21 5.3 15 7.5 4 5.0 4 3.4 1 1.9 7 8.9 3 2.0 5 5.4 5 10.9 5 82 20.4 39 19.4 3 3.8 2 1.7 3 5.8 2 2.5 21 14. 1 16 17.4 5 10.9 6 - - - - 1 1.2 2 1.7 2 3.8 2 2.5 6 4. 1 4 4.4 - - 7 7 1.7 4 2.0 26 32.6 31 26.3 14 26.9 19 24.0 23 15.4 10 10.9 11 24.0 8 32 8.0 18 8.9 5 6.2 8 6.7 4 7.7 8 10.2 _ _ _ _ _ _ 9 56 14.0 15 7.5 2 2.5 2 1.7 1 1.9 1 1.2 - - - - - _ 10 - - - - 5 6.2 15 12.7 8 15.4 5 6.4 _ _ _ _ _ _ 11 3 0.7 5 2.5 - - - - - - - - 19 12.8 7 7.6 1 2. 1 12 - - - - - - - - - - - - 2 1.3 _ _ _ _ 13 4 0.9 - - - - - - - - - - 5 3.3 15 16.3 11 24.0 14 - - - - - - - - - - - - 3 2.0 1 1.0 6 13.0 15 3 0.7 4 2.0 5 6.2 13 11.0 4 7.7 8 10.2 13 8.8 13 14.2 - - 16 12 3.0 7 3.5 2 2.5 5 4.3 2 3.8 1 1.2 19 12.8 8 8.7 1 2.1 17 138 34.4 76 37.8 4 5.0 7 5.9 - - 2 2.5 10 6.7 3 3.3 1 2. 1 18 11 2.7 6 2.9 2 2.5 4 3.4 - - 1 1.2 7 4.7 2 2.1 2 4.4 19 9 2.2 1 0.5 - - - - - - - - - - - - - _ 20 - - 1 0.5 - - - - - - - - 2 1.3 3 3.3 1 2.1 21 16 4.0 10 5.0 16 20. 1 20 17.0 10 19.3 18 22.9 1 0.6 - - - _ 22 - - - - 1 1.2 1 0.8 - - 1 1.2 - - - - - - 23 8 2.0 - - 4 5.0 4 3.4 3 5.8 4 5.1 - - - - - - Totals 402 100.0 201 100.0 80 100.0 118 100.0 52 100.0 79 100.0 149 100.0 92 100.0 46 100.0 214 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY Table 15.-Frequency of combined vessel forms and unusual appendages in sites of the Machalilla Phase {fig. 90) VOLUME 1 G-l 10: La Cabuya Combined vessel forms Cut 1 Cut 2 Cut 3 10-20 cm. 20-30 cm. 30-50 cm. 50-70 cm. 0-20 cm. 20-40 cm. 0-20 cm. No. % No. % No. % No. % No. % No. % No. % 1 1 4.3 3 10.4 6 13.1 9 20.0 23 9.7 19 8.5 2 8.7 2 2 8.7 6 20.8 3 6.5 1 2.2 22 9.3 23 10.3 1 4.3 3 1 4.3 4 13.8 - - 1 2.2 9 3.8 20 8.9 - - 4 1 4.3 2 6.9 1 2.1 2 4.4 17 7.1 13 5.8 2 8.7 5 1 4.3 2 6.9 3 6.5 6 13.4 24 10.2 39 7.5 2 8.7 6 1 4.3 - - - - - - 4 1.7 - - 1 4.3 7 - - 1 3.4 - - - - 8 3.4 2 0.8 1 4.3 8 1 4.3 - - - - 1 2.2 15 6.3 8 3.6 - - 9 - - - - - - - - 1 0.4 - - - - 10 - - - - - - - - 1 0.4 3 1.3 1 4.3 11 5 21.9 6 20.8 14 30.5 13 29.0 53 22.4 43 19.2 5 21.7 12 - - 1 3.4 1 2. 1 - - 1 0.4 - - - - 13 1 4.3 1 3.4 9 19.7 3 6.6 13 5.5 12 5.4 4 17.5 14 1 4. 3 - - - - - - 4 1.7 - - - - 15 4 17.5 2 6.9 7 15.2 7 15.6 20 8.4 29 12.9 4 17.5 Unci. 4 17.5 1 3.4 2 4.3 2 4.4 22 9.3 13 5. 8 - - Totals 23 100.0 29 100.0 46 100.0 45 100.0 237 100.0 224 100.0 23 100.0 Unusual Appendages: Stirrup Spout Cylindrical Spout - - - - - - - - 2 3 - 1 1 - - - Boss 1 WHOLE VOLUME APPENDIX 1 Table 15.-Frequency of combined vessel forms and unusual appendages in sites of the Machalilla Phase {fig. 90)-Continued 215 G-110: La Cabuya-Cont. M-28: Machalilla Cemetery Machalilla Phase trade sherds at Valdivia Phase sites Combined Tests and surface Trench Cut 1 vessel G-31: Combined levels of Cut J forms 0-20 cm. 20-40 cm. G-54: Misc. tests and surface No. % No. % No. % No. % No. % No. % 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Unci. 52 57 47 32 100 20 2 20 2 3 119 2 35 38 26 9.4 10.3 8.5 5.7 18.1 3.6 0.3 3.6 0.3 0.5 21.5 0.3 6.3 6.9 4.7 14 26 13 12 18 2 10 2 46 1 10 5 14 6 7.8 14.5 7.3 6.6 10.1 1.1 5.6 1.1 25.8 0.5 5.6 2.8 7.8 3.4 22 41 29 10 8 11 32 7 12 1(?) 1 23 11.2 20.8 14.7 5.1 4.0 5.6 16.3 3.5 6. 1 0.5 0.5 11.7 4 14 22 6 3 8 20 6 9 4 1 6 3.9 13.6 21.4 5.8 2.9 7.8 19.4 5.8 8.7 3.9 1.0 5.8 8 1 2 - 3 4 1 5 13 12 4 5 6.4 8.5 2.1 10.6 27.7 25.6 8.5 10.6 Totals 555 100.0 179 100.0 197 100.0 103 100.0 11 w 47 100.0 Unusual appendages: Stirrup spouts Cylindrical spouts Boss 5 2 2 - - - 4 3 2 - 1 - - - 1 - Sample too small for valid percentage calculations. 216 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Table 16.-Frequency of pottery types in surface collections and stratigraphic excavations at sites of the Machalilla Phase {fig. 89) M-28: Machalilla Cemetery G-l10: La Cabuya Pottery types Cut 1 Cutl 0-20 cm. 20-40 cm. 0-10 cm. 10-20 cm. 20-30 cm. 30-50 cm. No. % No. % No. % No. % No. % No. % Cabuya Plain 217 54 95 265 112 475 13 13.32 3.32 5.85 16.30 6.84 29.20 0.80 131 13 59 189 73 308 2 13.7 1.4 6.2 19.7 7.6 32.3 0.2 24 5 16 8 15 27.6 5.8 18.4 9.2 17.2 69 16 2 77 21 107 21.0 4.8 0.6 23.4 6.4 32.6 64 8 2 80 45 96 17.8 2.2 0.6 22.2 12.5 26.8 47 25 42 56 100 12.2 6. 5 PLAIN Machalilla Plain Machalilla Polished Plain .... Machalilla Striated Polished Plain . 10.9 14.6 26.0 Cabuya Finger Pressed Rim . . . Chorerra Incised Machalilla Burnished Line .... Machalilla Double-line Incised . . Machalilla Embellished and Red 1(?) 143 2 3 4 1 2 1 166 53 2 3 0.06 8.80 0.12 0. 18 0.24 0.06 0. 12 0.06 10.20 3.26 0.12 0. 18 1 1 54 1 94 26 1 1 0.1 0.1 5.6 0.1 9.8 2.7 0.1 0.1 3 13 2 3.4 15.0 2.3 8 1 1 5 2 1 14 4 1 2.4 0.3 0.3 1.5 0.6 0.3 4.3 1.2 0.3 2 9 2 3 31 15 1 0.6 2.5 0.6 0.8 8.6 4.2 0.3 4 2 2 11 4 1 3 60 25 2 1.1 0.5 0.5 2.9 1. 1 0.3 0.8 15.6 6.5 0.5 DECORATED Machalilla Embellished Shoulder . Machalilla Finger Punched .... Machalilla Incised and Punctate . Machalilla Incised and Red Zoned . Machalilla Punctate and Red Zoned Machalilla Red Banded: Machalilla Red Incised Unclassified Decorated: Chorrera Phase types Miscellaneous TRADE Valdivia Phase types: Valdivia Applique Fillet . . . Valdivia Broad-line Incised . . Valdivia (?) Corrugated . . . Valdivia Incised Valdivia Pebble Polished . . . Valdivia Punctate Plain Pottery types Kotosh-like Zoned Punctate . . . 2 4 2 1 5 2 0.12 0.24 0. 12 0.06 0. 31 0.12 1 1 2 0.1 0.1 0.2 1 1.1 - - 1 0.3 - - Subtotals : Decorated Types .... Trade Pottery 1,628 1,231 381 16 100. 00 958 775 179 4 100.0 87 66 20 1 100.0 329 292 37 0 100.0 359 295 63 1 100.0 384 270 114 0 100.0 WHOLE VOLUME APPENDLX 1 217 Table 16.-Frequency of pottery types in surface collections and stratigraphic excavations at sites of the Machalilla Phase {fig. 89)-Continued G-110: La Cabuya-Continued G-112 Cut 1 -Con. Cut 2 Cut 3 Pottery types Trench 1 Surface Grand Total 50-70 cm. 0-2C cm. 20-40 cm. 0-20 cm. No. % No. % No. % No. % No. % No. % -Cabuya Plain 55 14. 1 429 16.48 612 23.30 56 21.9 352 20.62 110 38.8 . Chorrera Plain 23 5.9 125 4.80 289 11.00 15 5.8 66 3.86 90 31.7 - z Machalilla Plain .... - _ 32 1.23 8 0.30 - - 25 1.46 - - - PLA Machalilla Polished Plain . . . 52 13.4 694 26.62 446 16.98 52 20.3 317 18.52 26 9.2 - Machalilla Polished Red .... 60 15.4 468 18.00 241 9. 17 37 14.4 42 2.45 32 11.3 - Machalilla Striated Polished Plain 89 22.8 485 18.65 580 22.04 26 10.2 537 31.42 3 1.1 _ Unclassified Plain - 6 1.5 80 3.08 37 1.41 9 3.5 10 0.58 - - - Cabuya Black-on-White .... - - 5 0.19 2 0.07 - - - - - - - Cabuya Finger Pressed Rim . . - - 5 0.19 3 0. 10 - - - - - - - Chorerra Incised 1(?) 0.2 - - 1 0.04 2 0.8 - - - - - Machalilla Burnished Line . . . 5 1.3 4 0. 15 3 0. 10 - - 2 0. 11 - - - Machalilia Double-line Incised . 1 0.2 8 0.32 15 0. 57 - - 3 0.17 - - - Machalilla Embellished and - - 3 0. 12 1 0.04 1 0.4 - - - - - Machalilla Embellished Shoulder 14 3.6 30 1 1.15 0.04 53 2.01 2 0.8 51 2.98 7 2.4 - P Machalilla Finger Punched . . . - Machalilla Incised - - 11 0.42 3 0.10 1 0.4 1 0.06 - - - DECORA Machalilla Incised and _ - - _ - - - - - - - - - Machalilla Incised and Red 3 0.8 7 0.27 23 2 0.87 0.07 4 1.5 5 0.29 3 1 1.1 0.3 - - Machalilla Punctate and Red 4 1.0 12 0.46 30 1. 14 1 0.4 9 0.52 2 0.7 - Machalilla Red Banded: Technique 1: Fine .... 42 10.8 126 4.86 208 7.93 34 13.3 199 11.62 5 1.7 Technique 2: Wide .... 26 6.7 57 2.19 56 2. 12 13 5.1 72 4.23 5 1.7 - Machalilla Red Incised .... - - 1 0.04 3 0.10 - - 1 0.06 - - - Unclassified Decorated: Chorrera Phase types . . . - - 1 0.04 1 0.04 - - - - - - - 8 2.1 10 0. 38 10 0.38 3 1.2 9 0.52 " " - Valdivia Phase types: Valdivia Applique Fillet . . - - 1 0.04 - - - - - - - - - Valdivia Broad-line 1 0.2 1 2 0.04 0.08 1 1 0.04 0.04 - - 6 0.35 - - - .ADE - Valdivia(?) Corrugated . . . - - - - - - - - - - - - - P5 H - - 1 0.04 - - - - 1 0.06 - - - Valdivia Pebble Polished . . - - 3 0. 12 - - - - 1 0.06 - - - - - - - - - - - - - - - - Plain Pottery types .... - - - - - - - - 1 0.06 - - - Kotosh-like Zoned Punctate . . - - - " 1 0.04 390 100.0 2,602 100. 00 2,630 100. 00 256 100.0 1,710 100. 00 284 100.0 11,617 Subtotals : 279 - 2,233 - 2,176 - 186 - 1,339 - 261 - 9,403 Decorated types . . . 110 - 361 - 451 - 70 - 362 - 23 - 2,171 Trade Pottery .... 1 - 8 3 0 9 0 43 218 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY Table 17.-Stone artifacts and natural stone from Machalilla Phase VOLUME 1 Levels in seriated V .*2 øo S5 Periods position based upon pottery types Abraders Blades Choppers Gravers Grinding Ston Hammerstones "Jaketown Perforators" Paint Stones Pebble Polishe Reamers Saws Scrapers Burnt Sandsto Fragments Coquina Fragments Mudstone Fragments Chert, Quartz Chalcedony Quartz Flak c 0 -*J5 Site Cut Level (cm.) Natura Wate Pebb G-110 1 10-20 1 1 1 1 4 4 12 21 9 2 0-20 1 5 1 2 - 2 - - 1 1 1 3 - - 15 100 10 3 0-20 - - - - - 1 - - - - - - - - 3 3 4 C 1 20-30 - 2 - 2 - - 1 - 4 - - 3 - - 3 8 1 1 30-50 - 2 - - - 3 - - - 1 - - - - 2 12 3 Trench 1 3 1 20-30 1 50-70 - - - - - - - - 1 - 1 - - - - - 2 - - - 1 _ _ _ 7 _ _ 1 _ _ 2 _ 6 1 2 20-40 - 4 1 1 - - 3 1 9 2 1 - 4 - 12 85 5 B G-112 Surface - - - - - - - - - - - - - - - - - 1 0-20 _ 2 _ 1 _ _ _ _ 1 _ 1 _ 1 _ 13 5 _ A M-28 1 20-40 - 1 - 1 - - - - = - 1 - - 14 - 1 G-l 10: Surface & Tests 2 - - - 1 - - - - - 7 - Totals 3 17 2 7 1 6 6 8 17 4 6 11 5 6 74 247 36 Appendix 2 SKELETAL REMAINS FROM SITES OF VALDIVIA AND MACHALILLA PHASES Juan R. Munizaga r~THE skeletal remains associated with the Valdivia and Machalilla Phases are of considerable poten- tial interest, in spite of the small size and poorly preserved condition of the sample, because they con- stitute the earliest such remains from the Ecuadorian coast. Material has been segregated by Phase, fol- lowing the organization of the cultural discussion, but this separation is justifiable also by the marked physical differences between the two populations in spite of their geographical and temporal juxtaposition. Although post-cranial fragments exist for many of the individuals, time allowed examination only of the skulls. The present study was undertaken in the Division of Physical Anthropology of the U.S. National Mu- seum during 1961. I should like to express my appre- ciation to the John Simon Guggenheim Foundation for fellowship support, and to the staff of the Depart- ment of Anthropology, U.S. National Museum, Smith- sonian Institution for many courtesies during my stay. Particular thanks are due Dr. T. Dale Stewart, who encouraged and directed the work, and who has checked the accuracy of the translation of this manu- script from the original Spanish and to Mr. Donald J. Ortner, Division of Physical Anthropology, who de- voted considerable effort to reconstruction and pho- tography of the skulls. Mr. Olaf Holm of Guayaquil kindly granted permission for inclusion of measure- ments and observations on skulls from the Valdivia Phase site of San Pablo, which were made during my brief stay in Guayaquil. THE VALDIVIA PHASE Discovery of human skeletal remains at the Buena Vista site (G-54) of the Valdivia culture has a dual importance for South American physical anthropol- ogy. First, it provides one of the few undeformed cranial series on the continent for which both cultural association and chronological position (by carbon-14 dating) are known. Second, the population repre- sented had a knowledge of potterymaking at so early a date that it may constitute the introducer of this craft. These factors suggest two possible lines of inquiry: 1) Was so marked a change in culture as the introduc- tion of potterymaking accompanied by a change in physical type? and 2) To what extent is the physical type of the Valdivians represented in later populations on the coast of Andean South America? In the fol- lowing pages, we will attempt to find answers to these two questions. Because of the extremely fragmentary condition of the material, requiring time-consuming restoration before study, and because of the limited time period available for the analysis, only the crania could be examined and compared. The poor state of preserva- tion makes detailed description of methods of analysis important, as a basis for judging the validity of morphological observations and for understanding the basis for selection of certain measurements. A description of each skull is also given in detail because of the potential value of the series for future comparative studies. Methods Reconstruction.-The skull fragments in general show clean breaks. Reconstruction consequently did not present a difficult problem. Where pieces were missing, and it was necessary to give the vault 219 220 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 more strength to prevent damage during study, gaps were filled with plaster. In one case only (to be described below), plaster was used for inter- pretative reconstruction of a frontal bone. Restora- tion was limited to the fitting together of fragments, with minimal reconstruction of missing parts, to eliminate as far as possible the subjective element and to give measurements and observations maximum degree of validity. Measurements, indices and morphological ob- servations.-Anterior-posterior diameter, maximum width and minimum frontal diameter were measured following the technique of Hrdlicka (1939). The anterior-posterior diameter differs slightly from that of other authors in that it is obtained by measuring the distance from glabella to the most posterior portion of the skull, whether this is in the midline or at the sides of the occipital. Basio-bregmatic height could not be determined because of the absence of the basion on all the specimens, and auricular-bregmatic height was therefore substituted. In the very incomplete skulls, this measurement was calculated by using outline drawings made with the Schwarz stereograph. Vault thickness was measured in two places on the parietal, one in the region of obelion adjacent to the superior longitudinal sinus, and the other at the point of maximum thickness between the midline and the parietal eminence. Cranial form was determined by the horizontal cranial index and the mean auricular height index. The latter index, elaborated by Stewart, is a varia- tion of the classic mean height index (Stewart, 1942) and differs from it in that auricular-bregmatic height is employed in place of basio-bregmatic height (pers. commun.). Stewart has developed for this index the following categories, which are employed in the present study: Low -66 Medium 67-71 High 72- Determination of skull size was difficult. Because of the poor condition of the crania neither capacity nor cranial module could be measured, although these are generally accepted as the most reliable indicators of size. In order to surmount this problem, we have employed a variant of the cranial module. Just as Stewart has elaborated his mean ear height index, we believe that it is possible to define an auricular cranial module. To calculate it, auricular-bregmatic height is substituted for basio-bregmatic height in the classic formula (length -J- width+height) /3. Although we have not as yet arrived at a subdivision of cranial size based on the auricular module, the range of variation represented in six North American indigenous populations was from 136 to 156 mm. This can be used as a basis for comparison of the sizes of the skulls of the Buena Vista series. Morphological observations.-In addition to the classical morphological characters, records were made of the presence or absence of such features as perfora- tions of the floor of the auditory meatus, ear exostoses, apical and Inca bones, parietal foramina, and the emissary branch of the middle meningeal artery. There is high probability that these five features follow rather simple genetic patterns. The last, the emissary branch of the middle meningeal artery (which will be abbreviated as "emissary of the M.M." in the following pages), has been the subject of recent study and appears to be a useful character for the study of precolumbian populations (Munizaga, 1963). Material Description of individual skulls.-The 15 skulls represented in the Buena Vista series have the following metrical and morphological characteristics: BV-1 (USNM 381930): Adult male, in- complete. In spite of the absence of the base, of several fragments of both right and left parietals and parts of the right sphenoid, almost complete recon- struction could be achieved (pi. 192 a-c). The principal metrical landmarks are present and the morphological characters are readily observable. The skull is undeformed, brachycranial (index, 81.3), high (mean auricular height index, 73.3), large (auricular cranial module, 150.3 mm.), with endo- cranial sutures obliterated and external sutures vestigial except for the right squamosal. With regard to morphological features, the Inca bone is absent, as are ear exostoses and the emissary of the M.M. The floor of the left auditory meatus contains a perforation of about 2 mm. diameter but this does not appear to be a normal dehiscence (pathological?). Parietal foramina symmetrical. Alterations of path- ological origin include a wrinkle-like depression in the median part of the frontal (cicatrice?). Thickness of the vault is 5 mm. at the median point and 6 mm. at the lateral point. BV-2 (USNM 381931): Young adult female; incomplete. This specimen is less complete than BV-1, lacking entirely the base, extensive portions of the lateral regions and the left half of the occipital; nevertheless, reconstruction was complete enough to permit description (pi. 192 d-f). Maximum anterior- posterior diameter was determined on the right half of the occipital, which does not appear to have dis- torted the result. Inspection shows the skull to be Whole volume APPENDIX 2 221 undeformed, brachycranial (index, 81.7), high (index, ^3.5), large (auricular cranial module, 150.0 mm.), with exocranial and endocranial sutures open both ecto- and endocranially. Absence of the relevant parts prevented observation of the floor and interior of the auditory meatus. The emissary of the M.M. is absent on the left side (the right side is missing). Neither parietal foramina nor an Inca bone are pres- ent; however, there is an apical bone. Pathological features include scarring of the orbital roof (pi. 195f) and arthritis in the right glenoid fossa. Thickness of the parietal is 5 mm. in the midline and 5 mm. at the lateral point. 3. BV-3 (USNM 381932): Young adult male; incomplete. The bones were coated with a hard calcium deposit, which was removed chemically. Reconstruction revealed the absence of the left half of the frontal, of both sphenoids, part of the left temporal squama, and the right side of the base (pi. 194 a-c). In spite of this, it was possible to iden- tify all the principal landmarks except glabella. The position of glabella could be estimated by means of the right superorbital curve which reaches to the mid- line. Inspection showed the skull to be undeformed, brachycranial (index, 82.2), high (index, 79.2), of medium size (auricular cranial module, 143.3 mm.), with the sutures beginning to close on the interior. Auditory exostoses, perforations in the floor of the auditory meatus, and the emissary of the M.M. are absent. Present are a lateral Inca bone (pi. 195e) and bilaterally asymmetrical parietal foramina (left> right). No pathological lesions were observed. Thickness of the parietal is 5 mm. in the midline and 7 mm. at the lateral point. BV-4 (USNM 381933): Young adult male; very incomplete. Absence of the bregmatic portion and the lower half of the occipital made determina- tion of the sagittal contour impossible. Also missing was all of the base (pi. 193 a-c). Inspection showed the skull to be undeformed, symmetrical, brachy- cranial (index, 81.3), apparently high and large. Endocranial sutures are obliterated, while those on the exterior remain open. Perforations in the floor of the auditory meatus and the emissary of the M.M. are absent. Each auditory meatus contains a small oblong elevation (exostosis?). The Inca bone is of classic type, and there are faint parietal foramina. The only pathological sign is a possible arthritic con- dition of the glenoid condyles. BV-5 (USNM 381394): Adult female; very fragmentary. Reconstruction permitted determina- tion of the sagittal contour, but not the horizontal contour because of the absence of the temporal and part of the left parietal. Also missing are the base and Bregma <- Glabella Figure 110.-Stereograph drawing of skull BV-5, G-54, Burial 8. 222 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 the sphenoids. Maximum width was consequently calculated by assuming symmetry. Auricular-breg- matic height was obtained from a stereographic drawing (fig. 110). Inspection showed the skull to be undeformed, brachycranial (index, 89.1), high (index, 72.6), medium sized (auricular cranial module, 142.6 mm.), with endocranial sutures completely closed and ectocranial sutures in the process of closure. Observa- tions of the right side showed absence of exostoses of the auditory meatus, and absence of perforations in the floor of the auditory meatus. Present are a double emissary of the M.M. and bilaterial parietal foramina. Neither an apical nor an Inca bone is present, and no pathological lesion was observed. Thickness of the parietal bone is 5 mm. at the midline and 8 mm. at the lateral point. 6. BV-6 (USNM 381936): Adult female; in- complete. Except for the base, almost complete reconstruction was possible (pi. 193 d-f). Auricular- bregmatic height was obtained from a stereographic drawing (fig. 111). The skull is undeformed, brachy- cranial (index, 83.6), high (index, 72.6), large (auricular cranial module, 147.6 mm.), with en- docranial sutures closed and ectocranial sutures in process of closure. Neither exostoses nor perforations of the floor of the auditory meatus are present. Existence of the emissary of the M.M. could not be determined in the absence of the relevant bones. A left parietal foramen was present. Inca and apical bones are absent. No pathological lesion was detected. Thickness of the parietal measured 5 mm. in the midline and 8 mm. at the lateral point. BV-7 (USNM 381937): Adult male; very in- complete. Only the sagittal curve could be recon- structed because of the absence of the temporals and lower portions of the parietals (pi. 195 a-c). The auricular point was estimated for determination of basio-bregmatic height. The skull is undeformed, and has a sagittal profile similar to those previously described. Morphological features of the temporal bones could not be observed because of the absence of these bones. Inca and apical bones are absent. Double parietal foramina were present on the right side. No pathological lesion was evident. BV-8 (USNM 381938): Adult female; incom- plete. Reconstruction made it possible to take meas- urements and observe the principal morphological characters in spite of the absence of the base and left temporal (pi. 194 d-f). Auricular-bregmatic height was determined from a drawing (fig. 112). The skull is undeformed, brachycranial (index, 81.1), of inter- mediate height (index, 69.4), on the small side (auricu- lar cranial module 138.3 mm.), with endocranial sutures closed and ectocranial sutures in the process Bregma Lambda Glabella -^ Upper Nuchal Line (Inion) Figure 111.-Stereograph drawing of skull BV-6, G-54, Burial 9. whole volume APPENDIX 2 223 01 closure. The right temporal exhibits neither ex- ostoses nor perforations in the floor of the auditory meatus. The emissary of the M.M. is present on the right side. There is no Inca bone. The parietal foramen is marked on the left side but slight on the right side. No pathological lesion was observed. Thickness of the parietal is 5 mm. at the midline and 7 mm. at the lateral point. 9. BV-9 (USNM 381939): Adult male; very in- complete. In spite of the highly fragmentary condi- tion, reconstruction was sufficient to permit definition of the sagittal profile. Absence of the midportion of the frontal and the parietals made it necessary to recreate these parts in plaster, which was done by taking as a basis certain points of these bones that approach the midline and using as a guide the con- tour of this region characteristic of the other skulls of the series. The position of bregma, consequently, is not entirely certain, but it is probably approximately correct. The horizontal contour could not be directly measured in the absence of part of the parietal and temporal, and the transverse diameter was calculated by assuming symmetry. Auricular-bregmatic height was calculated from a drawing. Inspection showed the skull to be undeformed, brachycranial (index, 78.1), high (index, 75.4), large (auricular cranial module, 149.6 mm.), with endocranial sutures com- pletely closed and ectocranial sutures only partly closed. The left temporal shows neither exostoses nor perforations of the floor of the auditory meatus. The existence of the emissary of the M.M. could not be observed in the absence of that portion of the skull. A parietal foramen was limited to the left side. Neither Inca nor apical bones occur. No pathological lesion was evident. Thickness of the parietal is 6 mm. in the midline and 7 mm. at the lateral point. BV-10 (USNM 381940): Adult male; very fragmentary. Restoration was limited to reconstruc- tion of the anterior part of the right parietal, the superorbital area and a fragment that joins these portions along the right external frontal line. The skull appeared similar to others in the series. Several fragments showed sutures closed on the interior but open on the exterior. Morphological features ob- servable include presence of the emissary of the M.M. on the left side (the right side is missing). Auditory meatuses do not show exostoses; the floor of the right meatus has a small perforation. The existence of an Inca bone and parietal foramina could not be ascer- tained. No pathological features were noted. BV-11 (USNM 381941): Adult female; very incomplete. Reconstruction permitted establishment of the sagittal and horizontal contours in spite of the absence of fragments of both lateral norms and the frontal. Also missing is the base and the lower half of the occipital. Basio-bregmatic height was cal- Bregma Glabella Upper Nuchal Line Union) Figure 112.-Stereograph drawing of skull BV-8, G-54, Burials 1-7. 224 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 culated from a drawing (fig. 113). The skull is undeformed, brachycranial (index, 88.6), of inter- mediate height (index, 69.2), on the small side (auricular cranial module, 141.3 mm.), with en- docranial sutures closed and ectocranial sutures in process of closure. The right temporal shows neither exostoses nor perforations in the floor of the auditory meatus (the left is missing). The existence of the emissary of the M.M. could not be ascertained. Neither Inca bone nor parietal foramina are present. No pathological features were observed. Thickness of the parietal is 5 mm. in the midline and 9 mm. at the lateral point. BV-12 (USNM 381942): Adult male (?); incomplete. Although only the base was missing, reconstruction was not possible because of the extreme distortion of the fragments and the thick coat of calcium covering the surfaces. Nevertheless it was possible to observe that the general form was similar to other skulls in the series. Morphological observations were restricted to the temporal regions: the auditory meatuses do not exhibit exostoses but have numerous small circular perforations in the floor. BV-13 (USNM 381943): Adult male (?); fragments of a skull. Reconstruction was limited to part of left side. The left temporal showed neither exostoses nor perforation of the floor of the auditory meatus. BV-14 (USNM 381944): Child; fragments of a skull. Condition did not permit measurements or observations. BV-15 (USNM 381946): Child; fragments of a skull. No reconstruction was possible. Size of the superorbital region indicated an age of under one year. Summary of the Data The measurements and observations obtained for the individual skulls show a number of consistencies that provide a basis for inference and comparison. These can be summarized under five general headings: Population.-The series is composed of 15 indi- viduals of which 13 are adults, 1 a child of undeter- mined age and 1 an infant of less than one year. Among the adults, 5 are certainly female, 5 are cer- tainly male, and 3 are probably male. The 10 adults of certain sex identification represent 1 young adult (female), 8 middle aged adults (4 female and 4 male) and 1 old adult (male). General morphological characteristics.- Similarities between the individuals comprising the series are remarkable. All are round as viewed from above (vertical norm). From behind (occipital norm) the parietals appear parallel and the roof of the vault rather flat. In profile (lateral norm), the vault is moderately high. From the front (frontal norm), a Lambda < Glabella Upper Nuchal Line (Inion) Bregma Figure 113.-Stereograph drawing of skull BV-11, G-54, Burial 8. WHOLE VOLUME APPENDIX 2 225 metopic crest can be clearly discerned. The super- orbital ridges show marked sexual dimorphism. The skulls in general have smooth lines with attenuated eminences and occipital crests. Metric characteristics.-Measurements and in- dices are summarized on table 1. Table 1.-Measurements, indices, and module of individual crania comprising the Buena Vista series Sex Identifi- cation No. Measurements {mm.) Indices Module {mm.) MAI MT AB MF CI MAUI ACM Male BV-1 182 148 121 99 81. 3 73. 3 150.3 Male BV-3 (169) 139 122 - 82.2 79.2 143.3 Male BV-4 182 148 - (91) 81.3 - Male BV-7 181 - (115) 88 - - Male BV-9 183 (143) 123 - 78. 1 75. 4 149.6 Male BV-10 - - - 87 - - - Female BV-2 (181) 148 (121) 93 81.7 73.5 150.0 Female BV-5 166 (148) 114 - 89. 1 72. 6 142.6 Female BV-6 177 148 118 94 83. 6 72. 6 147.6 Female BV-8 170 138 107 92 81.1 69.4 138.3 Female BV-11 167 148 109 97 88. 6 69. 2 141.3 Key: MAP, maximum antero-posterior diameter; MT, maximum transverse diameter; AB, auricular-bregmatic diameter; MF, mini- mum frontal diameter; CI, cranial index; MAHI, mean auricular height index; ACM, auricular cranial module. Figures in paren- theses are estimated. Particular morphological features.-The fre- quency of morphological features observed on the Buena Vista skulls is summarized on table 2. Pathology.-Signs of spongy hyperostosis (sym- metrical osteoporosis) were observed in 4 of 10 individuals examined (5 male and 5 female). All showing the condition are female. Three exhibited thickening in the obelionic region and the fourth a vestigial criba orbitaria. The glenoid cavity of the temporal shows marked degenerative changes (ar- thritis) in one female; only the glenoid condyles are affected in another. One male has a scarlike depres- sion on the frontal of unidentified origin (cicatrice? result of infection?). Table 2.-Frequency of independently variable traits in the Buena Vista population Trait Individuals examined Cases observed Frequency {percent) Ear exostosis 11 1? 9 Perforations of the floor of the auditory meatus Inca bone 10 10 2 2 20 20 Apical bone Emissary of the M.M. 10 5 2 3 20 60 Discussion Discussion of the skulls that compose this series can be organized under two frames of reference. First, we will compare the individuals forming the group, de- fining their physical type and characteristics as a population, including pathology, and second, we will attempt to indicate the differences or similarities that the group as a whole shows with other groups from the coastal Andean area. Group characteristics.-From inspection alone it is evident that there is great homogeneity in the skull type represented by individuals of the Buena Vista series. Characteristic are roundness when viewed from above, parallelism of the parietals and a rather flat vault when viewed from behind, a metopic crest when viewed from the front, and a general smoothness of contour. Metrically, all are brachy- cranial; cranial indices range between 78.1 and 89.1, with eight of nine individuals having an index above 81.1. It should be noted, furthermore, that the index of 78.1 ascribed to one skull was calculated from estimated reference points, making its mesocephalic condition hypothetical. In regard to height, the mean auricular index classifies all as high except two females, which have a medium high skull. When viewed as a population, this group shows several interesting characteristics. First, there are three generations, including one old person, twelve middle aged adults and two children. Second, the skulls show striking homogeneity in certain features (for example six of the nine measureable skulls have a maximum transverse diameter of 148 mm.) to such an extent that superposition of the stereograms shows them in several cases to be almost identical. Thirdly, there is an unusually high frequency in certain phenotypic traits whose genetic origin is extremely probable, such as the high percentage occurrences of apical and Inca bones (20 percent in each case) and of the emissary artery of the M.M. (60 percent). The latter is exceptionally high in comparison with that for precolumbian populations in Peru, which vary between 20 and 40 percent (Munizaga, 1963). These three kinds of facts lead to the inference that the in- dividuals represent an endogamous and possibly a family group. In respect to pathology, we wish to comment only on the differential distribution of spongy hyperostosis between males and females. It is curious that in a population presumably sharing the same living con- ditions only the females are affected, and furthermore that nearly all of the females (4 out of 5) have been. This finding makes it difficult to accept the explana- tion that the condition originates from malaria. 226 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Comparison with other coastal andean series.- Buena Vista and San Pablo: Since we have shown that the individuals from the Buena Vista site represent a homogeneous physical type, it is of interest to determine whether other skulls associated with Valdivia Phase cultural remains share the same physical characteristics. Some light can be shed on this question by comparing our series with that ob- tained by Carlos Zevallos Menendez and Olaf Holm from G-l 15: San Pablo, located to the south of G-54: Buena Vista and representing the same general cul- tural period (see pp. 95, 104). The San Pablo series comprises nine individuals, three male and six female. They were recovered in poor condition, but after reconstruction seven could be measured (table 3). Table 3.- -Measurements, indices and module of individual crania comprising the San Pablo series Sex Identifi- cation No. Measurements {mm.) Indices Module {mm.) MAP MT AB MF CI MAHI ACM Male SP-1 190 134 118 80 70. 5 72. 8 147.3 Male SP-2 178 133 - 96 74.7 - Male SP-3 188 132 - 94 70.2 - Female SP-4 182 131 110 90 71.9 70.2 141.0 Female SP-5 - (138) - - - - Female SP-6 168 132 - - 78.5 - Female SP-7 163 (138) - - 84.6 - Key: MAP, maximum antero-posterior diameter; MT, maximum transverse diameter; AB, auricular-bregmatic diameter; MF, minimum frontal diameter; CI, cranial index; MAHI, mean auric- ular height index; ACM, auricular cranial module. Figures in parentheses are estimated. Comparison of the measurements, indices and modules with those of the Buena Vista series reveals both biological and cultural differences and simi- larities. From the biological point of view, die most striking aspect of the San Pablo series is its hetero- geneity in contrast to that from Buena Vista, suggest- ing either that different groups are represented or that population mixture has taken place. The fact that cranial indices tend to group around disparate norms favors the inference that two different physical types are represented rather than the effect of inter- mixture. Culturally, a new trait is exemplified in an individual showing intentional cranial deformation, a treatment not encountered in the Buena Vista population. Since the same type of cranial deforma- tion has been observed in skulls associated with the Machalilla Phase, this individual can be ascribed with reasonable certainty to that Phase (see pp. 172, 229.) These differences and similarities between the two series suggest that the question regarding the physical type of the Valdivians may be answered in the following way: Brachycephaly was the pre- dominant skull form, with a small minority of dolico- cephalics. The distribution of indices suggests co- existence of two groups of different biological origin sharing the same culture. Can one of these physical types be designated as the basic Valdivia Phase population? In view of the larger number of brachy- cephalic individuals represented, and because of the evidence that the Buena Vista series comprises a social and possibly family group, we are inclined to believe that the original population resembled most closely that illustrated by the individuals included in the Buena Vista series. Buena Vista and Cabezas Largas: To obtain an answer to the question of whether the introductoin of pottery to this part of the South American continent was paralleled by the introduction of a new physical type, it is necessary to compare the Buena Vista series with preceramic populations of equivalent or greater antiquity. One group suiting this criterion has been described by Hartweg (1961) from Cabezas Largas on the south coast of Peru. Carbon-14 dates give this site an antiquity of about 5000 years (Engel, 1960), making it some 1000 years older than the Ecuadorian series. From the morphological point of view, none of the three types distinguished by Hartweg in his population of 28 individuals equates with the Buena Vista series. The delicate (fine) type, which is most frequent and encompasses 71.5 percent of the popula- tion, differs in possessing sphenoid bosses when viewed in the vertical norm, and a well marked sagittal crest. The Eskimoid type deviates in having an elipsoidal horizontal contour in addition to a well marked sagittal crest. The gross intermediate type is incomparable because of "its general grossness, the thickness of the bones and the prominence of the relief" (Hartweg, 1961, p. 124). In making metrical comparisons, we are limited to the horizontal cranial index. The mean indices for the different physical types making up the Cabazes Largas series (a mean for the series as a whole is not given) diverge markedly from that for the Buena Vista series, which is 82.9. Specifically, the delicate type (mean 75.13) differs by 7.8 points, the Eskimoid type (mean 71.33) by 11.6 points, and the gross type (mean 77.89) by 5.1 points. Examination of differences in distribution of the cranial indices in individuals representing these types makes the lack of comparability still more apparent, since the Cabezas Largas skulls are found to be dolico- to meso-cranial, while the Buena Vista skulls are decidedly brachy- cranial (table 2). WHOLE VOLUME APPENDIX 2 227 Visual examination by Hartweg of the Cabezas Largas skulls led him to the conclusion that they are high, although in the gross intermediate type this is not so evident. In one case, the following indices could be obtained: height/length 74.44 (orthocranial) and height/width 92.41 (acrocranial). The mean auricular height index showed the Buena Vista skulls also to be high, with the exception of two females. Table 4.-Comparative distribution of cranial index in populations of Cabezas Largas and Buena Vista Cranial index Cabezas Largas Buena Vista Delicate Eskimoid Gross Total Total 65-69. 9 1 1 70-74. 9 6(?) 4 - 10 - 75-79. 9 13(?) - 2 15 1 80-84. 9 - - 2 2 6 85-89. 9 - - - - 2 Totals 20 4 4 28 9 In summary, we are confronted with two basically different populations. The two brachycranial indi- viduals in the Cabezas Largas series, incorporated in the gross intermediate type, are of most interest be- cause they are the only ones that show any possible relationship to the Buena Vista population. Un- fortunately, since Hartweg (1961) does not specify the conditions in which they were found at the site, and Engel (1960) suggests three different hypotheses to explain the origin of the ossuary from which these skulls may have come, their chronological position is not completely clear. The third hypothesis, which Engel feels to be the most plausible, attributes them to intrusion during subsequent reoccupation of the site. A cultural unity between these and the other burials is implied by similar specific burial goods, especially hides and totora (Engel 1960, pp. 24-25). However, this type of treatment of the dead survived into later times, so that similarity in cultural context does not necessarily imply contemporaneity. In view of these factors, the possibility cannot be discarded that the brachycranial individuals represent later in- trusion. If so, the populations of the two sites are totally distinct. Buena Vista and later populations: Our second problem-the affinities of the Buena Vista physical type with that of later populations-can be ap- proached by comparison of cranial indices and ob- servations on size and height of the skull. In 1914, Hrdlicka noted that brachycrania was one of the characteristic features of coastal Peruvian groups. Newman (1943, 1947) has confirmed this impression, and Stewart (1943b) has also encountered brachycranial individuals, although in smaller fre- quency, among north coastal populations. At least in respect to cranial index, these populations resemble that of Buena Vista. For purposes of comparing the cranial indices of several coastal Peruvian series with ours, the indices have been grouped into two cate- gories, less than 79.9 and more than 80 (table 5). In respect to height, we have already noted that the Ecuadorian skulls are high, a feature that agrees with reports on precolumbian populations of the Peruvian coast (Stewart, 1942, 1943a). Although the series is too small to permit significant conclusions regarding size, the large size of some of the skulls is striking since what we knew of the later populations of the region had not led us to expect it. Table 5.-Distribution of individuals by two categories of cranial index in five coastal Andean series Peruvian Coast Ecuador Cranial index Central Coast Cupis- nique Mochica Chica- ma Buena Vista Total x-79.9 80-x 13 66 2 4 12 7 33 32 1 8 61 117 Totals 79 6 19 65 9 178 The Central Coast series (Newman, 1947) and that from the Chicama Valley (Stewart, 1943b) are com- posed entirely of males and represent several cultural periods. These factors, plus the large size of the samples, permit the derivation of valid conclusions from comparison. The Cupisnique and Mochica series, and our own from Buena Vista, on the other hand, represent a small number of individuals, both male and female, which makes comparative analysis unreliable. Nevertheless, these small series, well defined culturally and chronologically, permit us to approach the study of the physical types of the region from another point of view. They are of utility at the level of analysis that emphasizes their characteristic of constituting a social group. They may be viewed as endogamous populations, whose isolation in the coastal valleys contributed to the formation and maintenance of the brachycranial head form on the coast. They also may be viewed as representing a continuing migra- tion of small groups from a point of origin where this physical type was characteristic. One or both of these hypotheses may reflect the role played by the Buena 228 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Vista population in the racial history of the area, and account for the high proportion of brachycrany represented in table 5. Conclusions The Buena Vista series, associated with a pottery- making culture, represents a very homogeneous brachycranic type that probably corresponds to a family group. Total absence of any kind of intentional cranial deformation indicates that this practice was not of Valdivia Phase origin. The Buena Vista physical type is totally different from the earlier preceramic Peruvian coastal popu- lation of Cabezas Largas, leading to the inference that the introduction of pottery also brought a new physical type. This new group may be to a large degree the source of the brachycrany observed in populations associated with the later potterymaking periods on the Peruvian coast. THE MACHALILLA PHASE The skeletal remains of two individuals from the site of G-110: La Cabuya are of particular importance for physical anthropological studies in South America. On the one hand, they constitute the only skeletal material clearly associated with the Machalilla Phase, and on the other, they furnish the earliest evidence on the Pacific coast of South America for the presence of intentional cranial deformation. A sample containing only two individuals would be insufficient for making comparative studies of physical type, even if it were not for the alterations resulting from artificial deformation. Consequently, this report will be limited to description of the remains and of the type of cranial deformation. Subsequently, the distribution of this kind of deformation in time and space will be reviewed in order to evaluate the role played by the population associated with the Machalilla Phase in diffusion of the trait along the Pacific coast. Material The skeletal remains are very fragmentary. Time permitted detailed examination only of the crania, which have the following characteristics: 1. M-l (USNM collection: Young adult female (?), deformed (pi. 196). Base, sphenoids, lower portion of occipital and part of the face missing (fig. 114). The existence of broad suture areas and the fact that many of the breaks were clean permitted reconstruction of the vault. The most difficult adjustment was along the coronal suture, where Bregma Lambda Upper Nuchal Line (Inion) Glabella Figure 114.-Stereograph drawing of skull M-l, G-110, Cut 1. WHOLE VOLUME APPENDIX 2 229 contact between the frontal and parietals was reduced to a small part of the bregma tic zone. However, a small area between the middle of the right parietal and the frontal was complete, permitting correct positioning of the frontal. This factor is particularly important since observation of the inclination of the frontal bone is fundamental for identification of the kind and amount of deformation. Sutures remain open on both the interior and exterior of the vault. The teeth, with all four third molars erupted, show medium wear. In regard to physical type, it can be noted that, had deformation not been present, the individual would have been brachycranic. Morphological features include absence of osteoma of the auditory canal, and presence in both tympanic plates of large perforations. An interparietal bone is absent, but the lambdoid region shows probable vestiges of an apical bone with the transverse suture closed. Signs of pathology are absent. Both maxilla and mandible exhibit a dental anomaly in the form of crowding of the canines. This is sufficiently marked in the maxilla to result in an anteroposterior orientation of the canines (pi. 196f). 2. M-2 (USNM collection): Young (?) adult female, deformed. Limited to the nearly com- plete occipital and superior posterior angle of the left parietal. Sutures in the lambdoid region are open both on interior and exterior. Teeth show medium wear. Reconstruction was limited to the sagittal curve of the posterior third or quarter of the parietals to a point on the under side of the occipital torus. This curve permits recognition of the cranium as deformed. Apical and Inca bones are absent and no pathology was observed. Artificial Deformation of the Skull Identification of the type of deformation.- The sagittal outline and photographs of Skull M-l show clearly the presence of artificial deformation, resulting in a marked flattening of the occipital bone, especially the upper part (pi. 196 a, c-d). Pressure has altered the sagittal outline slightly on the frontal, strongly on the parietals (which have an index of curvature of 25*), and to an extreme degree on the occipital. The squamous portion of the latter bone, which under normal circumstances is more or less convex and inclined from front to back, has been *The mean values for the index of curvature of the parietal bone in normal skulls correspond, according to Aigner, to 22.44 for dolico- crania and 22.00 for brachycrania. Using this index, Dillenius (1910, p. 46) arrived at a mean value of 27.09 for skulls with fronto-occipital deformation, clearly separating them from the normal series. The value attributed to Skull M-l helps to give an idea of the intensity of deformation of the Machalilla Phase skulls. altered to a concave contour and a back to front inclination. Consideration of all evidence leads to the diagnosis of the type of deformation as fronto- vertico-occipital, or tabular erecta in the terminology of Imbelloni. The fragments of the parietal and occipital bones representing Skull M-2 show the same type of deformation. A third skull, which comes from the Valdivia Phase site of G- 115: San Pablo, can be identified as repre- senting an individual of the Machalilla Phase on the basis of its artificial deformation, which is the same type as that characterizing the examples from G-110 but more pronounced. The particular details of the two most complete skulls permit classification of the deformation as the "cuneiform" or "Chavm" variety of the fronto-vertico- occipital type (Weiss, 1961). The more intense flattening of the occipital bone and absence of asym- metric form place them in this subtype. Identification of the deforming apparatus.- The different shapes that a normal skull can acquire by deformation have been classified into types and varieties according to the manner in which pressures were applied. This permits us to infer from the characteristics of a skull the kind of apparatus by which it was molded. The tabular erecta deforma- tion shown by the Machalilla Phase skulls is the result of strong compression of the back of the head against a hard surface, to which it is fastened by appropriate binding. The "cuneiform" variety of this type indicates that the head was held so that the face invariably looked forward; the pronounced occipital flattening shows that the bindings were tighdy applied; the concavity of the occipital permits the inference that a pillow-like object was placed between the skull and the board. Weiss (1961) is the most recent to discuss this type of cranial deformation. After a review of the literature and reexamination of the skulls used to define the "cuneiform" variety, he encountered difficulties in reconstructing the apparatus by which it was pro- duced, but inclined to the view that it was not a cradleboard but rather a device fastened to the head alone. He illustrates two probable examples of such apparatus from Peruvian archeological sites, one of which is a pillow attached to the occipital by a complicated system of binding (Weiss, 1961, pi. 4) and the other a small cradle-like object placed on the head only, which because of its special structure produces the "cuneiform" type of deformation (op. cit., pi. 39). Because of the scarcity of both archeological and ethnological evidence relating to apparatus that might have been employed to produce tabular erecta deformation, it is worth calling attention to a figurine 230 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 from Lambayeque showing a board attached to the occipital by two bands, one sagittal and the other transverse (Dembo and Imbelloni, 1938, pp. 288-289), which could have produced this result. Distribution of Skull Deformation on the Pacific Coast of South America Analysis of the pattern of occurrence of the various types of cranial deformation in South America is a complicated task, because of the ambiguous manner in which different types are distributed through time, space and culture, and because of the small number of samples whose cutural and chronological context has been established with reasonable confidence. Nevertheless, it is possible to recognize three large geographical areas or centers of dispersion: 1) the Caribbean coast; 2) the coast of Argentina, and 3) the Pacific coast (Stewart, 1950). In certain localities in Peru, the pattern of distribution of the practice by cultural phases has made it possible to give a temporal dimension, but no clear picture is yet available of when and where deformation was introduced into South America or how it was diffused over the continent. Within the Pacific coastal area, several skeletal series have been described in recent years, which in conjunction with the information now furnished by the Machalilla Phase skulls makes it useful to reexamine the evidence for the origin and dispersal of the trait. The Machalilla Phase remains equate with Period C, with an initial carbon-14 date of 3320 ñ170 years ago (table G), and comprising the final portion of the Phase. The inception of the Phase, and by inference the introduction of the population repre- sented by the Period C skeletal remains, has been given a minimum antiquity of 4000 years ago by the archeologists (pp. 148-149). At this time, complexes on the coast of Peru had not yet adopted ceramics. In order to trace the antiquity of the practice of deformation as well as its association with the spread of potterymaking, it is necessary to review evidence for the Peruvian Preceramic and Formative (or Cultist) Periods. Complexes failing within these periods for which skeletal information is available are shown on figure 115. Skull deformation in the preceramic period.- Only three series are available for the Preceramic Period on the coast of Peru and none exist from the coast of Ecuador. The earliest is from Cabezas Largas on the south coast and represents the early part of the period lacking cotton, dated at about 5000 years ago (Engel, 1960). Artificial cranial deformation does not occur (Hartweg, 1961). More recent, but still associated with the Pre- ceramic, are two populations from the north and central coast: Culebras and Asia. Both represent the late part of the period, dating around 3200 years ago (Engel, 1963, p. 20), in which cotton was present. All the individuals comprising these two populations show artificial skull deformation of the fronto-vertico-occipital or tabular erecta type, and the apparatus used to produce the deformation is inferred to be the cradle (Hartweg, 1958, p. 182). The position in which the head of the child was fixed to the board was not always the same, resulting in a high frequency of asymmetrical skulls in both populations. In other words, the position of the head against the board was not culturally defined, in contrast to what appears to have been the situation in the Formative Period. Skull deformation in the formative period: Ecuador.-The Machalilla Phase remains are the earliest so far discovered on the Pacific coast of South America showing artificial deformation. The 24 skulls available from earlier and partly contemporary Valdivia Phase sites on the coast of Ecuador show no evidence of deformation with the single exception of a female from G-115: San Pablo. Since this in- dividual exhibits the type of deformation character- izing the Machalilla Phase skulls from G-110, it can confidently be considered a representative of the Machalilla Phase population. As such, it provides physical evidence to supplement cultural evidence (see pp. 147-148, 171-172) of contact between the two populations. These data permit us to infer that artificial cranial deformation was introduced onto the coast of Ecuador by individuals representing the cultural complex known as the Machalilla Phase some 4000 years ago, and that the type was fronto-vertico-occipital or tabular erecta. This type continues to be practiced in various later periods of Ecuadorian prehistory. Skull deformation in the formative period: peru.-On the north coast, Stewart (1943b) encoun- tered a sequence of cranial deformation beginning with the tabular erecta type in the Cupisnique (Coastal Chavin) Period. Newman (1947) found a similar situation on the central coast. A brief com- parison made by Stewart (1943b, p. 172 footnote) between his results and those of Newman indicates clearly the pattern in which this practice is distributed on the north and central coasts: As for fronto-occipital deformity, Doctor Newman found it to be present in practically all his earliest or "Shell Mound" crania. In the Early (Interlocking Style) Period the incidence falls to about one-third, but in the Middle Period it increases to about two-thirds. Finally, in the Late Period intentional deformity almost disappears, and the great majority of the whole volume APPENDIX 2 231 crania are either undeformed or show slight to moderate posterior flattening. This situation seems to parallel that in the Chicama-Moche-Viru region, where intentional deformity is present in the Cupisnique period, disappears in the subsequent Mochica period, and reappears later only to taper off ultimately. On the south coast, the situation is similar, with the tabular erecta type in its "cuneiform" or "Chavin Cavernas" variety occurring in Paracas Cavernas (Weiss, 1961, p. 72). Comparison of cranial deformation characteristic of populations of the early Peruvian Formative brings to light not only a great similarity but an identity in the type employed. The closeness of the resemblance is evident from examination of illustrations of the Machalilla Phase skull (pi. 196) in conjunction with those from Cupisnique (Stewart, 1943b, pi. 11), early Ancon-Supe (Newman, 1947, pi. la), and Paracas Cavernas (Weiss, 1961, pi. 22B). Added to this identity in type of deformation is identity in the deforming apparatus used to produce it. Consistent repetition of the same kind of occipital flattening cannot be explained except as the product of applica- ECUADOR PERU GUAYAS COAST NORTH COAST CENTRAL COAST SOUTH COAST B.P. 2450 - PARACAS CAVERNAS A A CUPISNIQUE A EARLY ANCON r 3000 - |~~ CULEBRAS A ASIA i _j A 3400 - MACHALILLA k I VALDIVIA i > 1 f 4000 - 4500 - 5000 - 1 ' CABEZAS m LARGAS   NO DEFORMATION A TABULAR ERECTA DEFORMATION A TAB. ERECTA DEFORMATION, CHAVIN VARIETY APPEARANCE OF POTTERY Figure 115.-Sites reflecting the first occurrence of skull deformation on the coasts of Ecuador and Peru. 232 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 tion of a deforming mechanism according to a rigidly defined cultural pattern. We may assume, as a result, that these populations shared not only an element of material culture (the deforming apparatus) but also a method of attaching to it the head of the child. Problems Having completed this brief review of the occurrence of cranial deformation on the Andean coast of South America, we may return to the problems raised in the introduction: Since the population of the Machalilla Phase is ac- cording to existing evidence the earliest to show cranial deformation on the coast of the Andean Area, and since the type of deformation closely resembles that characteristic of later populations, can it be con- sidered as the focal point for the diffusion of the prac- tice in this area? The fact that all the populations examined showed tabular erecta or fronto-vertico-occipital deformation suggests a single origin for the practice. This unity is even more strongly implied by the universality among Formative Period populations of the "cuneiform" variety, which reflects use of the same kind of deform- ing apparatus, making it seem probable that this variety of tabular erecta deformation was diffused from Ecuador to Peru during the Formative Period. On the other hand, an inference of similar diffusion toward the end of the Preceramic presents certain difficulties because of the absence of the "cuneiform" variety and the high frequency of asymmetrical skulls in Peru. This situation can be accounted for by one of the two following possibilities: Diffusion of the idea of deformation without simultaneous diffusion of the deforming apparatus: Examination of the type and variety of deformation represented by the Machalilla Phase skulls suggests that it was produced by a deforming apparatus at- tached to the head. Using similar evidence, Hartweg (1961) has inferred that a deforming apparatus in- volving the whole body (a cradle) was utilized in the late Preceramic in Peru. Even if both these inferences are correct, diffusion of the idea of intentional defor- mation of the skull can still be postulated with modifi- cation in the method employed by the receiving population. Simultaneous diffusion of the deforming apparatus and the idea of cranial deformation: Acceptance of Hartweg's inference that deformation among preceramic populations of Peru was accom- plished by use of a cradle would not permit acceptance of this alternative hypothesis. However, although archeological and ethnographic evidence suggests that tabular erecta deformation was produced by use of the cradle, from the strictly technical point of view there is no reason to reject the possibility that such deformation may have been produced by use of a board or pillow applied to the back of the head and held in place by bindings. If this alternative is allowed, the earlier asymmetrical form of deformation can also be viewed as a diffusion from Ecuador to Peru in the Preceramic. The difference in symmetry can be explained by the hypothesis that during the Formative Period a board or pillow held by a system of bindings was employed to hold the head of the child in a forward position, while during the Preceramic this apparatus was employed without attention being paid to keeping the head in a position that would result in symmetrical deformation. The latter hypothesis, which does not depend on the use of the cradle as the deforming apparatus, is in keeping with what is known of the archeology of these periods. Engel (1957, 1960, 1963) does not mention the cradle as one of the numerous cultural elements recognized from the Preceramic periods in Peru. Furthermore, Weiss (1961) in a review of the deforming apparatus employed in precolumbian Peru does not mention the cradle during the Formative, while listing several other kinds of cranial deformation apparatus. Although he distinguishes three periods during which cradleboard deformation was practiced (Preceramic of Asia and Culebras; Late Chimu, Chancay and other contemporary cultures; and Inca), archeological evidence for the use of the cradle is available from the latter two only (Weiss, 1961, pp. 19-25). Do differences between the Preceramic and Forma- tive varieties on the coast of Peru indicate a diverse origin for the practice? The answer appears to be negative. If the differences reflect separate origins, a possible center of diffusion can be sought on the Atlantic coast. Skulls from Intihuasi cave, Province of San Luis, Argentina, dating from 6000-8000 years ago exhibit deformation of the circular or pseudo-circular type, but not tabular erecta (Gonzalez, 1960, p. 160). These finds indicate the presence of a tradition of cranial deformation of greater antiquity than that on the Pacific coast, but the possibility of diffusion from this area is minimized by the fact that circular de- formation is produced by binding whereas the tabular erecta type results from the use of a board. Conclusions Cranial deformation appears for the first time on the Pacific coast of South America in Ecuador at about 4000 years ago, as part of the Machalilla Phase. WHOLE VOLUME APPENDIX 2 233 The type of deformation is fronto-vertico-occipital or tabular erecta, of the variety known as "cuneiform" or "Chavin." Tabular erecta deformation appears along most of coastal Peru in the late Preceramic, at a date subse- quent to its occurrence in Ecuador, permitting the inference that it was diffused from north to south. The north-south diffusion current is most clearly demonstrated by the Formative Period complexes (Machalilla, Cupisnique, Early Ancon-Supe, Paracas Cavernas), in which identity in the variety of deforma- tion implies diffusion of the same deforming appara- tus, a rather complicated system that seems to be correlated with this time period and zone. Diffusion of cranial deformation from the Formative Machalilla Phase of Ecuador to the late Preceramic complexes of coastal Peru (Asia and Culebras) is highly probable but not as definite, since although the tabular erecta type is shared, the "cuneiform" variety is absent. To explain this, it has been suggested that the same type of deforming apparatus was used but differently applied during die two periods. In the Formative, a rigidly prescribed pattern seems to have existed in attachment of the head of the child to the board, producing a symmetrical and homogeneous result. In contrast, during the Preceramic, there ap- pears to have been greater latitude in the use of the apparatus resulting in examples of asymmetrical deformation. The possibility of influence from the ancient center of deformation in Argentina is rejected on the basis that the type and by implication the deforming apparatus was totally distinct from that used on the Pacific coast during the Preceramic and Formative Periods. LITERATURE CITED Dembo, Adolfo, and Jose Imbelloni 1938. Deformaciones intencionales del cuerpo humano de caracter etnico. Humanior, Bibliografia americana moderna, sec. A, vol. 3, Buenos Aires. Dillenius, J. A. 1910. El hueso parietal bajo la influencia de la deformacion fronto-occipital. Publ. de la Seccion de Antropologia no. 7, Facultad de Filosofia y Letras, Buenos Aires. Engel, Frederic 1957. Sites et etablissements sans ceramique de la cote peruvienne. Journ. de la Soc. des Americanistes, vol. 46, pp. 67-155, Paris. 1960. Un groupe humain datant de 5000 ans a Paracas, Perou. Journ. de la Soc. des Am6r- icanistes, vol. 49, pp. 7-35, Paris. 1963. Notes relatives a. des explorations archeologiques a Paracas et sur la cote sud du Perou. Travaux de l'Institut Frangais d'Etudes Andines, vol. 9, pp. 1-72, Paris. Gonzalez, A. Rex 1960. La estratigrafia de la Gruta de Intihuasi (Prov. de San Luis, R.A.) y sus relaciones con otros sitios preceramicos de Sudamerica. Revista del Instituto de Antropologia, Cordoba, Argentina. Hartweg, Raoul 1958. Les squelettes des sites sans ceramique de la cote du Perou. I. Etude descriptive. Journ. de la Soc. Americanistes, vol. 47, pp. 179-198, Paris. 1961. Les squelettes des sites sans ceramique de la cote du Perou. II. Etude descriptive de documents nouveaux. Journ. de la Soc. des Americanistes, vol. 50, pp. 111-133, Paris. Hrdlicka, Ales 1914. Anthropological work in Peru in 1913, with notes on the pathology of the ancient Peruvians. Smithsonian Misc. Coll., vol. 61, no. 18, Washington. 1939. Practical anthropometry. Philadelphia. Munizaga, Juan R. 1963. Estudio de un rasgo de variacion discontinua en las poblaciones Americanas. Notas del Centro de Estudios Antropologicos no. 7, Universidad de Chile. 234 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 1 Newman, Marshall T. 1943. A metric study of undeformed Indian crania from Peru. American Journ. of Physical Anthropology, new series, vol. 1, no. 1. 1947. Indian skeletal material from the central coast of Peru: An archeologically oriented study in physical anthropology. Papers of the Peabody Mus. of American Archaeology and Ethnology, Harvard Univ., vol. 27, no. 4. Stewart, T. Dale 1942. Anthropometric nomenclature. II. The indices of head height. Amer. Journ. Phys. Anthrop., vol. 29, no. 1. 1943a. Distribution of cranial height in South America. American Journ. Phys. Anthrop., new series, vol. 1, no. 2. 1943b. Skeletal remains with cultural associations from Chicama, Moche, and Viru Valleys, Peru. Proc. U.S. Nat. Mus., vol. 93, pp. 153-185. 1950. Deformity, trephining, and mutilation in South American Indian skeletal remains. In Handbook of South American Indians. Bur. of Amer. Ethnol. Bull. 143, vol. 6. pp. 43-48. Weiss, Pedro 1961. Osteologia cultural, Segunda Parte: Practicas cefalicas, Lima, Peru. PLATE 1 Typical views of the Guayas coast, a, Playas Bay looking southward, h. Low sandy beach typical of coast from Santa Elena Peninsula north to G-88: Palmar Norte. 7G7-S41-Co PLATE 2 Typical views of the Guayas coast, a, Puerto Chanduy, a modern fishing village on the coast east of Santa Elena Peninsula, b, High cliff on which G-l 10 is located (scale is indicated by the pickup truck on the beach at the center right). Typical topography and xerophytic vegetation of the coast of Guayas Province, a, Inland terrain in the vicinity of the modern village of Palmar, b, Margin of Lagarto Salitre. :m& X ? -'?, "'r;v.':^ v' ">--' ?.; ^^?p^k- .^>-y?-- ; >^ -v- ? , m**rz^ *S¯ '-&" mfi"'. - f ¯.fcr We."'- <7* V-''' 'v**-. V- ' '" ¯. V* -, "œ: -.>.- i^v- - .J-* "* :,.i'.i* f (I ? . -"-'-^ll.-.-u-" '..<** .S*^i ? ? ?  ?  ".- ->*.-.'¯, ; , ... - -B ** *i ® .¯ y--M ;R \ v-  -V- 4-^*:^ ^*^*-U¯. ^b,"^^ PLATE 10 Stratigraphy of G-31, Cut J, southeast face, a, Sloping levels in Section B. b, Sections D and F. PLATE 11 G-54: Buena Vista, a Period C site of the Valdivia Phase, showing topography and modern vegetation, a, View from the first terrace toward the bank at the north edge of the site; the dark upper layer is sterile overburden (see fig. 6). b, Looking west with Cut 2 in the foreground. PLATE 12 G-54, Burials 1-7 during excavation, a, General view of skeletons, b, Closeup of polished ax (pi. 19r) adjacent to arm. The environment of the Posorja region, a, Edge of the hill occupied by G-84 and the modern town of Posorja, looking toward the Gulf of Guayas (shell refuse represents a late Manteno occupation), b, View from the same hill toward Pesqueria Salitre. 5w :*& 'v ? '"^'^ 0 .- " - > x ^?v -.^. ' ? ?*> _ ^iMXi W ? ??- b PLATE 14 Environment of Palmar Salitre. a, View from the south edge looking across the salitre toward the north, b, Northwest margin showing the hill on which G-88 is located. 3 CM PLATE 15 Flake blades or knives from Valdivia Phase sites, with the cutting edge downward. .Vh>-. . ' PLATE 16 Miscellaneous stone artifacts of the Valdivia Phase, a-b, Bowls, c, Pebble chopper, d-f, Grinding stones. PLATE 17 Pebble choppers from the Valdivia Phase. 767-841-65 18 PLATE 18 Valdivia Phase stone tools, a-j, Gravers, k-w, Scrapers. 0 I 2 3 4 CM L. 0 J I I 2 CM PLATE 19 Miscellaneous objects from Valdivia Phase sites, a-j, Natural pebble polishers, k-n, Clay lumps, o-r, Ground and polished axes. %v i I 1 1 ' ' 0 I 2 3 4 CM PLATE 20 Fishhook reamers from Valdivia Phase sites showing uniformity in size and form. PLATE 21 Miscellaneous shell objects from the Valdivia Phase, a-l, Cut segments of pearl oyster shell, probably for preparation of fishhook blanks. m-p, Abraders and polishers of Group A. q-t, Scoops, spoons or spatulas. --?- J ?????*? *.-,.. ® ., * ??-'" ? -. '?- '' ??? ' ?.<7l j v-V * i - " - A'.'s^ '; r:,. />;: 'JtS&yt-* - K< ®?'?'. "''-?'.i--v ?; ? Mr \ ' ' / \.?'? >? i'|W r,+- v v j* <".u~*.y>' > ?;..-?? ¯0vs I L 0 I 2 3 CM a I I I I 0 I 2 3 cm b-e 0 I 2 CM f "-.- PLATE 22 Shell artifacts of the Valdivia Phase, a-b, Abraders. c-e, Unidentified objects, f, Bowl. p q r PLATE 23 Shell ornaments of the Valdivia Phase, a-b, d-h, Clam shell pendants. c,j-k, Shaped shell pendants. I, Pecten shell pendant, m-r, Beads. PLATE 24 Shell fishhooks of the Valdivia Phase, a-d, g, Roughed-out blanks, e-f, Drilled blanks, h-q, Complete and fragmentary fishhooks showing range of size. m} ?;!?'?, W¯?1 B] : V '? Ik I ^^IB- 1 1 ^H ^H B 1 ^?'?r;i . 1 ; (J^| I HI f t ? -li' ' III PLATE 25 Bone and tooth artifacts of the Valdivia Phase, a-d, Awls of ander. e, Antler tip projectile point, f-j, Awls of saw fish teeth. 4 CM PLATE 26 Type sherds of Punta Arenas Incised. 2 3 CM PLATE 27 Type sherds of Valdivia Applique Fillet, vertical parallel bands on body wall. PLATE 28 Type sherds of Valdivia Applique Fillet, curvilinear and intersecting patterns on body wall. PLATE 29 Type sherds of Valdivia Applique Fillet, variations in rim treatment. 0 I 2 3 4 CM PLATE 30 Type sherds of Valdivia Broad-line Incised, Motif 1. PLATE 31 Type sherds of Valdivia Broad-line Incised, Motif 1. I ' I I 1 0 1 2 3 4 CM PLATE 32 Type sherds cf Valdivia Broad-line Incised, Motif 2. J I L 0 I 2 3 4 CM 767-841-65 19 PLATE 33 Type sherds of Valdivia Broad-line Incised, Motif 2. 0 I 2 3 4 CM PLATE 34 Type sherds of Valdivia Broad-line Incised, Motif 3. <**.. 3 CM PLATE 35 Type sherds of Valdivia Broad-line Incised, Motif 3. #;t\\W&& :-i V* c A04mm mm- V ?' " ' #v * ^l * ft® ?'/ !IiJSPr '? f Ilk 'Win J I 1 0 1 2 3 4 CM PLATE 45 Type sherds of Valdivia Brushed, Technique 1, patterned variant. PLATE 46 Type sherds of Valdivia Brushed, Technique 2. PLATE 47 Type sherds of Valdivia Carved. PLATE 48 Type sherds of Valdivia Combed, Technique 1. \ 2 3 4 CM i b Mr \v wy 1 767-841-C5 20 PLATE 49 Type sherds of Valdivia Combed, Technique 1. PLATE 50 Type sherds of Valdivia Combed, a-n, Technique 2. o-s, Technique 3. 1&&^Z^J& 4&&*B*****& 0 I 2 3 4 CM PLATE 51 Type sherds of Valdivia Cord Impressed, a-l, Braid impressed, m-n, Cord impressed. 0 I 2 3 CM \ PLATE 52 Type sherds of Valdivia Corrugated. m PLATE 53 Type sherds of Valdivia Corrugated, a-e, g, i, Plain, f, h,j-n, Finger pressed. PLATE 54 Valdivia Phase decorated types, a, Valdivia Corrugated jar. b-i, Type sherds of Valdivia Pseudo-Corrugated. saw a s )%Sm * > V J I 1 J 0 I 2 3 4 CM ->. L_J I I-I 0 2 4 CM PLATE 55 Type sherds of Valdivia Cut and Beveled Rim. a-j, Motif 1. k-l, Tetrapod vessels of Motif 2. PLATE 56 Type sherds of Valdivia Cut and Beveled Rim. a-j, Motif 2. k-n, Motif 3. J I-I 0 I 2 3 4 CM PLATE 57 Type sherds of Valdivia Embossed. 1 I I I I 0 1 2 3 4 CM PLATE 58 Type sherds of Valdivia Excised, a-m, Motif 1, anthropomorphic faces, n-q, Unclassified motifs. 0 I 2 3 4 CM PLATE 59 Type sherds of Valdivia Excised, a-c, Motif 2. d-g, Motif 3. h-l, Motif 4. PLATE 60 Type sherds of Valdivia Excised, a-e, Motif 5, gouged-out scallops, f-j, Motif 6, undulating bands, k-l, Motif 7, minor addition to broad-line incised designs. I I I I 1 0 I 2 3 4 CM PLATE 61 Type sherds of Valdivia Fine-line Incised, Motif 1. I I I I I 0 I 2 3 4 CM PLATE 62 Type sherds of Valdivia Fine-line Incised, Motif 1. 4 CM PLATE 63 Type sherds of Valdivia Fine-line Incised, a-k, Motif 2. l-q, Motif 3. 4 CM PLATE 64 Type sherds of Valdivia Fine-line Incised, a-d, Motif 4. e-g, Motif 5. h-j, Motif 6. k-q, Unclassified motifs. I I I I I 0 I 2 3 4 CM PLATE 65 Type sherds of Valdivia Finger Grooved, a, b, e-i, n-o, Plain grooving, c-d, m, Additional decoration by fingertip or fingernail punctate '-/, Additional decoration in Valdivia Modeled technique. 767-841-65 21 PLATE 66 Type sherds of Valdivia Fingernail Decorated, a-m, Technique 1, fingernail, n-u, Technique 2, fingertip. - - * 0 I 2 3 CM PLATE 67 Type sherds of Valdivia Incised, Motif 1, zigzag, a, Rim of Form 1. b-d, Rims of Form 2. e, Rim of Form 6. f-h, Rims of Form 3. i-k, Rims of Form 4. PLATE 68 Type sherds of Valdivia Incised, Motif 1, zigzag. PLATE 69 Type sherds of Valdivia Incised, Motif 1, zigzag, a, c, Rims of Form 2 with incision continuing onto exterior thickening. Form 4 with nicked or finger-pressed lip. p, Spout. b, d-o, Rims of 4 CM PLATE 70 Type sherds of Valdivia Incised, Motif 1, zigzag. PLATE 71 Type sherds of Valdivia Incised, Motif 1, zigzag with row of punctation at base of neck. PLATE 72 Type sherds of Valdivia Incised, a-c, Motif 1, zigzag, vertical variants, d-g, Rare vessel Form 2 with small vertical handle, h, Rare vessel Form 1, superimposed jars. I L. ...J ' I 0 I 2 3 4 CM PLATE 73 Type sherds of Valdivia Incised, Motif 2, Crosshatch. ^-j^ZTZ. . ? \.~,i ^-¯"WT ' . ': ' PLATE 74 Type sherds of Valdivia Incised, Motif 3, horizontal parallel lines. 2 3 4 CM PLATE 75 Type sherds of Valdivia Incised, Motif 3, horizontal parallel lines. PLATE 76 Type sherds of Valdivia Incised, a-g, Motif 4, branched lines, h-k, Motif 5, vertical parallel lines. 2 3 4 CM PLATE 77 Type sherds of Valdivia Incised, a-f, Motif 6, network, g-l, Motif 7, zoned punctate. ^^^"w'^^l^^ - ir PLATE 78 Type sherds of Valdivia Modeled, a-e, Exterior, f-j, Interior. 4 CM PLATE 79 Type sherds of Valdivia Multiple Drag-and-Jab Punctate. b * * iff994 4H ^ PLATE 80 Type sherds of Valdivia Multiple Drag-and-Jab Punctate. f^W^  --.'- *.- 172 A'-- ?- . :. ?'%? 'itiQtf i i i i i 0 I 2 3 4 CM ,.< \ * h h, % % % ?.* & ** ^*^^4^^*f*Tr$~>i PLATE 81 Type sherds of Valdivia Nicked Broad-line Incised, single line on exterior adjacent to rim or carination. 767-841-65 22 '. f IX "XX txx ^4M:MM% I I I I I 0 I 2 3 4 C MX] PLATE 82 Type sherds of Valdivia Nicked Broad-line Incised, single line on exterior adjacent to rim. PLATE 83 Type sherds of Valdivia Nicked Broad-line Incised, a-j, l-m, o, t-u, Single line adjacent to rim or shoulder, k, ®, p-q, Associated with broad-line incision. V. j j m^r^'**uksdtar"> yK& ? - : - ----- i 'h. '4^^S. PLATE 84 Type sherds of Valdivia Nicked Broad-line Incised, a-b, d, Integration of plain and nicked incisions to form the design, c, e, h,j, Multiple nicked incisions, f-g, i, Single nicked incision in broad-line incised design. PLATE 85 Type sherds of Valdivia Nicked Rib or Nubbin, rims of Form 1. * is. xmwgj^ ? * 9 PLATE 86 Type sherds of Valdivia Nicked Rib or Nubbin, rims of Form 1. PLATE 87 Type sherds of Valdivia Nicked Rib or Nubbin, rims with nubbins, a-c, Form 2. d-l, Form 1. PLATE 88 Type sherds of Valdivia Nicked Rib or Nubbin, body sherds with vertical ribs. PLATE 89 Type sherds of Valdivia Nicked Rib or Nubbin, shoulder of Form 1 with nubbins. I I I I I 0 4 CM PLATE 90 Complete vessels of Valdivia Nicked Rib or Nubbin, a, Rare Form 1, with nubbins above shoulder, b, Rare Form 2, with ribs above shoulder, c, Rare Form 1, with ribs above shoulder, d, Form 1, with ribs on exterior rim camber. I I I L- 0 I 2 3 4 CM a-h PLATE 91 Type sherds and a complete vessel of Valdivia Pebble Polished, Variant A. a-h, Rims of Form 5-7. i, Complete bowl. 4 CM PLATE 92 Type sherds of Valdivia Pebble Polished, Variant A. ???-?~*®*^^.yt 3 CM PLATE 93 Type sherds of Valdivia Pebble Polished, Variant A, with supplementary decoration by broad-line incision, or nicked broad-line incision. PLATE 94 Type sherds of Valdivia Pebble Polished, Variant B. Wmw$ ? a PLATE 95 Type sherds of Valdivia Phase unpolished plain types, a-g, San Pablo Plain, h-q, Valdivia Plain. PLATE 96 Type sherds of Valdivia Polished Plain, a-k, m, Bowl rims. /, Jar rim. n, Concave base. 0 1 2 3 4 CM PLATE 97 Type sherds of Valdivia Polished Red. a-b, Poorly polished Period A variety, c-d, j, I, n, Polished red bowl exteriors, e-i, Tetrapod feet. k, m, o, polished or striated polished bowl interiors. 767-841-65 23 PLATE 98 Typical bases of Valdivia Polished Red. a-b, Tetrapod. c, Concave. HBBBSDBHf PLATE 99 Complete vessels of Valdivia Phase plain pottery types, a, Valdivia Polished Red, Form 8. b, Valdivia Polished Red, Form 6. c, Valdivia Polished Red, Form 11. d, Valdivia Striated Polished Plain, Form 4. -V A V'iH JflHS ; , .. 3 V i -® ''?'*4'*?Km*Mm ?'^ p 1 I 0 1 2 CM CCiA.CCcco PLATE 100 Type sherds of Valdivia Punctate, a-n, Technique 1, circular, o-r, Technique 6, "pinpoint." e* -*V®.a-f*t^|Mfe, >v**¯ - *'- --'' v;'' ** PLATE 109 Type sherds of Valdivia Rocker Stamped, a-h, i-k, Single row applied adjacent to the rim on the exterior of bowls of Form 1 and 2. h, Double row above the carination on Form 3. PLATE 110 Type sherds of Valdivia Rocker Stamped. Supplementary decoration is made by broad-line incisions. PLATE 111 Type sherds of Valdivia Rocker Stamped. Bowls of Form 1-3, with additional decoration by broad-line incision. PLATE 112 Type sherds of Valdivia Rocker Stamped. Neck or rim exterior of jars of Form 4 and 5, some with additional decoration by incision or nicked rib. 2 3 4 CM . PLATE 113 Type sherds of several Valdivia Phase decorated pottery types, a-e, Valdivia Shell Stamped, Motif 1. f-i, Valdivia Shell Stamped, Motif 2. j-k Valdivia Shell Stamped, Motif 3. l-o, Valdivia Zoned Incised, Technique 1. p-r, Valdivia Zoned Incised, Technique 2. s, Valdivia Zoned Incised, unclassified. 767-841-65 24 I 1 I I I 0 I 2 3 4 CM PLATE 114 Type sherds of Valdivia Zoned Incised, a-p, Technique 3. q-u, Technique 4. 2 3 4 CM PLATE 115 Type sherds of Valdivia Striated Polished Plain. PLATE 116 Sherds of Valdivia Phase types from sites of the Machalilla Phase. a, Valdivia Applique Fillet, b-c, Valdivia Brushed, d-e, h, Valdivia Broad-line Incised, f-g, Valdivia Corrugated, i, m-n, Valdivia Incised, j-k, Valdivia Pebble Polished, o, Valdivia Pseudo-Corru- gated, p, Valdivia Punctate, q, Valdivia Polished Red, concave base. ?  m y. '"? m. ??- Sfc. I m i* rii m ? ' ' 0 I 2 3 4 5 CM dd ee PLATE 117 Stone figurines of the Valdivia Phase, a-q, Palmar Plain type, r-ee, Palmar Notched type. PLATE 118 Stone figurines of the Valdivia Phase, a-j, Palmar Incised type, k-r, Stone variant of the Buena Vista type. bo w a o ?rt O .y _>> , 3 ?si *8 o "ø -3 O 3 s * be cs œ o ?"T2 V " bo o .>> cS D fc "3. e a cS o CO O "So 2 !-. > <* > cS hn nes ion s- CS 3 be -a G <œ3 3 - .2 CJ oil 0 u a. JD a 3 V 3 M O 0 TD vN.v > be Ji* -1 O be œ I B ** bo CO +J .5 ®i _o be c cS tu u cS > bo c > o j= > &H PLATE 123 Figurines of the Valdivia type, showing variation in size, workmanship and stylistic detail. a-bb,ff-ii, Single headed, cc-ee, Double headed. o Q. 3 be U3. be a 2 I w -S 3 cS Pn 0 4 CM ? --???-::; i ..-.- -?',œ-.. -* ^&E  ,??*- v¯# rife i 9¯^l PLATE 127 Large grinding stone from G-l 10: La Cabuya, a Period C site of the Machalilla Phase. n u i-i-.-i 0 I 2 3 CM PLATE 128 Natural pebbles used as polishing stones from Machalilla Phase sites. PLATE 129 Worked shell from Machalilla Phase sites, a-p, Cut pieces, probably blanks for pendants or fishhooks, q, Bracelet fragment, r-s, Per- forated fishhook blanks. 767-841-65 25 PLATE 130 Shell fishhooks from the Machalilla Phase, showing range in size. PLATE 131 Type sherds of Ayangue Incised, Motif 1, triangles. PLATE 132 Type sherds of Ayangue Incised, a-h, Motif 1, triangles, i-y, Motif 2, slanting parallel lines. *."mM7Jy^¯^'':-'¯i ¯®¯*Jwi^fiV-~.~-.-.* Sl2 g PLATE 174 Jomon and Valdivia Phase sherds showing similar overall texturing by brushing or shell scraping, a-b, e, g, Natsushima. c, Todoroki. d, Ishizaka. f, Sobata. h-j, Valdivia Phase sites. PLATE 175 Jomon and Valdivia Phase sherds with patterned overall shell scraping, a-b, Todoroki. c, Natsushima. d-f, Mito. g-i, Valdivia Phase sites. PLATE 176 Jomon and Valdivia Phase sherds with similar decoration by finger grooving, a, Ataka. b-d, i, Izumi. e, Todoroki. f, Natsushima g-h, Mie. j-l, Valdivia Phase sites. *  ~au PLATE 177 Jomon and Valdivia Phase sherds with excised decoration in similar technique and motif, a, d, Nampukuji. b, Nishi Ataka. c, e, Izumi. f-i, Valdivia Phase sites. '^V^ PLATE 178 Jomon and Valdivia Phase sherds with similar types of decoration, a, Oomiyama. c, Ataka. d, Shirahama. b, e-g, Valdivia Phase sites. PLATE 179 Tomon and Valdivia Phase sherds with similar decoration by rocker stamping, a, c-e, g-h, Kitashikawa. b, Moroiso. /, Todoroki. i-n, Valdivia Phase sites. PLATE 180 Jomon and Valdivia Phase sherds with shell stamped decoration, a-e, g, Todoroki. f, Sobata. h-k, Valdivia Phase sites. PLATE 181 ]omon and Valdivia Phase sherds showing similar lobed rim treatment, a, c, Ataka. b, e-f, Izumi. d, Mie. g, Napukuji. h-m, Valdivia Phase sites. 767-841-65 20 PLATE 182 Jomon and Valdivia Phase sherds with folded-over plain or finger-pressed rim treatment, a, Mie. b, Iwasaki (after Kidder, 1957, pi. IV-4). C} e-h, Izumi. d, Kasugacho. i-l, Valdivia Phase sites. JJrTsT id}* *JL' $&œ **8' - ^V #fe-. i^*¯~*-v. ] PLATE 183 Jomon and Valdivia Phase sherds with similar fine-line incised and drag-and-jab punctate decoration, a-d, g, Mito. e-f, h-i, Valdivia Phase sites. ?7':-n PLATE 184 Tomon and Valdivia Phase pottery with similar relief decoration in the form of a stylized face, a, Okubo. b, Ooi (after Esaka, 1960, fig. 191). c, Omori. d, Valdivia Phase sites. *3!&&*r -5^C.i . i 3 -. * ? \ ' ??' PLATE 185 Jomon and Valdivia Phase sherds with similar decoration by cord impression, and Jomon vessels resembling Valdivia Phase examples. a, c-d, Ubayama (after Groot and Sinot6, 1952, pi. X-3, 7-8). b, Ichioji (after Kidder, 1957, pi. VI-5), e-g, Valdivia Phase sites. h, Oomiyama, maximum diameter 17 cm. i, Shimotakaido, height 12.2 cm. (courtesy of Meiji University). a b PLATE 186 Jomon vessels of typical Valdivia Phase shapes, a, Nakahosawa, mouth diameter 21.5 cm. b, Sobata, mouth diameter about 11.5 cm. 0 $' 3 CM PLATE 187 Pre-Jomon and Valdivia Phase stone figurines, a-f, Kamikuroiwa (courtesy of T. Esaka). g-j, Valdivia Phase sites. * liGHMBiflQi ? ' * < J.'l?* f t i* V' I PLATE 188 Decorated sherds from Puerto Hormiga. a-c, Finger grooved, d-e, Multiple drag-and-jab punctate. /, Shell stamped, g-h, Drag-and- jab punctate, i-k, Incised. /, Zoned parallel hachure (photos courtesy of G. and A. Reichel-Dolmatoff). PLATE 189 Jomon and Valdivia sherds with similar decoration by nicked and finger-pressed applique ribs, a-e, Kasugacho. f-i, Valdivia Phase sites. um fr "&'?+?$;, & PLATE 190 Jomon and Valdivia Phase sherds with similar nicked broad-line incised decoration, a, Katsusaka (after Kidder, 1957, pi. 1-5). b, Togarushi (after Kidder, op. cit., pi. V-4). c, Kasugacho. d-f, Valdivia Phase sites. mmmmmmm ? - "®-? ? ' I *>e?y "" '* "i ???¯ itimwrii¯ .. 111 ffHt m m q 'iilriihihiHkr. ipHIH I MM ? .- ? ¯ -tjir._ i ateo PLATE 191 Decorated sherds from Kotosh. a-b, Tiered jar. c-e, Zoned elongated punctate, f-k, Zoned circular punctate, l-m, Ring and dot and dot ended incision, n-s, Zoned parallel hachure. (After Izumi and Sono, 1964, pis. 52a, 62b, 76b, 77a, 83b.) o o CD - CVJ - o CD 3 cm pa p, ^ I > "O * * ' r  ø ¯tt< &U* I %, ^ 4" o o CvJ O *- -?-' O ;?* < o o - *- CVJ O i > pq ? CD ** eq c^ ?st- un W 1 H o < J ro" Ph 1 > PQ o tf H *%i.5 I 1* O O CD  O J3 CS h >^ LT, y be C\ Ch c - o CD HH ^9 a 1 OT H > r j < ® ^ 1 Ph Burial 8. d, bone, /", BV- o 0. , 4ft II* >* ¯ ^ .x. *~r * i y#0 U.S. GOVERNMENT PRINTING OFFICE i 1663